The metal oxide varistor is dual purpose here: it's used to prevent the HV spike kicking back into the circuit, and it limits the voltage of the generated HV pulse, should the lamp fail or gone missing. When ignition happens, the voltage across COM and N would be the peak of 240V plus the HV pulse of the unloaded section of the ballast winding (should not be more than 2-3 kV). The varistor in combination with the 120k resistor puts a slight load to the induced voltage spike, it therefore decreases the voltage but increases the duration of the pulse. The 220k resistor in combination with the 180nF capacitor enables the triac to turn off for sure by dropping the current below the holding current threshold within at least 20ms. So this ignitor wouldn't work very well outside the 50/60Hz range.
Well, for the yesterday answer I was the happy owner of a shaddow ban. I don't know why but that is done. I repost the yesterday answer in quotes: "The main and lamp connections for the igniter are reversed. The big capacitor make a brief connection to ground thru the triac shunting the ballast to ground. The 220K+120K and the current from the capacitor are not enough to mantain the triac in conduction so after a short period it will disconnect the capacitor from ground. This will create a voltage spike that will ignite the sodium lamp. Think like a boost circuit with ballast as inductor (essentially the ballast is an inductor). After lamp ignition the voltage on the divider side of the diac is lowered by the big capacitor under the trigger level. The divider in the triac gate rise the voltage of triggering over the lamp working voltage. That is what i see in the schematic. BTW, the dividers are there to assure that the firing is on the maximum level of the sinosoide to have the maximum chance to ignithe the lamp. PS: keep the good work Big Clive and give us, tinkerers, food for brain. :). Sorry for any misspeling, english is not my primary language." I'm adding the mention that with the term "divider" I'm reffering to the resistor dividers on the both side of the diac. I hope this post will not be shaddowed. PS1: for the moment I can reply to others posts. I'ts better than nothing. I can't make new answers, those are shaddowed by default... strange YT...
@@The_Stressmaker Perhaps you posted some derogatory comments somewhere on RUclips. The system seems to take offense to the slightest opinions that are made by people that it deems offensive, and it's quite sensitive, just like many of the snowflakes in the population today. Let's see if THIS comment gets shadow banned because I offended the bot.. 🤷
@@davelowets That comment was the first and only comment from this account in about 11 (eleven) years. Talk about minority report and preemptive strike. But the question remains... why i cannot initiate a new post but i am able to respond to another post. Maybe the answer is in my username or simply that the YT security bot is so stressed that a simple username like mine is triggering all the red flags and alarms... BTW, thanks for the comment.
Ballasts are definitely more expensive, I do lots of repairs to 2KW sports pitch lights. A 2KW 415V ballast costs about £230, an ignitor is around £35.
The LED's should be nearer the red spectrum they don't need to be as bright white as they are the disturb wildlife don't allow for proper rest patterns..
Some LED street lights I've noticed can be horrible. Not just the brightness as you travel from dark to light and light to dark but also the strobe effect can be quite pronounced if the council skimped on lights that don't have smoothing, this makes it look like objects are flickering as they move which is disconcerting
When I saw this video, I immediately went and got my Philips SX 26-01 ignitor (for a SOX-E 18 lamp) for reference since it looks similar, however, it has more components and only lamp and neutral connection. I can see a blue capacitor with the X2 marking on it, and the big blue one with 1000V, and the yellow "mov" component says "275V 593-8C 0101". It also has a triac BTA208.
There are also ignitors with HV transformers inside but these are usually completely potted. These are connected between the ballast and the lamp. There are also ignitors with a timer which stops the ignitor after few minutes of continuous striking.
Some time ago when the city was replacing almost all the street lights from sodium to LEDs I was able to snatch a complete set (ignitor, capacitor, ballast and lamp holder with reflector) and quite a few lamps
Brilliant Clive I’ve never seen the ignition unit before, I have had a few old car park sodium lights they were bright as hell, never opened them up though 😊
It turns the bulk of the ballast inductor and turns the coil into a HV pulse transformer of about 6kV output that ignites the lamp. Seen several schemes for ignition of these lamps as well.
High pressure and low pressure sodium lamps are a fascinating design, and many of the bulbs have instant restrike, if they are shut off hot. The difference between the two is differences in color rendering, although neither of them are a pure white. Metal halide, on the other hand, is a better option, particularly for sporting events, and the like. However, the restrike time may take as long as 15 minutes. Sodium lights have one advantage to flora, and that is that it doesn't stimulate plants to grow at night like MH and MV lamps tend to do. However, our eyes perceive the light to be brighter than MV lamps.
the eyes of us humans are designed to work with sunlight. sodium lights might produce a lot of light but because it is not the same as sunlight they are effectively less efficient. its to do with the rods and cones inside our eyeballs.
MH also has the disadvantage of shorter bulb life. "only" 10k hrs instead of north of 20k in many HPS. This, and a modest advantage in efficiency for HPS, mark the major reasons why standard issue street lighting is (or, sadly, was) HPS. The restrike situation was addressed in some MH models by having a secondary lamp in the fixture that would power up while waiting for the main MH to cycle. My work did not have these, and chose to replace their high bay MH with HO fluorescent. That they are now replacing with LED. (Yes, I grabbed both an MH and HO-F fixture.) I cant speak for MV, though the same research that praises LEDs suggest that they would also have this advantage, and could even have favorable photopic efficiency (accounting for varying color sensitivity at night) than LED. I don't know for sure though, because many places have banned MV due to its high mercury content. I wouldn't mind owning one though.
@@FerralVideo , i think anything with mercury needs to be avoided. when schoolchildren 'spill' it over benches to watch it run they (read me) are actually ingesting vapors of the stuff. i dont understand how such a heavy metal can evaporate but it seems it does.
Interesting what you say about instant re-strike. I often wondered why they didn't do it anyway - all hot and vapoury. I don't know, but i am guessing green light wouldn't stimulate plants much - and may even be a useful colour for people to tend plants at night without disturbing them - but allowing them to see what they are up to! It's curious. i used to think that the low pressure sodium lamps were used because our eyes were most sensitive to that yellowy colour - and in fact of course, it was just a lot of lumens per watt at the time. It was only as I started playing increasingly with lasers where the power output is measured only in light OUT and nothting to do with power IN that i realised human eyes are optimised in terms of sensitivity to green. A 523nM green laser at 1mW APPEARS brighter than a 10mW red laser at 680nM As such you get the slightly odd situation where a bright green laser SEEMS more dangerous than a much "dimmer" red laser but in fact, assuming output power is equal, the extra perceived brightness just triggers blink reflex much faster and makes it SAFER. Indeed you can get an almost invisible 808nm "pump" laser diode and if youmake the grim mistake of looking into it, you can see (just the once) a deep dark cherry red glowing dimly. Sadly blink reflex isn't triggered and you are down to just the one working eye. It's why a "white light laser" usually has a MUCH stronger RED and BLUE diode than the green one. Eg: 3W Red 1W Green 6W Blue
Most ballasts in the US are multi-tap ballasts with autotransformers (HX and CWA) that usually have 120V, 208V, 240V, 277V, and sometimes a 480V tap as well. Some ballasts can also be reactor type ballasts for a single voltage, although they are rare.
Reactor/choke ballasts for high pressure sodium lamps are very common as 35W to 100W as well as S55 150W HPS lamps have arc potentials of 52V to 55V and thus 120V chokes are a simple and efficient solution.
@@ThornBeta5 Problem is low pressure sodium lamps require skilled lampmakers and are thus much harder to offshore to where they can be made by child slaves.
I do like how you explain how the light woks. I do prefer the high and low pressure sodium bulbs as the (LED lights are to me (Bloody useless) as the light output is basically crap. You can still get the bulbs even though they are (ES 40 ) (Edison screw) cap I would like to get hold of (Electronic igniter ) to power the bulbs as (Son - E) have internal igniter. While (Low Pressure ) (400 / 600 / 1,000 Watt) do not have internal staters. Very good Clive keep up the good and outstanding work you do.
An important terminology distinction for American/Canadian ballasts is that common always refers to the connection common to the supply, ballast, ignitor (if used for that lamp type) and lamp. This will by convention be the neutral if the fixture is fed by 120V, 277V, 347V or less commonly single hot 240V (which we have in corner grounded 240V delta and 240V/480V single phase 3 wire services, the latter used by some municipalities and utilities for lighting services). The lamp connection is always marked "LAMP" on ballasts where the lamp connects directly to the ballast.
@@Choukai_Kai Multitap ballasts are actually very common for mercury, metal halide and S66/200W and higher high pressure sodium ballasts. The most common multitap ballast type is 120V/208V/240V/277V (commonly called 'quad tap') which covers every common lighting operation potential. 5 tap 120V/208V/240V/277V/480V ballasts are also common in fixture repair kits as that way a single ballast will work in all cases. Another common one is dual tap 120V/277V ballasts, which are usually lower power as they are meant for applications that would use a 100W or 200W 120V or 277V GLS lamp; all of mine are 50W S68 HPS or 50W M102 MH. The Canadians also use 120V/277V/347V tri tap ballasts on single hot lighting circuits, as a regulatory requirement there is that the shell of the lampholder must be at or near ground potential; fixtures there connected in single phase delta (208V, 240V, 480V or 600V) or to 240V derived from a 120V/240V single phase 3 wire supply use isolated secondary ballasts. The Mexicans use 127V/220V/254V/277V quad tap ballasts as Mexico uses 127V/220V and 127V/254V as standard 250V class services. The weirdest multitap ballasts I've seen are 120V/220V/240V 50cy/sec ballasts, which from what I can tell are mainly for export fixtures and overseas military installations where everything is wired to American norms but only 50cy/sec grid electricity is available.
It reminds me of a dimmer circuit. I think I've seen people drive car ignition coils from mains voltage in a similar way. It's like a capacitive discharge ignition driver
It's amazing how for so long we've used lighting for large areas that was basically a trapped bolt of lightning, causing the gases inside a capsule, tube or bulb to glow & emit the UV light needed to stimulate phosphors, makes you wonder what lighting technology could come once LEDs become just as old... :\
Particulary for street lighting , after HPSV & Mercury Vapour lamps came the " INDUCTION LAMP " these worked on 2-3 GHz freq. hence did not need physical any connection.. energy transfer to light the lamp was through an induction coil encircling the lamp in the middle.. Dimmable down to (0-1)% ..and had a continous run time that could allow operation upto 40+ years without any loss of brightness over that time span.... Development work was still on to bring down controller & mfg. cost..but LED's development came along a lot faster & cheaper.....
