Seems like I made a rather big mistake in the video. My transformer should not be switched on during the zero crossing point but at the voltage peak in order to decrease the current flow. While recording the current waveforms on the oscilloscope, everything appeared good to me and I am not sure why my circuit breaker stopped tripping when I switched at the zero crossing point. But many viewers pointed me in this direction and after doing research, it seems to be correct. Sorry about this. But if you want to build the softstarter you can still try it out whether it works for you and if not then you can always change the Arduino code in order to switch at a different time point. Sorry again for the inconvenience. I do make mistakes from time to time since this is a one man production. Stay creative :-)
@@aks8403 You have to change the timer value OCR1B. In my case I firstly used 590 which is around 9.44ms after the zero crossing point. 625 would be 10ms. So simply use half of 625 so around 313 in order to get a 5ms value at which point a voltage peak should appear after the zero crossing point.
Tripping circuit breaker? Hah, I don't have such weakness with my 30-years-old house wiring. The circuit resistance is high enough to limit any inrush current.
That was my first thought, but i guess it isn't a viable solution if transformer shares breaker with other devices. Type C is really slow and if wires aren't properly sized it can cause all sorts of problems.
I had a similar issue with a 3000 watt transformer for reducing power from 220 to 120 volts. I later learned that the type of circuit breaker used makes a difference. There are some breakers rated for inductive loads and some for resistive loads. Most home breakers are of the resistive loads found in most homes. Replace breaker with an inductive rated breaker. Did this and no problems now.
The current into an inductor equals the integral of the voltage accross it, in formula: i = 1/L.integral(Vdt). If you switch it on at the (rising) zero crossing the voltage remains positive for 180 degrees. This causes the magnetizing current to rise to approx. 2x its normal value. However the inductor core is not made for this 2x magnetizing current and thus the core will saturate. This causes the coil inductance to drop, which increases the current even (a lot) more. If you switch at the maximum voltage, the current also starts to rise, following the integral relation, but after 90 degrees the voltage polarity reverses and the current starts to go down again. In this way the current starts of nicely lagging the voltage by 90 degrees, as it should be in the stationary situation, with peaking and tripping the breaker. Best regards, Ben.
"If you switch it on at the (rising) zero crossing the voltage remains positive for 180 degrees." ....... But in normal operation, the voltage remains positive for 180 degrees and then becomes negative again for 180 degrees. What's the difference now?
@@Thomas72B The difference is exactly what you say: in normal operation a positive periode is always preceded by a negative period, and visa versa. An inductor stores current in a similar way as a capacitor stores voltage. Therefore the history is important. But at startup there is no history (current) stored in the inductor. When you then start with a full positive (or negative) period, the current becomes too high. This is only a temporary problem, after a number of periods the current normalizes, due to the series resistance of the inductor. Regards, Ben.
@@beneindhoven3424 Not always ! If you put the plug in the socket you cannot know where the sine curve is. 0 degrees, 130 degrees or 174 degrees you don't know .... That's why I still don't quite understand where the problem is supposed to be.
@@Thomas72B Yes, that’s correct, you never know where you start when you connect the plug. That’s why the circuit breaker not always trips, but only sometimes. Switching in near the zero crossing is the worst case situation. Having said that, I have to admit that I don’t understand why Great Scott’s tests at zero cross switching didn’t trip the breaker. I tried this same experiment a few months ago, and could clearly see the current peaks are maximum at the zero crossing. And minimum when switching at the voltage peaks. Regards, Ben.
Ohhhh.... Brilliant comment! I think it's also just answered a question I've had about adding energy into a LC tank. Thanks for sharing the knowledge. Edit: Yes! I've been trying to get my head around this for days. You're a catylist for understanding good sir. Double thanks!
Another much easier and cheaper way which I used succesfuly in the past was a 1s timed relay in parallel with a high power resistor with just enough ohms to limit the current to under 16A. Both of that in series to the load will limit the current for 1s and then the relay shorts the resistor. Worked like a charm :-)
good idea, but using an ntc instead of a power resistor is better, since it can prevent from burning down your house in case of failure of the relay/connection..