I was literally just running out of quality content to watch on RUclips, and up you popped Big Clive, just in the nick of time! EDIT: Apparently, we're all going to go blind (no, NOT because of that), because of blue light exposure from screens and LED lighting! 😎
Some of the first organisms that evolved in the oceans of the earth possessed a very clever survival technique: Blue light is one of the last visible light wavelengths to vanish when sunlight strikes the surface of the ocean. These micro-organisms hide in deep water during the day and move up to depths closer to the surface once the blue wavelength vanishes from the water. A cue for these micro-organisms that it is safe to feed. This trigger exists in the human DNA even today and can affect our behaviour. Whilst there is still blue light around we are tricked into thinking that it is still day time. Many computer screens emit blue light so staring at a screen can lock us into this "day-time" mode and prevent us from wanting to go to sleep. This is a source of fatigue and we dont even realise why we aren't tired and yearning for sleep and rest. There are screens and eye glasses that block the blue wavelength light in order to limit this effect.
@@PetraKann I have a blue light filter on my Tablet set to maximum. It turns the screen a sort of sepia colour. I also take a Lutein supplement, which helps protect the eyes.
@@marcse7en Blue light doesnt actually damage your eyes. It tricks the brain into thinking it is still daytime and so our instinct to sleep and rest is delayed. AN old evolutionary biology and adaptation that occurred over a billion years in a marine micro-organism. That genetic chracteristics is still with hominid species today and can be triggered in people that are exposed to the blue light wavelength for extended periods of time. So it's not quite right that Blue light directly makes us tired through our eyes. Casinos and other entertainment establishment can and do exploit this by ensuring blue light wavelength is plentiful and intense at gambling machine screens - keeping people alert for longer periods in order to gamble (if they still have any money left). There are lots of examples in the Human DNA where ancient genes are still present and can be dormant and/or triggered by certain cues in the environment. I have blue filter treatment on my glasses when working in front of a computer screen. Also a good idea to be exposed to natural light every day (not only for your Vitamin D) Big Pharmaceutical corporations are the most corrupt and devious organisations on the planet - they make the mining industry and the banking cartel appear like angels in comparison. CHeers
@@PetraKann , there is some research which seems to indicate that this blue light you are suspicious of may help alleviate depression in high latitudes. a replacement for sunlight. i am one of those who feels depressed from not enough sunlight.
@@vsvnrg3263 That research wouldn't surprise me. Many phenomena can have both beneficial and detrimental effects on the human body. Usually it's a matter of balance. It also varies with each person's sensitivity. What do you mean by High Latitudes? I know here in Australia there were huge health issues during the Pandemic lockdowns due to a lack of sunshine (Vitamin D) and just a lack of fresh air etc. Cheers
There are three common types. A parallel two wire type which goes across the lamp and produces a 750v kick, usually for metal halide lamps designed for use on mercury vapour ballasts. Semi-parallel type which is depicted here which is favoured by Philips that is usually good for 3-4kv and then the more common super-imposed type which is good for 5kv. I have plenty of spare super-imposed types I could send but they are all potted!
When I was in the diecast model trade, dead streetlight balasts made great heavy haulage truck loads, after painting them red oxide primer. I got them for nowt off a customer that worked for the council
I am puzzled with the two components you mentioned too, but i had wondered how the 1KV capacitor charged - what it was referenced to. I could see its Live connection via the inductor, but where is its ground if not via the triac, other than the two oddly placed components. i admit I am still puzzled.. Also I love the less efficient but oh-so-simple self ballasted lamps which are available in 160W and 400W flavour. I mention this for the interest possibly of others as I am sure you know more about it than me, but instant light is provided by a filament which also heats the discharge tube and then a bimetallic switch puts the discharge lamp in series with the filament (IIRC). A bit more epensive but they were awesome easy to deploy quickly where you just needed lots of light with minimal kit / weight / effort. E40 bulb holder and a 400w self ballast and off you went - the 160s just plugged into a regualr B22 socket :) Not as efficient I guess as a properly ballasted lamp, and a bit dearer - but for some applications, that didn't matter too much. :)
I believe you may have swapped the Line and Neutral connections to the igniter module. The only connection to the lamp is the ballast (there is no connection to the igniter module here). The ballast also functions as an auto-transformer. The traic is connected to the centertap via the 1kV capacitor and fuse, and to Line on its other side. The triac gate drive is also connected to Line. The 120k resistor is connected to Neutral. When trying to ignite the lamp, you need a high voltage pulse across the lamp (several thousand Volts, more when the lamp is hot or old (not cold)). The ballast, functioning as a boosting auto-transformer can produce that pulse by discharging the 1kV capacitor (charged to near peak line voltage) through the low impedance 220V/240V ballast winding. The ballast, functioning as an auto-transformer, boosts the discharge voltage to the thousands of volts need to initially beak down the sodium vapor lamp (striking voltage) to form a low resistance arc. As the lamp warms up, the arc resistance continues to decrease, but lamp current is limited by the ballast now functioning as an inductive choke. How does the capacitor discharge across the 220/240v winding? The triac (and its gate drive) should be connected to Line rather than Neutral. When the triac fires, the capacitor is discharged through the low impedance side of the auto-transformer. The high impedance side of the auto-transformer develops the high voltage pulse. How does that capacitor charge? The 120K resistor should be connected to Neutral. (There are no other connections to the lamp / ballast node.) The capacitor charges from Neutral through the 120K resistor, the 220k/thermistor pair, and fuse to near peak line voltage. When the PTC thermistor gets hot, its resistance rises, reducing charging current and the capacitor voltage. You can estimate the capacitor charging current by assuming the thermistor is either open when hot or shorted when cold. The triac is triggered near the peak AC line voltage by the diac breaking down when the 200nF capacitor has almost reached peak line voltage. This circuit attempts to ignite the lamp every half cycle. However, due to the thermistor's high resistance when hot, the resulting pulses become weak after a few seconds and the lamp will fail to ignite until the igniter has cooled.
I'm super fascinated with this tech, and i try to collect as many different variations of MH, HPS and Hg lamps as possible. Thus i would love to see more about these things. I have MH/HPS electronic ballasts that aren't potted, as well as quite a few ignitors. None of the ignitors i have looked into contain any HV transformer, only the "electronic ballasts" However, the circuit for the electronic ballasts (CMH at least) is very strange and foreign to me, first time i saw it, i was deeply perplexed because it looked like no other electronic circuit i had ever before seen, but for some reason, the CFL electronic ballasts doesn't have the same overall design (lower ignition voltage i suppose) It is also interesting how different the NA design is to the EU one, because the lamps operate around ~100v, it is actually pretty difficult to run them on a 115v supply, so the NA design is typically much more elaborate. Fun facts for the nerds out there: CMH exists all the way down to 20w (HPS down to 50w) There is a warm-white sodium lamp, although quite niche, they are known as SDW. SON(HPS) is still to this day, more efficient in terms of lumens per watt than most LEDs, especially when factoring in investment costs - this makes you wonder about the math behind switching from HPS to LED street lighting.
Heh, that makes two of us. My collection is in its infancy though. I have one stray MH ballast, three MH fixtures, and five HPS lamps. HPS: Four little 35w wall packs ($5 ea at yard sale), one 70w flood. One 35w lamp had a bulb that broke inside (arc tube cracked). MH: Two High bay fixtures, multi-voltage, 400w, probe-start. Heckabright. The stray ballast is a weird one. Maybe you can help me here. It's labelled 70w, but the ANSI code is for 100w bulbs. Which size of bulb should I put in it? Can I safely run 70s? I could install it in the 70w HPS flood fixture if I wanted to. One porch light of unknown but highish wattage. The markings on the bulb came off years ago. Looks like a 250. I wouldn't mind adding an Hg lamp or a few to the collection, and a companion in a group I follow is actively trying to source one. Both of my 70w are multi-tap magnetic ballasts with an electronic ignitor unit and a PFC cap. The 35s are single voltage 115v with a simple coil and the electronic ignitor and cap. I'd have to pull one apart again to remind myself exactly how they're wired - I think it might actually just be an inductor. The high-bays are probe start, no ignitors. The one 35w HPS that failed was actually really interesting, in that the ignitor was able to strike an arc across the leads going to the discharge tube near the socket, creating a faint blue glow there. I didn't run it long in this state as I didn't want to damage anything. As for the switch to LED, the thing that drives me the most bonkers is that most discharge lamps were designed to be easily serviced. LEDs are explicitly designed to be unserviceable, despite having comparable service life to one or two HPS bulbs if not overdriven. (You watch BigClive, you know that they're all overdriven) They say the switch is because of luminous efficiency. The spectrum that an LED produces is more in line with our eyes' sensitivity than HPS, resulting in more *visible* light efficiency. I question this, but this comment is already long enough. I'll leave off by just remarking that I wouldn't be quite so bitter about them if they weren't explicitly designed to be thrown away complete at EOL.
Maybe the varistor/resistor combination discharges the 220nf capacitor in the timing circuit. That would reset the capacitor voltage so the timing is consistant. ?
It works like this: When the triac triggers, it formes an autotransformer with the tapped inductor )AND a resonant circuit with 180nF and input part of ballast inductor from L to fuse. That will oscillate until current drops below triac holdoff current. Triac does NOT turn off instantly! (and high dv/dt triggers it too and high di/dt keeps it conducting!) It needs some time with low current to turn off. Not so obvious. Look more closely to datasheets to see cutoff current and period it is needed. Period was not documented on couple random datasheets I checked, so check several to see an example. Because triac has a high frequency ringing current (L-C resonant from part of ballast and 180nF), it won't turn off until current has dropped low and been low for a bit of time ie. ringing faded. (at 50Hz it may seem instant but it is not) Then either the lamp has started with high voltage pulse (high current ringing in 240-220 part of autotransformer) and the high frequency current faded (been below cutoff current for cutoff time ie. triac not conducting) and you got lamp on keeping it from re-triggering. If lamp did not start the cycle starts over by charging 220nF and 180nF for a new trigger. I think NTC probably is part of charging circuit and perhaps needs to cool off before the system can trigger? You cannot trigger a hot lamp with a few kilovolts, you need some time to cool down the lamp before starting to ignite it. (Did not analyze thoroughly, it could also be part of 180nF charging circuit) I think 220nF has the resistor in parallel only as a bleader. Basically it is a 120k + 220nF network charging that triggers the triac. Quite a complex thing to figure out unless you know the details and can also think about the resonances (that also make it have more than mains voltage at tapped section due to LC resonance. Peak voltage at inductor is multiplied by Q of LC circuit). So, you get out something like Vinitial(in 180nF) * Q * ratio_of_autotransformer = usually several kilovolts.
For mere mortals 593 series varistors are only available between 30V and 550V but if they used the same logic for a custom voltage it should be: 593=9 mm diameter 9=something related with packing 00=RMS operating voltage multiplicand (100) 1=RMS operating voltage multiplier (x10) 6=something related with leads shape/size/separation So it's probably a 1000V varistor.
With a ( redundant ) 400 Watt SON fitting , the ballast and 40 μF capacitor can be reused to run a 3 phase bench grinder from a single phase supply ... connect neutral to 1st phase , ballast to 2nd phase , cap to 3rd phase , then common the other side of cap and ballast and connect this to single phase live ... hey presto , grinder is now running well ! ... ( tried - n - tested ) ...... DAVE™ 🛑
The MOV is shorting out the inductot when the triac fires. You have an RLC loop with the 180 nF cap, the ballast, and then 120kΩ resistor. By shorting out the 240kΩ resistor, it limits the voltage seen by the semiconductors.
The 220k seems to be the bleeder resistor for the 180 nF when the power is off. But is the thermistor a PTC or a NTC? NTC for a soft start, I would guess. BTW In the US most commercial lighting is 277 VAC in order to use smaller gauge copper wire and save $$$ Thanks, Clive. 👍
I thought it was a thermistor at first - maybe as a simple timing function to shut off the ignition spikes after a suitable time, but it appears to be a voltage dependent resistor.
The main and lamp connections for the igniter are reversed. The big capacitor make a brief connection to ground thru the triac shunting the ballast to ground. The 220K+120K and the current from the capacitor are not enough to mantain the triac in conduction so after a short period it will disconnect the capacitor from ground. This will create a voltage spike that will ignite the sodium lamp. Think like a boost circuit with ballast as inductor (essentially the ballast is an inductor). After lamp ignition the voltage on the divider side of the diac is lowered by the big capacitor under the trigger level. The divider in the triac gate rise the voltage of triggering over the lamp working voltage. That is what i see in the schematic. BTW, the dividers are there to assure that the firing is on the maximum level of the sinosoide to have the maximum chance to ignithe the lamp. PS: keep the good work Big Clive and give us, tinkerers, food for brain. :). Sorry for any misspeling, english is not my primary language.
I've always wondered why this tech never used slayer exciters to start. I've had a 160 watt MH lamp light up 200 cm away from a cheap plasma globe. Maybe it's reliability? Semiconductor electronics don't like heat, as I'm sure many of the dead LED street lights can affirm. Also, I've had ideas about a auto variable ballast that slowly turns up the current over time to squeeze more life out of the lamps. I think that's why they cycle indefinitely at their EOL stage? The electrodes recede to the point where the current required to sustain the discharge is above what the ballast supplies. Is that why, or am I missing something as I don't have one of these fixtures to probe around on?
The company I work for has just swapped to PIR led lights throughout, trouble is they turn off after 60 seconds so they often turn off mid job And I now have to work a 170 ton press with only 150 lux of illumination 😎
I'd imagine that the MOV connected that way is some sort a transient clipper. The MOV turns on, and in doing so, forces the TRIAC to fire. The firing of the TRIAC forces the transient to be shunted to common. This would move the brunt of the transient clipping to the TRIAC which should lengthen the life of MOV, while TRIACs don't seem to care much about high currents or peak power dissipation.
Sodium lighting is appaling, the amount of Hangars I've worked in and you couln't see bugger all. The funny thing is though, when at RAF Valley in Gaydon hangar. We were told to wear bump caps due to the sodium bulbs that had been replaced bursting and showering glass everywhere. Not sure about this, but the lads in the hangar said that the workmen had replaced the wrong voltage bulbs.
the eyes of humans are designed to work best under daylight sort of lighting, not sunset sort of lighting. its got to do with the rods and cones within our eyes. i also hate sodium coloured lighting.
When the triac conducts, the capacitor basically grounds that "20v" winding, viciously reversing the existing magnetic flux in the core, with obvious results.. The MOV is likely a clamp in case the lamp isn't fitted.
MOV is going to trigger the triac on overvoltage, to get the triac conducting, making the sharp voltage gradient to start the lamp. Diac charges to threshold if lamp does not strike, as it gets to the peak of the mains, not the normal 100VAC or so of the running tube. Then triac fires, shorting out the reactor for the remains of the cycle, and storing voltage in the capacitor. On nect half cycle at mains peak the voltage is enough across MOV to start it breaking down, and thus turn triac on hard, shorting capacitor and dumping charge into the reactor, making a series LCR circuit, with a high voltage at the far side of the coil, as current is flowing through the coil. Coil rings at least a few times, and lamp should strike a little, next cycle it charges capacitor again, and then triggers another set of rings, each time heating up the gas, till the lamp finally ionises and strikes, keeping the voltage on the 120k resistor low, and thus not enabling the triac to fire any more. Even on 220V the inductance of the wire, combined with the high rate of change, means that it will strike, unless you have a very short ultra low impedance wire direct to transformer, which for most lamps you never get, typically 10m of 1mm wire does provide more than enough impedance. EOL the triac fires multiple times, till it gets the fuse warm enough to melt, generally after a few hours of the current pulses, or if the capacitor goes short circuit, on the first cycle. Open circuit capacitor it just does not work.
@@LDuncanKellyFinally, something I can understand. As usual, most of what Clive says and the comments go way over my head, but your response is easy to understand, even for me.
I suspect that when the 20V coil is dumped by the triac the voltage spike in the 220v section will jump dangerously high, much higher than is actually needed for igniting the lamp. So the MOV is there to clamp the top of the spike. Just a guess though.
That's sort of my thought. But I would surmise its there to shunt the pulse around the timing network and thru the the triac itself in order to keep spikes out of the timing side of the circuit each time it fires to strike the lamp. Given the whole thing is behind a 120k resistor its probably not going to pass much current, but could be enough to cause false triggering (or even potentially damage the gate), so it and the snubber network are probably belt and braces to ensure it triggers properly. That's what I think its for, anyway.
I was told 35 years ago, that it is OH to measure the socket voltage of a Metalarc, but never try it with a HP Sodium - it will blow up your meter in your hand, Scare me enough that I never tried. Used a light probe to see if they were powered up. Metalarc were close to 600V though. (In the states & Canada)
My missus prefers the "flick the switch, witchcraft happens and the light comes on" 😂😂 I spent nearly a year playing with this kind of thing in a large factory. It was never ending!
This autoformer effect (single winding with taps) is neat - it works pretty much like a voltage divider when the supply comes to the end taps and the load involves one or more of the center taps. But when the supply involves one or more of the center taps, it can step voltage up. It is more efficient than a resistive voltage divider which is the only option for DC and only allows stepping down. An autoformer can also use thinner wire than a respectively powerful normal transformer and many voltage adaptors, especially cheaper ones, rely on the design. Most variacs are of the autoforner type and do not offer isolation from the mains. Which comes to today's adventure at work. We have a plentitude of light fixtures that were installed 15 years ago when fluorescent lights were still very common and the white LED technology was only slowly making itself known. Many years ago I bought a sh*tload of G23 tubes that fit in these fixtures - I got them dirt cheap from a traffic station store that was clearing products they were planning to discontinue. I also got a heap of windshield wipers that were old stock. But, and this is a big but. The tubes I got were mostly the 11W variety and the fixtures are designed for 2 x 9W. I can get a 11 and 9W tube in series to strike 50% of the time, luckily they are mostly on day in day out. The 11W tubes fit without the cover bur will not strike with two in series. There are no taps on the ballast to adapt for bigger tubes. What I found to work fairly reliably is a 7 and 11W tube in series. Also the 7W tubes seem to be the cheapest, must try to acquire some when they are still available. The supply is dwindling as they are apparently no longer produced, and the LED replacements for those are F****G expensive at the moment.
Where I used to work we had really bad tappings on the sun station for the factory so our high tech fluke portable appliance tester would often give an error because it was receiving 260VAC. Depending on what was running in the factory it would vary slightly but normally be 259VAC mains voltage. How would something like this effect this sort of device? Because it would be putting nearly double the voltage across the short winding of the ballast.
Hi from Brasil, I was wondering if you could do a teardown in the Salt Water Lamp Emergency Lighting Outdoor Lamp No Battery to see if this is a misleading product like it's just closing the circuit with the hidden battery using the water and salt. I did not find anyone opening this so I thought of you!
It uses the salt water as the electrolyte. The real work is done by the two different metal electrodes, that will gradually dissolve. It is not the water that creates the power.
Discharge resistors perhaps? Although it does seem strange having a varistor and a fixed value resistor in parallel like that. I don't believe they're there for protection against voltage spikes, as the varistor's resistance would drop significantly under such circumstances and surely the current would dump through either the triac or the resistor+capacitor setup depending on where it is in its cycle (probably the triac tbf), which seems bad. Besides, isn't that what the fuse is for? I'm stumped, although this is hardly my area of expertise. I'm a chemist by training. Disconnect them and see what happens. Another hypothesis: as the voltage moves from 0 to the peak of the sine wave, current will flow less via the 120k resistor and more via the 220k + var until we start moving back to 0, charging both capacitors. Then, when the 220?nF activates the triac pretty much all the current now flows through the 180nF fully charging it until we hit 0 again (or thereabouts), and then it *dis*charges on the next half of the sine wave once the triac activates and shunts the current now going the other way through it. But even then, they seem... Superfluous? Perhaps they found they needed to charge that big capacitor a little more and so they allowed it to connect to the neutral via something other than the triac to give a bit of extra juice in the upswing of the sine. But like I say, chemist not electrician.
General Electric often uses a kind of cordwood construction for their 3 wire ignitors using plastic frames with the components soldered or crimped to each other or to spade terminals for connecting to the fixture. More complicated 2 wire ignitors and universal 3 wire ignitors have a circuit board in this frame, but it isn't potted as there is usually a shunt wire to cut if the ignitor is used with 90V to 130V arc potential lamps (150W S56 HPS, all 200W to 400W HPS and all pulse start MH up to 400W). American Electric also used bare circuit board style ignitors in their HPS fixtures a lot.
I don't know if you guys have constant wattage auto transformers over there but here in the US if you have a CWA ballast and the cap fails you will likely lose the ballast too.
Clive, you're into ozone and all that. Would it be reasonable to use an ozone generator to disinfect an entire property? Basically close all the windows and ventilation ports, open all interior doors, plug in a powerful ozone generator and leave the place empty for a few days? Would this have problems with regard to ozone leaking into neighbours properties and since it's an oxidizer, would it cause damage to objects prone to oxidation such as metals?