@@nicolaswannen1743 can you get ntc with lots of watt dissipating ? Better to use onetime thermal fuse on the resistors.. If relay is dead resistors heats and kills it series connection to the load by melting thermofuse once... Before the relay and resistor an inline glas tube as slow 16A is needed as well... No ntc/ptc disc that ages and sets it self on fire, just metal case grounded and fuse will do better job maybe?
In most cases you can also use a normal relay and connect it's coil parallel to your load. Once the current draw drops the voltage rises and the relay closes. You may have to match your resistor and relay a little. But that's probably the easiest way to limit inrush current fairly efficiently (the only power loss being the holding current of the relay).
not gonna lie but this is exactly something I need to make for about 10-15 past days and here you go, guess a trip to DIY electronic store is required!
Your explanation of the inrush current on an inductive load isn't quite correct; what you describe is more like how a capacitive load acts than anything. For motors, the inrush current is due to the mechanical inertia; when first turned on, the motor isn't spinning and so there's no back EMF to limit current flow, but once it gets up to speed the back EMF lowers the effective voltage across the coils. For transformers, it's actually caused by connecting *at the zero crossing*, which can even saturate the core and cause its inductance to fall dramatically, further spiking the current upward. I'm not sure how your zero-crossing switch fixed the problem, actually... The ideal moment to switch a purely inductive load on is actually at the peak of the AC supply voltage. Your transformer may have a significant capacitive component to it, perhaps? Could be a parallel capacitor added for power factor reasons; I doubt it's parasitic winding capacitance.
Sorry @GreatScott!, your theory behind switching inductive loads at voltage zero crossing is wrong, it should be switched at current zero crossing (that happens at peak voltage), look around for documentations, the reason behind breaker tripping should be elsewhere. (Maybe Arduino and/or SSR are taking up time and they are switching correctly? Or is it simply because of the soft-start that it works?) Try to capture also the mains voltage with the oscilloscope in sync with the surge current and we will see. Thanks!
Sorry to be the one to tell you that, but your calculations about the inrush current are not really correct. The inrush current in transformers is caused by the magnetic remanence (magnetization of the core). When you turn off a transformer it usually has a magnetization in the core afterwards. If you then connect the transformer again, this "offset" can drive the transformer into saturation, which creates losses in the core. This offset only lasts after a couple of cycles, until its gone. This explains the unreliable tripping of the circuit breaker, since the stored magnatization is very much dependent on the timing when it was shut off during a cycle. The DC resistance only limits the inrush current into the parasitic capacitance of the transformer (brings down the quality factor Q of the whole system). So technically your not wrong, just not the whole story. It's probably rather hard to find a balance between entertainment and bone dry mathematical theory. But all in all a great well explained video as usual. :) Keep it up.
Finally somebody who knows what they are talking about! My simple solution is a resistor and a timer relay. The resistor needs to be small enough to drive current into the primary inductance but large enough to limit the current if it saturates. The time delay need only be a few cycles, so 100-200ms is sufficient.
Thanks for admitting the error. Thats news to me, since I don't normally work with inductors, so zero point made sense originally. Thanks to all the Commenters for the additional explanations!
On the large DC power supplies we design at work, we just use a 20 ohm NTC inrush current limiting resistor, and a contactor that bypasses it after the main bulk caps are charged. Which could be done with an RC timer + Comparator
You know what you need to make? An infinite-loop, programmable, loss-suppressing contest repeater tuned for an MSO ScopeNmultimeter that's compatible with a pre-century reserved shelf-gap filled with transient notta-waves that keeps inducing negative optical feedback?