Ozone is often used after flooding to try and remove some of the odour. It depends on the reason for wanting to disinfect the place. If it's deep seated mould then it may require furnishings to be changed. But if it's just a lingering odour then it may work. It takes a long time with continuous exposure to damage metals. Rubber materials are more prone to surface oxidation.
I was a little surprised to see no R/C snubber across the triac because it is driving an inductive load. Perhaps that's exactly what the MOV is doing. If so I would expect the MOV to become worthless over time due to experiencing repeated activation.
I think you are right about the MOV. I have seen them used in kiln controllers (where there is often some small inductance), many units last for decades of use.
Ultimately the sodium reacting with the aluminium oxide arc tube over its life depletes the elemental sodium content to a point where end-of-life tubes may strike ok from cold, but as they heat up, the internal pressure rises to a point where the voltage required to sustain the arc exceeds the supply. The lamp extinguishes and has to cool right down before the internal pressure is low enough to allow the arc to restrike. I have seen lamps in redundant areas at my plant cycling like this for 10 to 15 years without failing completely.
@@kyoudaiken That's described by the sulphur content loss as it reacts with the aluminium oxide arc tube under high heat and pressure over the years. Less sodium requires higher voltage to sustain arc, higher temps causing higher pressures results in higher voltage required to sustain arc. Age equals lower sodium so there becomes a point where the voltage required to sustain the arc exceeds the supplied voltage and the arc extinguishes, cools down, reignites, warms up, rinse, repeat.
Initially 120k & 18k and the full winding complete the 220nF charging ckt. thereby limiting the peak charge voltage across the 18k/220nF parallel combo. This ckt.is also effective in repeated restrikes if the ignitor doesn't fire the lamp... Once proprely lit , the lamp low voltage retrigger is through the series ckt. of 220v winding , F2H , 180nF , 220k and 18k/220nF ....this is useful in hot restrikes in case the lamp goes off momentarily.. In the cold startup sequence , once triac triggers , current flows through 220v winding ,completing the 180nF cap. path to return via the triac. Later when triac turns off , current must continue to flow in the same direction through the boost winding..since the HPSV lamp initial presents an open ckt. , a (5-6 )kV voltage is generated that can breakdown the potential barrier across the lamp to ignite it.
E39 (US)/E40 (EU), and you are correct. "In areas following the U.S. National Electrical Code, general-use lamps over 300 W cannot use an E26 base and must instead use the E39 base."
Starters are for preheat fluorescent lamps. The job of a starter is to provide heating current to the lamp electrodes to facilitate easy ignition as well as cause an inductive pulse when the starter opens. Integrated ignitor HPS and MH lamps often use glowbottle thermal switches like those used in fluorescent starters (just built for much higher currents), but they are prone to failing short from contact arcing.
Curious ... I have a sodium vapour lamp but it only has a ballast, no igniter.. lights just fine. I am pretty sure there is nothing in the ballast but a big fat coil, no extra parts... Are there lamps that don't require an igniter ? And if so.. why does your (that) lamp need one ?
Always open your black box Ballasts, boo, a single filament would provide it's own ballast, no components required 😅 I did once put a 150w uv center section of one of those bulbs across straight 230v,it went pop, but, didn't explode, strange! Make your own ignitor with a taser module? 👀
Maybe it’s a 90,016 volt varistor? Maybe 900.16 volts? Maybe it’s not a varistor at all; maybe it’s a metastatic boomufoomulator. 😅 Some experimentation is needed here. Could it be a thermistor that’s 900 ohms hot, and 16 ohms cold (or vice versa)? It’s a real pity, though, that this module doesn’t contain the obligatory 8-pin micro. 😢
One of the local hotels changed out all of their inefficient halogen spotlights that run through the pedway. Probably 50 new led potlights. The first day one was already flickering. Today... probably 4 months in... half of them maybe more are on ghetto disco mode. The homeless squatters are not impressed.
@@NiHaoMike64 I just walked through there and thought I would count. I exaggerated. The string of 15 had 5 flickering away. Its quite an effect. I didnt count the other strings... its 4 rows of these. I wonder if that was a bad batch or if they take other others on a circuit with all the noise/spikes.
Hmm - 4:48. I was looking at your worksurface and realised that if someone showed me that as an image; on their phone, or what-not; I would immediately be able to say, "Oh - that's big Clive's worktop". I've seen it for so long now, it's like an old friend. Have you ever considered releasing a high-definition image of it for use as computer desktop wallpaper? Or, do you not realise the familiarity we all have with it, or the nerd-niche kudos that would give? (*As always - great video - Thank you.*) [edit] Maybe, for just your own general interest, do a poll on whether people would be able to identify that desktop as being from your videos, if shown it alone. I for one ... would be interested in knowing if it's just me (and I'm weird) or whether, the community has the same response - "yup, that's Big Clive's main worktop".
I think this might be "advance sight" by those that support his channel - and freelocaders like me wait an extra week or two :) Awesome content for free seems a pretty good deal. I might be totally wrong, but that would be my guess :)
I'm deliberately building a buffer of videos for some work I'll be doing soon. During the initial load in and testing I may not be able to make videos.
As example , assuming the MOV is rated at 240( rms) , it begins short circuiting once the voltage across it goes above (900-950)V..{ typical clamp voltage for 240v mov from it's data sheet } but current is limited by the120k resistor....this allows the boost winding to generate sufficiently high voltage (5-6)kV to ignite the lamp.., circuit completes its path of current flow in same direction via. boost winding , 120k , mov , 180nF , & fuse .in that order.
Every time a MOV takes a hit, it's life is shortened. For UL 1449 listed SPDs they only have to be effective 8 times to pass the listing test.When I worked in high power/voltage electronics we never used a MOV as a snubber.
Hi Clive, another great video. Thank you. I have an eBay item that I would like to send to you for a tare-down. Can you tell me where to send it? It is a 433.92MHZ Keyfob Remote that was faulty on arrival. I have ordered this remote from different sellers at different times and they all have the same fault. Clearly, they are leaving the factory like this and they must know that these units are faulty because a simple test is all that is required to show the fault. If you could find the fault, that would be great. If you could find a fix for the fault, I am sure hundreds, if not thousands, of people, would be grateful, including me! Cheers Clive 🙂
@@simontay4851 ....I'm afraid that this matter no longer resides in my hands. Scotland Yard are involved and most likely the UN will be dragged into this rather unfortunate international incident. 😁
I think you've drawn this wrong. Shouldnt the F2H & the cap connect across the bulb, not the ballast tapping? It isnt making much sense to me otherwise.
The triac is in the ignitor circuit that's fired up/triggered only when lamp is to be turned on.....& that's most likely once a day..... exception being repeated strikes for retrigger attempts in case of non ignition. ..also these triacs , most likely , will be of high reverse breakdown volts( 2-3) kV , so snubber may be ineffective .
You can use this to set light to somebodys son? 😳 (the symbol for Sodium is Na 😉) Of course, Sodium lamps are also a type of metal halide (usually NaI, Sodium Iodide aka Anayodin). 🙂👍
Am I the only one that liked the sodium lights? Sure, they were crap for color reproduction, but seeing the sea of amber across the cityscape or highway was kinda cool.
The metal oxide varistor is dual purpose here: it's used to prevent the HV spike kicking back into the circuit, and it limits the voltage of the generated HV pulse, should the lamp fail or gone missing. When ignition happens, the voltage across COM and N would be the peak of 240V plus the HV pulse of the unloaded section of the ballast winding (should not be more than 2-3 kV). The varistor in combination with the 120k resistor puts a slight load to the induced voltage spike, it therefore decreases the voltage but increases the duration of the pulse.
The 220k resistor in combination with the 180nF capacitor enables the triac to turn off for sure by dropping the current below the holding current threshold within at least 20ms. So this ignitor wouldn't work very well outside the 50/60Hz range.
Well, for the yesterday answer I was the happy owner of a shaddow ban. I don't know why but that is done. I repost the yesterday answer in quotes:
"The main and lamp connections for the igniter are reversed. The big capacitor make a brief connection to ground thru the triac shunting the ballast to ground. The 220K+120K and the current from the capacitor are not enough to mantain the triac in conduction so after a short period it will disconnect the capacitor from ground. This will create a voltage spike that will ignite the sodium lamp. Think like a boost circuit with ballast as inductor (essentially the ballast is an inductor). After lamp ignition the voltage on the divider side of the diac is lowered by the big capacitor under the trigger level. The divider in the triac gate rise the voltage of triggering over the lamp working voltage. That is what i see in the schematic.
BTW, the dividers are there to assure that the firing is on the maximum level of the sinosoide to have the maximum chance to ignithe the lamp.
PS: keep the good work Big Clive and give us, tinkerers, food for brain. :). Sorry for any misspeling, english is not my primary language."
I'm adding the mention that with the term "divider" I'm reffering to the resistor dividers on the both side of the diac. I hope this post will not be shaddowed.
PS1: for the moment I can reply to others posts. I'ts better than nothing. I can't make new answers, those are shaddowed by default... strange YT...
@@The_Stressmaker Perhaps you posted some derogatory comments somewhere on RUclips. The system seems to take offense to the slightest opinions that are made by people that it deems offensive, and it's quite sensitive, just like many of the snowflakes in the population today.
Let's see if THIS comment gets shadow banned because I offended the bot.. 🤷
@@davelowets That comment was the first and only comment from this account in about 11 (eleven) years. Talk about minority report and preemptive strike. But the question remains... why i cannot initiate a new post but i am able to respond to another post. Maybe the answer is in my username or simply that the YT security bot is so stressed that a simple username like mine is triggering all the red flags and alarms...
BTW, thanks for the comment.
Ballasts are definitely more expensive, I do lots of repairs to 2KW sports pitch lights. A 2KW 415V ballast costs about £230, an ignitor is around £35.
The price of ballasts can only get higher. This is where the simpler ballast and an external high voltage ignitor module may have an advantage.
For an owner of a British classic car, saying that you've been sent a lighting module from Lucas (prince of darkness), gave me a laugh.
The different spelling keeps the magic smoke in :)
Ah yes, years ago battling with dynamo voltage regulators...argh.
Being an American who inherited a MGB I can relate to this in a way I never knew possible.
There are, apparently, standard recipes for Bosch conversions.
It could be worse, it could be magneti marreli!, magnetti marreli and the West of Scotland (where I am) is a bad mix 😂
All our soft orange glow sodium street lights were replaced with bright, cool, LED's.
It's just not the same, looks harsh at night.
It also looks quite dim to me in Estonia._
The LED's should be nearer the red spectrum they don't need to be as bright white as they are the disturb wildlife don't allow for proper rest patterns..
Some LED street lights I've noticed can be horrible. Not just the brightness as you travel from dark to light and light to dark but also the strobe effect can be quite pronounced if the council skimped on lights that don't have smoothing, this makes it look like objects are flickering as they move which is disconcerting
Not to mention, they can screw up your sleep cycle
When I saw this video, I immediately went and got my Philips SX 26-01 ignitor (for a SOX-E 18 lamp) for reference since it looks similar, however, it has more components and only lamp and neutral connection. I can see a blue capacitor with the X2 marking on it, and the big blue one with 1000V, and the yellow "mov" component says "275V 593-8C 0101". It also has a triac BTA208.
There are also ignitors with HV transformers inside but these are usually completely potted. These are connected between the ballast and the lamp. There are also ignitors with a timer which stops the ignitor after few minutes of continuous striking.
Some time ago when the city was replacing almost all the street lights from sodium to LEDs I was able to snatch a complete set (ignitor, capacitor, ballast and lamp holder with reflector) and quite a few lamps
Fascinating teardown/reverse engineering 👍👍, changed hundreds of these things over the years on lighting repairs
Yeah. They seem to give up when a lamp fails.
Omg so fucking close to one mill .. from the depths of my heart. Congratulations!.
Brilliant Clive I’ve never seen the ignition unit before, I have had a few old car park sodium lights they were bright as hell, never opened them up though 😊
It turns the bulk of the ballast inductor and turns the coil into a HV pulse transformer of about 6kV output that ignites the lamp. Seen several schemes for ignition of these lamps as well.
High pressure and low pressure sodium lamps are a fascinating design, and many of the bulbs have instant restrike, if they are shut off hot. The difference between the two is differences in color rendering, although neither of them are a pure white. Metal halide, on the other hand, is a better option, particularly for sporting events, and the like. However, the restrike time may take as long as 15 minutes. Sodium lights have one advantage to flora, and that is that it doesn't stimulate plants to grow at night like MH and MV lamps tend to do. However, our eyes perceive the light to be brighter than MV lamps.
I generally switched from HPS to MH for flora when I wanted to trigger flowering of a plant (along with a shortening of the 'day' length).
the eyes of us humans are designed to work with sunlight. sodium lights might produce a lot of light but because it is not the same as sunlight they are effectively less efficient. its to do with the rods and cones inside our eyeballs.
MH also has the disadvantage of shorter bulb life. "only" 10k hrs instead of north of 20k in many HPS.
This, and a modest advantage in efficiency for HPS, mark the major reasons why standard issue street lighting is (or, sadly, was) HPS.
The restrike situation was addressed in some MH models by having a secondary lamp in the fixture that would power up while waiting for the main MH to cycle. My work did not have these, and chose to replace their high bay MH with HO fluorescent. That they are now replacing with LED. (Yes, I grabbed both an MH and HO-F fixture.)
I cant speak for MV, though the same research that praises LEDs suggest that they would also have this advantage, and could even have favorable photopic efficiency (accounting for varying color sensitivity at night) than LED.
I don't know for sure though, because many places have banned MV due to its high mercury content. I wouldn't mind owning one though.
@@FerralVideo , i think anything with mercury needs to be avoided. when schoolchildren 'spill' it over benches to watch it run they (read me) are actually ingesting vapors of the stuff. i dont understand how such a heavy metal can evaporate but it seems it does.
Interesting what you say about instant re-strike. I often wondered why they didn't do it anyway - all hot and vapoury. I don't know, but i am guessing green light wouldn't stimulate plants much - and may even be a useful colour for people to tend plants at night without disturbing them - but allowing them to see what they are up to!
It's curious. i used to think that the low pressure sodium lamps were used because our eyes were most sensitive to that yellowy colour - and in fact of course, it was just a lot of lumens per watt at the time. It was only as I started playing increasingly with lasers where the power output is measured only in light OUT and nothting to do with power IN that i realised human eyes are optimised in terms of sensitivity to green. A 523nM green laser at 1mW APPEARS brighter than a 10mW red laser at 680nM
As such you get the slightly odd situation where a bright green laser SEEMS more dangerous than a much "dimmer" red laser but in fact, assuming output power is equal, the extra perceived brightness just triggers blink reflex much faster and makes it SAFER. Indeed you can get an almost invisible 808nm "pump" laser diode and if youmake the grim mistake of looking into it, you can see (just the once) a deep dark cherry red glowing dimly. Sadly blink reflex isn't triggered and you are down to just the one working eye. It's why a "white light laser" usually has a MUCH stronger RED and BLUE diode than the green one. Eg: 3W Red 1W Green 6W Blue
Awesome Video big clive
Most ballasts in the US are multi-tap ballasts with autotransformers (HX and CWA) that usually have 120V, 208V, 240V, 277V, and sometimes a 480V tap as well. Some ballasts can also be reactor type ballasts for a single voltage, although they are rare.
Reactor/choke ballasts for high pressure sodium lamps are very common as 35W to 100W as well as S55 150W HPS lamps have arc potentials of 52V to 55V and thus 120V chokes are a simple and efficient solution.
@@randacnam7321 You're right. most of the other reactor ballasts not for 55V high pressure sodium lamps will usually need a voltage higher than 120V.
Out of context but sox slays
@@ThornBeta5 Problem is low pressure sodium lamps require skilled lampmakers and are thus much harder to offshore to where they can be made by child slaves.
They were made (the uk ones) in GB until they were discontinued
I do like how you explain how the light woks. I do prefer the high and low pressure sodium bulbs as the (LED lights are to me (Bloody useless) as the light output is basically crap.
You can still get the bulbs even though they are (ES 40 ) (Edison screw) cap I would like to get hold of (Electronic igniter ) to power the bulbs as (Son - E) have internal igniter. While (Low Pressure ) (400 / 600 / 1,000 Watt) do not have internal staters.
Very good Clive keep up the good and outstanding work you do.
An important terminology distinction for American/Canadian ballasts is that common always refers to the connection common to the supply, ballast, ignitor (if used for that lamp type) and lamp. This will by convention be the neutral if the fixture is fed by 120V, 277V, 347V or less commonly single hot 240V (which we have in corner grounded 240V delta and 240V/480V single phase 3 wire services, the latter used by some municipalities and utilities for lighting services). The lamp connection is always marked "LAMP" on ballasts where the lamp connects directly to the ballast.
@@Choukai_Kai Multitap ballasts are actually very common for mercury, metal halide and S66/200W and higher high pressure sodium ballasts. The most common multitap ballast type is 120V/208V/240V/277V (commonly called 'quad tap') which covers every common lighting operation potential. 5 tap 120V/208V/240V/277V/480V ballasts are also common in fixture repair kits as that way a single ballast will work in all cases. Another common one is dual tap 120V/277V ballasts, which are usually lower power as they are meant for applications that would use a 100W or 200W 120V or 277V GLS lamp; all of mine are 50W S68 HPS or 50W M102 MH.
The Canadians also use 120V/277V/347V tri tap ballasts on single hot lighting circuits, as a regulatory requirement there is that the shell of the lampholder must be at or near ground potential; fixtures there connected in single phase delta (208V, 240V, 480V or 600V) or to 240V derived from a 120V/240V single phase 3 wire supply use isolated secondary ballasts.
The Mexicans use 127V/220V/254V/277V quad tap ballasts as Mexico uses 127V/220V and 127V/254V as standard 250V class services.
The weirdest multitap ballasts I've seen are 120V/220V/240V 50cy/sec ballasts, which from what I can tell are mainly for export fixtures and overseas military installations where everything is wired to American norms but only 50cy/sec grid electricity is available.
It reminds me of a dimmer circuit. I think I've seen people drive car ignition coils from mains voltage in a similar way. It's like a capacitive discharge ignition driver
It's amazing how for so long we've used lighting for large areas that was basically a trapped bolt of lightning, causing the gases inside a capsule, tube or bulb to glow & emit the UV light needed to stimulate phosphors, makes you wonder what lighting technology could come once LEDs become just as old... :\
Particulary for street lighting , after HPSV & Mercury Vapour lamps came the " INDUCTION LAMP " these worked on 2-3 GHz freq. hence did not need physical any connection.. energy transfer to light the lamp was through an induction coil encircling the lamp in the middle..
Dimmable down to (0-1)% ..and had a continous run time that could allow operation upto 40+ years without any loss of brightness over that time span....
Development work was still on to bring down controller & mfg. cost..but LED's development came along a lot faster & cheaper.....
I was literally just running out of quality content to watch on RUclips, and up you popped Big Clive, just in the nick of time!
EDIT: Apparently, we're all going to go blind (no, NOT because of that), because of blue light exposure from screens and LED lighting! 😎
Some of the first organisms that evolved in the oceans of the earth possessed a very clever survival technique: Blue light is one of the last visible light wavelengths to vanish when sunlight strikes the surface of the ocean. These micro-organisms hide in deep water during the day and move up to depths closer to the surface once the blue wavelength vanishes from the water. A cue for these micro-organisms that it is safe to feed.
This trigger exists in the human DNA even today and can affect our behaviour. Whilst there is still blue light around we are tricked into thinking that it is still day time. Many computer screens emit blue light so staring at a screen can lock us into this "day-time" mode and prevent us from wanting to go to sleep. This is a source of fatigue and we dont even realise why we aren't tired and yearning for sleep and rest.
There are screens and eye glasses that block the blue wavelength light in order to limit this effect.
@@PetraKann I have a blue light filter on my Tablet set to maximum. It turns the screen a sort of sepia colour. I also take a Lutein supplement, which helps protect the eyes.
@@marcse7en Blue light doesnt actually damage your eyes. It tricks the brain into thinking it is still daytime and so our instinct to sleep and rest is delayed.
AN old evolutionary biology and adaptation that occurred over a billion years in a marine micro-organism. That genetic chracteristics is still with hominid species today and can be triggered in people that are exposed to the blue light wavelength for extended periods of time. So it's not quite right that Blue light directly makes us tired through our eyes.
Casinos and other entertainment establishment can and do exploit this by ensuring blue light wavelength is plentiful and intense at gambling machine screens - keeping people alert for longer periods in order to gamble (if they still have any money left).
There are lots of examples in the Human DNA where ancient genes are still present and can be dormant and/or triggered by certain cues in the environment.