Did a similar thing for the mystrey rating variable voltage transformer I got. Unknown what the ratings are, but it seems fine to handle 15A output current. Just used a high power resistor in the input line, 4 10R 10W in parallel, and with a 100C thermal fuse between them, with a relay fed from a simple resistor, bridge rectifier and capacitor smoothing, so the relay turns on around 5 cycles after switch on and shorts out the resistors. Also used a winding tap that gave 60VAC (lowest one on it) that fed a half wave rectifier, smoothing capacitor, and then a 47V zener regulator, that in turn feeds a 33V zener regulator, providing power to a LM723 set to provide an output of 6V, that can sink 10mA if needed. This is the one side of a simple voltmeter, so that the scale starts at 60VAC, and goes all the way to 360VAC, which is the range of the output at maximum. All made using parts either repurposed or lying around, including the box, which was the case the variable voltage transformer came in, just with a coat of paint, and the assorted outlet sockets attached. 60VAC to 360VAC because the surplus meter I had came with a 1mA FSD coil, and was marked with 0 to 30 already, so quick to simply use a black drawing pen to relabel the numbers, after a painting over of the originals.
I rly wanna thank you for everything!!! I learned so much from your Videos! I started watching your Videos like back in 2017 and learned almost everything about Electronics including English. Please never stop making Videos☺️☺️
May I suggest that you add a relay to bypass the triac after completion of the load soft starting in order to limit the power loss in triac and saving it from overheating. And you might want to add a variable resistor to enable the user to set the soft starting phase duration. Great job
Informative video. I didn't realize that Solid State Relays have a zero crossing circuit. This should fix my Siglent Bench PS from tripping the UPS on my router.
Great video, although using a microcontroller, a triac and a lot of other components seems like a massive overkill. A NTC thermistor with a relay that bypasses the thermistor after around 1-2 seconds would do the job as well, but with way less components. But of course, with the microcontroller route you learn a lot in the process.
Thanks Great Scott!, I’ve learned so much from your videos. I’m having troubles with this one though, and as I’m building the circuit I’m hoping someone can help me with a few questions. I’ve researched using the timer registers to the point I feel I understand it, how clock cycles dependant on the prescaler settings correlates to actual time. I believe I’ve confirmed that OCR1B = 590 (9.44ms prescaled 256) is based on 50 Hz mains power cycle. If I understand the sketch correctly, everytime mains crosses 0 VAC on a rising portion of the main’s cycle, the sketch checks to see if the button was depressed. If the button was depressed the value of OCR1B starts to decrease until i = 587. What I can’t comprehend is why 9.44 ms is the chosen time, as by my math 1/50Hz * ½ a cycle = 10 ms. This would be the time for the mains to go from passing 0 VAC rising to a half cycle later when mains is at 0 VAC falling, and the triac is triggered for the first time. So why 9.44 ms, and not 10 ms, and why holding OCR1B at 3 (590 - i of 587), and not 0? If a person has 60 Hz mains and wants to trigger the triac at the peak of a mains cycle, as has been suggested, OCR1B would start at 260 (1/60Hz * ¼ = 4.17 ms) to trigger the triac for the first time, at the peak of the cycle after passing 0 VAC rising. In the original sketch the triac is triggered over half a cycle from 0 VAC falling to 0 VAC rising. If triggering the triac at the peak of the cycle would OCR1B be adjusted to still equal half a cycle duration, but starting from the peak, this time peak to peak which would include passing O VAC falling? I hope I’m correct in my explanations, and my questions are worthy. Thanks!
Newton's Third Law. The torque that turns the rotor has to push against something, which in this case was the frame of the motor, causing it to move as well.
@@thereynolds2725 Agree. But it looked like the same jolt as an across-the-line start of a motor that's not bolted down. Didn't expect it would be noticeable under soft start especially because the armature would have only had turned a few degrees to align itself. Cant make out and shaft movement from the video.
by using a primary to secodary winding in a toroidal transformer and make the output winding voltage divided chould prevent the spike all together making the output winding like a auto transformer. the primary winding would have nearly zero core loss in a toroidal transformer so there woudl be no initial overload by powerup.
Large transformers are always a problem for standard house breakers, if permanently installed the type of breaker can be changed to a C or even D, but then the tripping time has to be correlated with local disconnection times as dictated by local law. If the appliance is portable, then a soft start (cheaper than a full inverted) can be used to reduce the inrush, many are available commercially and are mainly used for motors. I am a little concerned that the motor jumps into life at the button press then returns to slowly accelerating, not something one would expect from an off the shelf solution!