I have blue filter treatment on my glasses when working in front of a computer screen.
Also a good idea to be exposed to natural light every day (not only for your Vitamin D)
Big Pharmaceutical corporations are the most corrupt and devious organisations on the planet - they make the mining industry and the banking cartel appear like angels in comparison.
CHeers
@@PetraKann , there is some research which seems to indicate that this blue light you are suspicious of may help alleviate depression in high latitudes. a replacement for sunlight. i am one of those who feels depressed from not enough sunlight.
@@vsvnrg3263 That research wouldn't surprise me. Many phenomena can have both beneficial and detrimental effects on the human body.
Usually it's a matter of balance. It also varies with each person's sensitivity.
What do you mean by High Latitudes?
I know here in Australia there were huge health issues during the Pandemic lockdowns due to a lack of sunshine (Vitamin D) and just a lack of fresh air etc.
Cheers
There are three common types. A parallel two wire type which goes across the lamp and produces a 750v kick, usually for metal halide lamps designed for use on mercury vapour ballasts. Semi-parallel type which is depicted here which is favoured by Philips that is usually good for 3-4kv and then the more common super-imposed type which is good for 5kv. I have plenty of spare super-imposed types I could send but they are all potted!
Potting always makes teardowns much more time consuming.
When I was in the diecast model trade, dead streetlight balasts made great heavy haulage truck loads, after painting them red oxide primer. I got them for nowt off a customer that worked for the council
I am puzzled with the two components you mentioned too, but i had wondered how the 1KV capacitor charged - what it was referenced to. I could see its Live connection via the inductor, but where is its ground if not via the triac, other than the two oddly placed components. i admit I am still puzzled..
Also I love the less efficient but oh-so-simple self ballasted lamps which are available in 160W and 400W flavour. I mention this for the interest possibly of others as I am sure you know more about it than me, but instant light is provided by a filament which also heats the discharge tube and then a bimetallic switch puts the discharge lamp in series with the filament (IIRC).
A bit more epensive but they were awesome easy to deploy quickly where you just needed lots of light with minimal kit / weight / effort. E40 bulb holder and a 400w self ballast and off you went - the 160s just plugged into a regualr B22 socket :)
Not as efficient I guess as a properly ballasted lamp, and a bit dearer - but for some applications, that didn't matter too much. :)
Nice device to be used in my 4th of July Firework's show...
The TRIAC snubber is there. It is formed by 220k + 220nF.
I believe you may have swapped the Line and Neutral connections to the igniter module. The only connection to the lamp is the ballast (there is no connection to the igniter module here).
The ballast also functions as an auto-transformer. The traic is connected to the centertap via the 1kV capacitor and fuse, and to Line on its other side. The triac gate drive is also connected to Line. The 120k resistor is connected to Neutral.
When trying to ignite the lamp, you need a high voltage pulse across the lamp (several thousand Volts, more when the lamp is hot or old (not cold)). The ballast, functioning as a boosting auto-transformer can produce that pulse by discharging the 1kV capacitor (charged to near peak line voltage) through the low impedance 220V/240V ballast winding. The ballast, functioning as an auto-transformer, boosts the discharge voltage to the thousands of volts need to initially beak down the sodium vapor lamp (striking voltage) to form a low resistance arc. As the lamp warms up, the arc resistance continues to decrease, but lamp current is limited by the ballast now functioning as an inductive choke.
How does the capacitor discharge across the 220/240v winding? The triac (and its gate drive) should be connected to Line rather than Neutral. When the triac fires, the capacitor is discharged through the low impedance side of the auto-transformer. The high impedance side of the auto-transformer develops the high voltage pulse.
How does that capacitor charge? The 120K resistor should be connected to Neutral. (There are no other connections to the lamp / ballast node.) The capacitor charges from Neutral through the 120K resistor, the 220k/thermistor pair, and fuse to near peak line voltage. When the PTC thermistor gets hot, its resistance rises, reducing charging current and the capacitor voltage. You can estimate the capacitor charging current by assuming the thermistor is either open when hot or shorted when cold.
The triac is triggered near the peak AC line voltage by the diac breaking down when the 200nF capacitor has almost reached peak line voltage.
This circuit attempts to ignite the lamp every half cycle. However, due to the thermistor's high resistance when hot, the resulting pulses become weak after a few seconds and the lamp will fail to ignite until the igniter has cooled.
I've replaced many of these on railway level crossings in the past. I never really knew what was inside.
I'm super fascinated with this tech, and i try to collect as many different variations of MH, HPS and Hg lamps as possible.
Thus i would love to see more about these things.
I have MH/HPS electronic ballasts that aren't potted, as well as quite a few ignitors.
None of the ignitors i have looked into contain any HV transformer, only the "electronic ballasts"
However, the circuit for the electronic ballasts (CMH at least) is very strange and foreign to me, first time i saw it, i was deeply perplexed because it looked like no other electronic circuit i had ever before seen, but for some reason, the CFL electronic ballasts doesn't have the same overall design (lower ignition voltage i suppose)
It is also interesting how different the NA design is to the EU one, because the lamps operate around ~100v, it is actually pretty difficult to run them on a 115v supply, so the NA design is typically much more elaborate.
Fun facts for the nerds out there:
CMH exists all the way down to 20w (HPS down to 50w)
There is a warm-white sodium lamp, although quite niche, they are known as SDW.
SON(HPS) is still to this day, more efficient in terms of lumens per watt than most LEDs, especially when factoring in investment costs - this makes you wonder about the math behind switching from HPS to LED street lighting.
Heh, that makes two of us. My collection is in its infancy though.
I have one stray MH ballast, three MH fixtures, and five HPS lamps.
HPS: Four little 35w wall packs ($5 ea at yard sale), one 70w flood. One 35w lamp had a bulb that broke inside (arc tube cracked).
MH: Two High bay fixtures, multi-voltage, 400w, probe-start. Heckabright. The stray ballast is a weird one. Maybe you can help me here. It's labelled 70w, but the ANSI code is for 100w bulbs. Which size of bulb should I put in it? Can I safely run 70s? I could install it in the 70w HPS flood fixture if I wanted to. One porch light of unknown but highish wattage. The markings on the bulb came off years ago. Looks like a 250.
I wouldn't mind adding an Hg lamp or a few to the collection, and a companion in a group I follow is actively trying to source one.
Both of my 70w are multi-tap magnetic ballasts with an electronic ignitor unit and a PFC cap. The 35s are single voltage 115v with a simple coil and the electronic ignitor and cap. I'd have to pull one apart again to remind myself exactly how they're wired - I think it might actually just be an inductor. The high-bays are probe start, no ignitors.
The one 35w HPS that failed was actually really interesting, in that the ignitor was able to strike an arc across the leads going to the discharge tube near the socket, creating a faint blue glow there. I didn't run it long in this state as I didn't want to damage anything.
As for the switch to LED, the thing that drives me the most bonkers is that most discharge lamps were designed to be easily serviced. LEDs are explicitly designed to be unserviceable, despite having comparable service life to one or two HPS bulbs if not overdriven. (You watch BigClive, you know that they're all overdriven)
They say the switch is because of luminous efficiency. The spectrum that an LED produces is more in line with our eyes' sensitivity than HPS, resulting in more *visible* light efficiency.
I question this, but this comment is already long enough.
I'll leave off by just remarking that I wouldn't be quite so bitter about them if they weren't explicitly designed to be thrown away complete at EOL.
Maybe the varistor/resistor combination discharges the 220nf capacitor in the timing circuit. That would reset the capacitor voltage so the timing is consistant. ?
It works like this: When the triac triggers, it formes an autotransformer with the tapped inductor )AND a resonant circuit with 180nF and input part of ballast inductor from L to fuse. That will oscillate until current drops below triac holdoff current. Triac does NOT turn off instantly! (and high dv/dt triggers it too and high di/dt keeps it conducting!) It needs some time with low current to turn off. Not so obvious. Look more closely to datasheets to see cutoff current and period it is needed. Period was not documented on couple random datasheets I checked, so check several to see an example.
Because triac has a high frequency ringing current (L-C resonant from part of ballast and 180nF), it won't turn off until current has dropped low and been low for a bit of time ie. ringing faded. (at 50Hz it may seem instant but it is not) Then either the lamp has started with high voltage pulse (high current ringing in 240-220 part of autotransformer) and the high frequency current faded (been below cutoff current for cutoff time ie. triac not conducting) and you got lamp on keeping it from re-triggering. If lamp did not start the cycle starts over by charging 220nF and 180nF for a new trigger.
I think NTC probably is part of charging circuit and perhaps needs to cool off before the system can trigger? You cannot trigger a hot lamp with a few kilovolts, you need some time to cool down the lamp before starting to ignite it. (Did not analyze thoroughly, it could also be part of 180nF charging circuit)
I think 220nF has the resistor in parallel only as a bleader. Basically it is a 120k + 220nF network charging that triggers the triac.
Quite a complex thing to figure out unless you know the details and can also think about the resonances (that also make it have more than mains voltage at tapped section due to LC resonance. Peak voltage at inductor is multiplied by Q of LC circuit). So, you get out something like Vinitial(in 180nF) * Q * ratio_of_autotransformer = usually several kilovolts.
For mere mortals 593 series varistors are only available between 30V and 550V but if they used the same logic for a custom voltage it should be:
593=9 mm diameter
9=something related with packing
00=RMS operating voltage multiplicand (100)
1=RMS operating voltage multiplier (x10)
6=something related with leads shape/size/separation
So it's probably a 1000V varistor.
With a ( redundant ) 400 Watt SON fitting , the ballast and 40 μF capacitor can be reused to run a 3 phase bench grinder from a single phase supply ... connect neutral to 1st phase , ballast to 2nd phase , cap to 3rd phase , then common the other side of cap and ballast and connect this to single phase live ... hey presto , grinder is now running well ! ... ( tried - n - tested ) ...... DAVE™ 🛑
The MOV is shorting out the inductot when the triac fires. You have an RLC loop with the 180 nF cap, the ballast, and then 120kΩ resistor. By shorting out the 240kΩ resistor, it limits the voltage seen by the semiconductors.
"The MOV is sitting out the inductive does when the triac did of" What - Huh? WTF? That doesn't make ANY sense at all.
@@simontay4851 Autocorrupt strikes again.
The 220k seems to be the bleeder resistor for the 180 nF when the power is off.
But is the thermistor a PTC or a NTC? NTC for a soft start, I would guess.
BTW In the US most commercial lighting is 277 VAC in order to use smaller gauge copper wire and save $$$
Thanks, Clive. 👍
I thought it was a thermistor at first - maybe as a simple timing function to shut off the ignition spikes after a suitable time, but it appears to be a voltage dependent resistor.
The main and lamp connections for the igniter are reversed. The big capacitor make a brief connection to ground thru the triac shunting the ballast to ground. The 220K+120K and the current from the capacitor are not enough to mantain the triac in conduction so after a short period it will disconnect the capacitor from ground. This will create a voltage spike that will ignite the sodium lamp. Think like a boost circuit with ballast as inductor (essentially the ballast is an inductor). After lamp ignition the voltage on the divider side of the diac is lowered by the big capacitor under the trigger level. The divider in the triac gate rise the voltage of triggering over the lamp working voltage. That is what i see in the schematic.
BTW, the dividers are there to assure that the firing is on the maximum level of the sinosoide to have the maximum chance to ignithe the lamp.
PS: keep the good work Big Clive and give us, tinkerers, food for brain. :). Sorry for any misspeling, english is not my primary language.
Saying "I don't know" is the sign of a good scientist, or in this case a Doctor 🙂
I've always wondered why this tech never used slayer exciters to start. I've had a 160 watt MH lamp light up 200 cm away from a cheap plasma globe.
Maybe it's reliability? Semiconductor electronics don't like heat, as I'm sure many of the dead LED street lights can affirm.
Also, I've had ideas about a auto variable ballast that slowly turns up the current over time to squeeze more life out of the lamps. I think that's why they cycle indefinitely at their EOL stage? The electrodes recede to the point where the current required to sustain the discharge is above what the ballast supplies. Is that why, or am I missing something as I don't have one of these fixtures to probe around on?
Thank you.
The company I work for has just swapped to PIR led lights throughout, trouble is they turn off after 60 seconds so they often turn off mid job And I now have to work a 170 ton press with only 150 lux of illumination 😎
I'd imagine that the MOV connected that way is some sort a transient clipper. The MOV turns on, and in doing so, forces the TRIAC to fire. The firing of the TRIAC forces the transient to be shunted to common. This would move the brunt of the transient clipping to the TRIAC which should lengthen the life of MOV, while TRIACs don't seem to care much about high currents or peak power dissipation.
Sodium lighting is appaling, the amount of Hangars I've worked in and you couln't see bugger all. The funny thing is though, when at RAF Valley in Gaydon hangar. We were told to wear bump caps due to the sodium bulbs that had been replaced bursting and showering glass everywhere. Not sure about this, but the lads in the hangar said that the workmen had replaced the wrong voltage bulbs.
the eyes of humans are designed to work best under daylight sort of lighting, not sunset sort of lighting. its got to do with the rods and cones within our eyes. i also hate sodium coloured lighting.
When the triac conducts, the capacitor basically grounds that "20v" winding, viciously reversing the existing magnetic flux in the core, with obvious results.. The MOV is likely a clamp in case the lamp isn't fitted.
MOV is going to trigger the triac on overvoltage, to get the triac conducting, making the sharp voltage gradient to start the lamp. Diac charges to threshold if lamp does not strike, as it gets to the peak of the mains, not the normal 100VAC or so of the running tube. Then triac fires, shorting out the reactor for the remains of the cycle, and storing voltage in the capacitor. On nect half cycle at mains peak the voltage is enough across MOV to start it breaking down, and thus turn triac on hard, shorting capacitor and dumping charge into the reactor, making a series LCR circuit, with a high voltage at the far side of the coil, as current is flowing through the coil. Coil rings at least a few times, and lamp should strike a little, next cycle it charges capacitor again, and then triggers another set of rings, each time heating up the gas, till the lamp finally ionises and strikes, keeping the voltage on the 120k resistor low, and thus not enabling the triac to fire any more. Even on 220V the inductance of the wire, combined with the high rate of change, means that it will strike, unless you have a very short ultra low impedance wire direct to transformer, which for most lamps you never get, typically 10m of 1mm wire does provide more than enough impedance.
EOL the triac fires multiple times, till it gets the fuse warm enough to melt, generally after a few hours of the current pulses, or if the capacitor goes short circuit, on the first cycle. Open circuit capacitor it just does not work.
Those make good grow lights, for your home grown.😀
Interesting, I had expected to see something more like the starter from a florescent tube fixture.
Big difference between a 40w fluorescent tube and a 1000W HPS globe 😎
@@LDuncanKellyFinally, something I can understand. As usual, most of what Clive says and the comments go way over my head, but your response is easy to understand, even for me.
I suspect that when the 20V coil is dumped by the triac the voltage spike in the 220v section will jump dangerously high, much higher than is actually needed for igniting the lamp. So the MOV is there to clamp the top of the spike. Just a guess though.
That's sort of my thought. But I would surmise its there to shunt the pulse around the timing network and thru the the triac itself in order to keep spikes out of the timing side of the circuit each time it fires to strike the lamp. Given the whole thing is behind a 120k resistor its probably not going to pass much current, but could be enough to cause false triggering (or even potentially damage the gate), so it and the snubber network are probably belt and braces to ensure it triggers properly. That's what I think its for, anyway.
I have seen ignitors that are in between ballast output and lamp with common neutral and some that are in parallel with the bulb with two wires._
MOV and 220k are there to discharge the capacitor maybe.
Could the mov be protecting against back emf from the coils?
900 volt MOV prevents spike from damaging the gate o the triac ( clamping)
You should do schematics left to right on first one
I was told 35 years ago, that it is OH to measure the socket voltage of a Metalarc, but never try it with a HP Sodium - it will blow up your meter in your hand, Scare me enough that I never tried. Used a light probe to see if they were powered up. Metalarc were close to 600V though. (In the states & Canada)
My missus prefers the "flick the switch, witchcraft happens and the light comes on" 😂😂 I spent nearly a year playing with this kind of thing in a large factory. It was never ending!
This autoformer effect (single winding with taps) is neat - it works pretty much like a voltage divider when the supply comes to the end taps and the load involves one or more of the center taps. But when the supply involves one or more of the center taps, it can step voltage up. It is more efficient than a resistive voltage divider which is the only option for DC and only allows stepping down. An autoformer can also use thinner wire than a respectively powerful normal transformer and many voltage adaptors, especially cheaper ones, rely on the design. Most variacs are of the autoforner type and do not offer isolation from the mains.
Which comes to today's adventure at work. We have a plentitude of light fixtures that were installed 15 years ago when fluorescent lights were still very common and the white LED technology was only slowly making itself known. Many years ago I bought a sh*tload of G23 tubes that fit in these fixtures - I got them dirt cheap from a traffic station store that was clearing products they were planning to discontinue. I also got a heap of windshield wipers that were old stock.
But, and this is a big but. The tubes I got were mostly the 11W variety and the fixtures are designed for 2 x 9W. I can get a 11 and 9W tube in series to strike 50% of the time, luckily they are mostly on day in day out. The 11W tubes fit without the cover bur will not strike with two in series. There are no taps on the ballast to adapt for bigger tubes. What I found to work fairly reliably is a 7 and 11W tube in series. Also the 7W tubes seem to be the cheapest, must try to acquire some when they are still available. The supply is dwindling as they are apparently no longer produced, and the LED replacements for those are F****G expensive at the moment.
Where I used to work we had really bad tappings on the sun station for the factory so our high tech fluke portable appliance tester would often give an error because it was receiving 260VAC. Depending on what was running in the factory it would vary slightly but normally be 259VAC mains voltage. How would something like this effect this sort of device? Because it would be putting nearly double the voltage across the short winding of the ballast.
It would result in higher lamp current and more heat from the ballast.
Hi from Brasil, I was wondering if you could do a teardown in the Salt Water Lamp Emergency Lighting Outdoor Lamp No Battery to see if this is a misleading product like it's just closing the circuit with the hidden battery using the water and salt. I did not find anyone opening this so I thought of you!
It uses the salt water as the electrolyte. The real work is done by the two different metal electrodes, that will gradually dissolve. It is not the water that creates the power.
Discharge resistors perhaps? Although it does seem strange having a varistor and a fixed value resistor in parallel like that. I don't believe they're there for protection against voltage spikes, as the varistor's resistance would drop significantly under such circumstances and surely the current would dump through either the triac or the resistor+capacitor setup depending on where it is in its cycle (probably the triac tbf), which seems bad. Besides, isn't that what the fuse is for? I'm stumped, although this is hardly my area of expertise. I'm a chemist by training. Disconnect them and see what happens.
Another hypothesis: as the voltage moves from 0 to the peak of the sine wave, current will flow less via the 120k resistor and more via the 220k + var until we start moving back to 0, charging both capacitors. Then, when the 220?nF activates the triac pretty much all the current now flows through the 180nF fully charging it until we hit 0 again (or thereabouts), and then it *dis*charges on the next half of the sine wave once the triac activates and shunts the current now going the other way through it. But even then, they seem... Superfluous? Perhaps they found they needed to charge that big capacitor a little more and so they allowed it to connect to the neutral via something other than the triac to give a bit of extra juice in the upswing of the sine. But like I say, chemist not electrician.
Where do I send electronics for Clives teardowns?
I've never seen an ignitor that isn't potted before!
The ones with high voltage transformers in them usually are potted.
General Electric often uses a kind of cordwood construction for their 3 wire ignitors using plastic frames with the components soldered or crimped to each other or to spade terminals for connecting to the fixture. More complicated 2 wire ignitors and universal 3 wire ignitors have a circuit board in this frame, but it isn't potted as there is usually a shunt wire to cut if the ignitor is used with 90V to 130V arc potential lamps (150W S56 HPS, all 200W to 400W HPS and all pulse start MH up to 400W). American Electric also used bare circuit board style ignitors in their HPS fixtures a lot.
It just struck me. How many drafts of the schematic do you usually go through before you get to what you show us? 😛
Usually just one rough one and then the final one.
I don't know if you guys have constant wattage auto transformers over there but here in the US if you have a CWA ballast and the cap fails you will likely lose the ballast too.
Only if the cap fails short or low impedance, as it is what's doing the current limiting.
Is that the ballast people carry on submarines and hot air balloons? Where do they go when they're released?
I can tell you one thing if you want a lamp that'll lasts for what's seems like forever.. choose an HPS.
could 90016 mean 900v 16A?
Amazing these are being replaced by truck loads/lorry loads
Clive, you're into ozone and all that. Would it be reasonable to use an ozone generator to disinfect an entire property? Basically close all the windows and ventilation ports, open all interior doors, plug in a powerful ozone generator and leave the place empty for a few days? Would this have problems with regard to ozone leaking into neighbours properties and since it's an oxidizer, would it cause damage to objects prone to oxidation such as metals?