As you mentioned, even a zero volt point initial turn-on is not always curing the problem. The reason is the remanent magnetism in the transformer core. If the magnetism left in the core at previous turn-off matches the direction of the start time voltage growth slope, there still is a big inrush. The theoretical analysis considers the fact that there is a 90 degree phase shift between the voltage and the current (on the primary). So, at the positive going zero crossing of the voltage, the usual transformer operation condition would have actually the negative peak current at this moment. You don't have that, if you make the primary connection at the voltage zero crossing moment. Your gradual increase of the voltage is indeed the way to prevent the breaker popping inrush. As funny as it may sound, if the transformer was turned OFF gradually, you would get near zero remanence and then the optimum time to turn it again on would be near the voltage peak (either positive or negative) !!! Too bad, you generally cannot know the state of the remanence, nor force it to any known state. I hope this explanation can serve other people who may suffer from the same issue and are wondering why.
ah, the old "dim bulb" great for slowly powering up something like an old radio or TV that has wax caps in them. light bulb lights up brightly you shut it the hell off.
Nice project I have a couple questions. Does this work for 120V? I've seen a couple other soft start projects which claim that a relay is a more reliable long term solution than a triac? Also would like to see a soft start which starts up automatically when the load device is powered on. Thanks.
You probably have a core saturation problem during the initial transient, this most likely to occur when you start close to a zero crossing, if you start close to a voltage maximum you flux peak will be lower.
Wieso so kompliziert? Du kannst einfach ein NTC verwenden und ein RC Glied das ein Relais über den NTC nach paar Sekunden schaltet. (NTC ist kalt beim einschalten wird selbst warm bei 230v und dadurch leitender). War schon so in den Audio Verstärkern in den 80er Jahren und die funktionieren heute noch😉.
or put one NTC in series and short it with relay contact with one transistor with RC constant in base, 5 parts... arduino in such a simple project ? what is gonna be next, how to turn on house bulb using arduino ?
6:22 It's the opposite, it's better that the voltage is high at the beginnig, it's a complex fact but basicly because in inductance courrente are 90° late than voltage so maxmum voltage = minimum courrent.
Why not introduce a input resistor and bypass it after a set amount of time. That would not waste power all that much and prevent high inrush as well. Alternatively you could use a MCB with a higher TCC curve.
Since you have a 3-phase induction motor lying around, why don't you make a soft-starter for that as well? Or do a DIY or BUY video for a 3-phase induction motor soft-starter?
Note that an autotransformer is just one with a single shared winding, yours is a variable autotransformer. Also, the input voltage is only across part of the winding, which is what gives it the ability to output a higher voltage than the input.
Great project! For softstart I used a "Finder DPDT Multi Function Timer Relay" set to 1 second with a thermistor over the "off" position. The first second you power up the thermistor heats up then the relay clicks to the "on" position so the attached device is powered directly. The switching time seems fast enough so far.
@@gblargg yes but they will become very hot, not ideal for long-term use. Also when you switch off and on before the ntc had time to cool off the resistance is still too low. This is why I used a time relay in combination with an ntc.
@@o0arend0o this solution still has the lingering heat from previous on time. Unless chosing time delay longer allowing ntc to cool down enough to make a difference.
@@muthumanikya1847 the ntc only heats up for less then a second when power is switched on for first time. The rest of the time the ntc cools off while the device is on.
Isn't it better to simply connect a resistor in series with the inductive load that is later bypassed by a relay (exact same model as your DC softstarter, using a small smps to generate a low DC voltage)?
@@greatscottlab mine is simpler, you don't have to program an arduino. It is also a lot cheaper, and doesn't dissipate anything in a triac. 230V 3000W through such a triac is probably going to blow up.
I have a 15A variac behind a 15A capable isolation transformer (isolated and adjustable voltage is sometimes handy). All in parallel with a second lower power isolation transformer. I use a high power resistor and a timer relay. The inrush current is limited by a simple resistor in series on the mains side and after a second or two, the resistor is bypassed by a timer relay. Works a treat.
If you decide to use your small variac again you might want to use ring terminals and nuts instead of directly soldering wires. It should be safer and more durable.
Great job! Only question is what is the motor jump at 11:17. Just at the button push there is a jump and than slow speed up. Looks like a short maximum peak at the initial. It shoudn't be. Am I right? Did you fix it?
When you pushed the button at 11:16, why did the motor jump before it started spinning? I ask, because I also designed a soft start some years ago (it's still in my table saw), and it does the same thing - not always, but frequently. Probably depends how far it is from zero crossing when I switch it on. I could investigate it, but it's a pig to get at, so I have taken the easy option and left it. It's mentioned at ruclips.net/video/9HDLc40Gmac/видео.html
could you just add a furnace blower motor capacitor in line to the autotransformer to resolve the quick power surge from pulling all the power from the breaker?
Only feedback I'd have for you would be to move those LEDs, put the red one near the back, as it indicates power in, move the other one to out outlet side and make it light up red when the button is pushed and go green when the sequence is done (those looked to be three or 4 legged LEDs, so either dual color or maybe RGB?). This way it is much easy to know what the state of the thing is by a quick visual glance, and therefore a bit safer to use.
I like your design. I had similar problem for putting an AC capacitor in parallel with load to correct for power factor. When you want to connect the capacitor to AC line you might get a huge "inrush current", which is similar to “locked rotor current” in motors. -- By the way you can measure inrush current with many types good multimeters these days, and you do not need to use oscilloscope. -- I solved the problem by putting a small inductor in series with the capacitor, so the total impedance is still capacitive at the frequency of operation. To find out how much inductor you need you should following laws: Current in the inductor is continues and does not change instantaneously. Voltage in a capacitor is continues and does not change instantaneously. Rate of change of inductor current is d(I)/d(t) = V/L So at worse time is when you connect at the peak of the cycle: Vp/L is the maximum rate of change . To find out what is maximum current you have to integrate d(I)/d(t) for 1/4 of frequency cycle. so I=Integral of Vp/L Cos(2 pi f t) d(t) where t from 0 to 1/(4f) where f is frequency of the AC line, Vp = SQRT(2) Vrms of the line. In your case you could have put a huge AC capacitor in series with your auto transformer. Other solutions I found out in the web: Using NTC thermistors: When power is first applied, the Negative Temperature Coefficient thermistor is cold, so it has a high impedance and effectively limits the current. The NTC thermistor heats up very rapidly due to its own losses when it used in series with the load, and as result its resistance decreases. *** Problem ambient temperature dependent *** Using fixed resistors After the electrolytic capacitors on the input of the power supply are charged, the resistor is bypassed. Trailing edge phase dimming:
Why make it so complicated? I made softstarter to my 10kVA transformer by pre magnetizing the coil through power resistors for some milliseconds before directly powering the coils. Very simple and works.
I would like to see you create single sided PCBs (use wire links on the component side where necessary) - for all of your projects rather than use matrix perf board.
@@flflflflflfl It makes better sense. Now I am not saying that I am an engineer or even a novice one, but I have been floating around the scene armed with an soldering iron, scope and other necessary tools for over 30-years now. Surely my opinion is not entirely worthless. ☆Magic☆
@@flflflflflfl I owe you no explanation other that my opinion that reflects my efforts with academia and actual real-world experiences. Now if you want me to build you a rocket, then I am going to require a lot of money to keep my Wife satisfied. Thank you. Please come again and have a ☆Magical Day☆
A commercial soft start is about $300. It works automatically (eg for air conditioners), i have a bunch of inductive loads that I would love to put soft starts on. $300 each is a bit much. Is there a product opportunity for you here? Either a kit, or a PCB? Example: www.gonewiththewynns.com/product/air-conditioning-off-grid
Thanks so much for posting this! I've needed an AC soft starter for quite some time but was hesitant to spend the $300+ that most places want for a decent one. Thought for sure the circuit would be much more complicated than it is! I'll be using it for an air conditioner instead of an autotransformer.