Ozone is often used after flooding to try and remove some of the odour. It depends on the reason for wanting to disinfect the place. If it's deep seated mould then it may require furnishings to be changed. But if it's just a lingering odour then it may work.
It takes a long time with continuous exposure to damage metals. Rubber materials are more prone to surface oxidation.
I was a little surprised to see no R/C snubber across the triac because it is driving an inductive load. Perhaps that's exactly what the MOV is doing. If so I would expect the MOV to become worthless over time due to experiencing repeated activation.
I think you are right about the MOV. I have seen them used in kiln controllers (where there is often some small inductance), many units last for decades of use.
Actually there is a snubber circuit there. It is formed by 220k + 220nF.
When SON lamps age, the arc gets often interrupted. But it takes a long time until the lamp strikes again. If there also a varistor in the lamp?
Ultimately the sodium reacting with the aluminium oxide arc tube over its life depletes the elemental sodium content to a point where end-of-life tubes may strike ok from cold, but as they heat up, the internal pressure rises to a point where the voltage required to sustain the arc exceeds the supply. The lamp extinguishes and has to cool right down before the internal pressure is low enough to allow the arc to restrike. I have seen lamps in redundant areas at my plant cycling like this for 10 to 15 years without failing completely.
@@v8snail Makes sense! I always thought it's some protection mechanism. Why does the arc get interrupted despite the same light output after aging?
@@kyoudaiken
That's described by the sulphur content loss as it reacts with the aluminium oxide arc tube under high heat and pressure over the years. Less sodium requires higher voltage to sustain arc, higher temps causing higher pressures results in higher voltage required to sustain arc. Age equals lower sodium so there becomes a point where the voltage required to sustain the arc exceeds the supplied voltage and the arc extinguishes, cools down, reignites, warms up, rinse, repeat.
Well that was Exciting 😆
I'd query the values of the 120k and 18k.
;)
Initially 120k & 18k and the full winding complete the 220nF charging ckt. thereby limiting the peak charge voltage across the
18k/220nF parallel combo.
This ckt.is also effective in repeated restrikes if the ignitor doesn't fire the lamp...
Once proprely lit , the lamp low voltage retrigger is through the series ckt. of 220v winding , F2H , 180nF , 220k and 18k/220nF ....this is useful in hot restrikes in case the lamp goes off momentarily..
In the cold startup sequence , once triac triggers , current flows through 220v winding ,completing the 180nF cap. path to return via the triac. Later when triac turns off , current must continue to flow in the same direction through the boost winding..since the HPSV lamp initial presents an open ckt. , a (5-6 )kV voltage is generated that can breakdown the potential barrier across the lamp to ignite it.
There's an issue in America, some street lighting is failure turning blue! Supposed to be white lights hut the LED mfg failed now they are bue/ purple
The phosphor layer detached from the violet LEDs.
What kind of Edison screw size would that lamp be? I'd suspect not E27 just to make sure things that aren't compatible aren't compatible :D
E39 (US)/E40 (EU), and you are correct. "In areas following the U.S. National Electrical Code, general-use lamps over 300 W cannot use an E26 base and must instead use the E39 base."
Bigger lamps are GES and smaller ones are standard ES.
is the ignitor what we used to call a starter in the US?
Starters are for preheat fluorescent lamps. The job of a starter is to provide heating current to the lamp electrodes to facilitate easy ignition as well as cause an inductive pulse when the starter opens. Integrated ignitor HPS and MH lamps often use glowbottle thermal switches like those used in fluorescent starters (just built for much higher currents), but they are prone to failing short from contact arcing.
A 1000v capacitor is an extra-extra 5G Death Beam device!!
Hi,
I found only 2 connection in one of my 250w hps ignitor fixtures.
Is that normal?
There are a wide variety of starting systems.
@@bigclivedotcom
Thank you
Would you purchase a LPS street light and do a tare down of it? Would be seriously cool :)
Hmmmmmm Tare - deduction from the gross weight of a substance and its container made in allowance for the weight of the container. 🤔
@@brucepickess8097 sorry I’m not following?
@@ThornBeta5
Tare = a weight term used in shipping and haulage
Tear = tear down
It was a poor attempt at subtle humour
@@mrglide7078 evidently! Sadly spelling isn’t my strong point 🤣🤣
Curious ... I have a sodium vapour lamp but it only has a ballast, no igniter.. lights just fine. I am pretty sure there is nothing in the ballast but a big fat coil, no extra parts... Are there lamps that don't require an igniter ? And if so.. why does your (that) lamp need one ?
Some have an extra ignition electrode or fluorescent style shunt starter inside the lamp.
Always open your black box
Ballasts, boo, a single filament would provide it's own ballast, no components required 😅
I did once put a 150w uv center section of one of those bulbs across straight 230v,it went pop, but, didn't explode, strange!
Make your own ignitor with a taser module? 👀
Maybe it’s a 90,016 volt varistor? Maybe 900.16 volts? Maybe it’s not a varistor at all; maybe it’s a metastatic boomufoomulator. 😅 Some experimentation is needed here. Could it be a thermistor that’s 900 ohms hot, and 16 ohms cold (or vice versa)? It’s a real pity, though, that this module doesn’t contain the obligatory 8-pin micro. 😢
One of the local hotels changed out all of their inefficient halogen spotlights that run through the pedway. Probably 50 new led potlights. The first day one was already flickering. Today... probably 4 months in... half of them maybe more are on ghetto disco mode. The homeless squatters are not impressed.
It's a feature, not a bug! :)
@@NiHaoMike64 I just walked through there and thought I would count. I exaggerated. The string of 15 had 5 flickering away. Its quite an effect. I didnt count the other strings... its 4 rows of these. I wonder if that was a bad batch or if they take other others on a circuit with all the noise/spikes.
Hmm - 4:48. I was looking at your worksurface and realised that if someone showed me that as an image; on their phone, or what-not; I would immediately be able to say, "Oh - that's big Clive's worktop". I've seen it for so long now, it's like an old friend. Have you ever considered releasing a high-definition image of it for use as computer desktop wallpaper? Or, do you not realise the familiarity we all have with it, or the nerd-niche kudos that would give? (*As always - great video - Thank you.*) [edit] Maybe, for just your own general interest, do a poll on whether people would be able to identify that desktop as being from your videos, if shown it alone. I for one ... would be interested in knowing if it's just me (and I'm weird) or whether, the community has the same response - "yup, that's Big Clive's main worktop".
I have noticed that if the thumbnail doesn't show my usual background, less people view the video because they didn't see it was one of mine.
upoloaded: 3 weeks to others... 2 mins to me. i dont mind. but the growth between us grows ever wider. thanks anyway
I think this might be "advance sight" by those that support his channel - and freelocaders like me wait an extra week or two :) Awesome content for free seems a pretty good deal. I might be totally wrong, but that would be my guess :)
I'm deliberately building a buffer of videos for some work I'll be doing soon. During the initial load in and testing I may not be able to make videos.
@@bigclivedotcomWhat are we going to do without your soothing voice to calm the hectic day?
@@bigclivedotcom understood, but why create such a large time separation ? and then delete my comment?
igniter module in a streetlamp!? that sounds like a 5g death beam item to me!!
Wouldn't the MOV have a fixed life if it's constantly being active in the circuit?
I think it probably only clips spikes during ignition.
As example , assuming the MOV is rated at 240( rms) , it begins short circuiting once the voltage across it goes above (900-950)V..{ typical clamp voltage for 240v mov from it's data sheet } but current is limited by the120k resistor....this allows the boost winding to generate sufficiently high voltage (5-6)kV to ignite the lamp.., circuit completes its path of current flow in same direction via. boost winding , 120k , mov , 180nF , & fuse .in that order.
Every time a MOV takes a hit, it's life is shortened. For UL 1449 listed SPDs they only have to be effective 8 times to pass the listing test.When I worked in high power/voltage electronics we never used a MOV as a snubber.
The fuse is in the wrong place?
I think it's just to protect against capacitor failure.
Hi Clive, another great video. Thank you. I have an eBay item that I would like to send to you for a tare-down. Can you tell me where to send it? It is a 433.92MHZ Keyfob Remote that was faulty on arrival. I have ordered this remote from different sellers at different times and they all have the same fault. Clearly, they are leaving the factory like this and they must know that these units are faulty because a simple test is all that is required to show the fault. If you could find the fault, that would be great. If you could find a fix for the fault, I am sure hundreds, if not thousands, of people, would be grateful, including me! Cheers Clive 🙂
Philips numbering defies all logic the only thing it refers to is the data sheet or the stock code.
How does a "big fat resistor" charge a capacitor?
Slowly.
@@simontay4851 ....so a resistor is really a battery?
Political correctness police. Morbidly Obese resistor.😏
@@PetraKann sorry, misread charge as discharge. Of course a resistor doesn't _charge_ a capacitor at all. It discharges it.
@@simontay4851 ....I'm afraid that this matter no longer resides in my hands. Scotland Yard are involved and most likely the UN will be dragged into this rather unfortunate international incident.
😁
Did Lukas write his name on it to remind you who sent it in, or did you? ;)
I wrote it on as a reminder.
I think you've drawn this wrong. Shouldnt the F2H & the cap connect across the bulb, not the ballast tapping? It isnt making much sense to me otherwise.
Can't really go wrong with self ballasting lamps.
Sigh. It’s 10:22 pm. I almost have enough time to watch the whole thing.
Sounds more like an impulse generator
LED floodlights seem quite childish compared to the explosive technology used here
Looks like a Philips ignitor.
It is.
Hi
a snuber network would be logical because your switching an inductive load.Altho the placement looks weird to me
The triac is in the ignitor circuit that's fired up/triggered only when lamp is to be turned on.....& that's most likely once a day..... exception being repeated strikes for retrigger attempts in case of non ignition. ..also these triacs , most likely , will be of high reverse breakdown volts( 2-3) kV , so snubber may be ineffective .
I hate HIDs. LED all of them!
*cough* Reliability. *cough*
You can use this to set light to somebodys son? 😳 (the symbol for Sodium is Na 😉)
Of course, Sodium lamps are also a type of metal halide (usually NaI, Sodium Iodide aka Anayodin). 🙂👍
Just so you know there's an ad in the middle of this video
I'm doing some tests on adverts.
@@bigclivedotcom what tests
Am I the only one that liked the sodium lights? Sure, they were crap for color reproduction, but seeing the sea of amber across the cityscape or highway was kinda cool.