This is the worst decision. «The severity of the Switching Transients in Transformer is related to the instant when the voltage wave is switched on; the worst conditions being when the applied voltage has zero value at the instant of switching», link: www.eeeguide.com/switching-transients-in-transformer/ What you did with your zero-crossing schematics is just adding triac finite resistance in serial with transformer. If you want to avoid current inrush in inductance, you must: 1) turn it on at peak voltage; 2) turn it off at zero current to avoid residual core saturation. The worst scenario nonetheless not only switching transformer at zero voltage moment but with the flux reversal of the beforehand saturated core which can result in 10-100 numbers of inrush current. That possibly describes your time-to-time circuit breaker tripping.
GREAT video!!! I think for a normal house.. all circuit breaker are of for normal devices..!!! you had this problem but... not every person has a autotransformer at home!!! ajaja
Why such a complicated circuit? You could've done something simple, like adding a resistor in series with the transformer and adding a relay which shorts the resistor when it closes, if you connect the coil parallel to the transformer (so it only closes once the current draw goes below a certain level and the voltage rises). There you go, same thing achieved with two components... Of course you can also do it with one component by just using an NTC. But with a higher power loss and a cool down time.
7:11 that explanation makes no sense to me. The transformer should be a an impedance which is an inductor (Lm) in series with another leakage inductance (Lr) finally in series with a resistive component being the resistance of the wires. The current waveform at steady state is of course going to lag the voltage, the amplitude of which (apparent current) easily found with phasor analysis. But with a step voltage in to an inductor should result in i=0 initially and build up from there, an inductor current is proportional to volt-seconds, the integration of volts with respect to time. So why is there inrush to an unloaded transformer? Is the inter-winding capacitance playing a significant role or something? Motors are quite complex loads so I can understand inrush for them
I had the same question/concern. The inductance will resist the change in current as long as the core isn't saturated. Interwinding capacitance would allow some initial current spike. ..or maybe there is an input filter cap at the autotransformer? Or maybe it depends on the load attached to the autotransformer? This might be an idea for a follow-up video.
@@SkyhawkSteve My guess is there is a sizable parallel resonant capacitor meant to improve the power factor by resonating with the magnetizing inductance. Such a capacitor would certainly draw a massive inrush when power is applied at the peak of mains, limited by series inductance and resistance of the house wiring and impedance of the grid.
A transformer is modelled as an inductor Lm *parallel* to a series combination of Lr and coil resistance r. The point you raised regarding current being proportional to volt second is true only as long as the inductor is in its linear range of operation(not saturated). When a transformer is switched on at the zero volt point, the flux set up theoretically rises upto 2 times the peak value of normal operating sinusoidal flux. Most transformers operate near to its saturation to optimise iron usage and hence when flux can go upto 2 times the peak, the core undergoes heavy saturation and depending on the saturation curve of the core, the transformer will draw a very large current to generate that flux.
@@manjunathvishnumoorthy7479 Oh, I see. Because of the "integral windup" from a previous cycle the definite integral from t=0 to the first zero crossing would indeed see a current that is in fact double, that makes sense! Edit: Oh and I get your point about why you mention saturation, that's something I'll have to keep in mind thanks for sharing! --On the transformer model I was assuming open-circuit secondary and thus the 3 series impedances (really an open circuit transformer is just an inductor or ballast) there is also some distortion and core saturation + hysteresis losses causing a what I recall was 3rd and 5th order harmonics and higher than expected peak currents, sure.--
Not real sure of how to fix it, but from what I understand is that the core material has memory. When turned off all the lines get lined up in one direction, and if you turn the power back on, and the phase is 90 deg. from the lines of force, it acts like a short circuit. It is kind of hard to explain, so I hope it is a little bit clearer. Michael from Washington State, near Seattle.
Nice video! One recommendation, make sure your diy pcb can handle up to 3000 watt or preferred 3600 watt. The small screw connectors are not designed for this loads.
Tripping circuit breakers? Ah I just remembered when I was 10yo I plugged in a copper wire in socket assuming it was nichrome wire to get heated Suddenly I heard a trip and tv was off Conclusion:-great scott teaches very good
Similar story about 10 year old me: I figured that if I shorted out an AC/AC adapter, it would short out the mains and blow the fuse. When I finally felt brave enough to try out my new found super powers, it just blew the fuse in the adapter without even making a noise or smoke.
Will this soft starter work for a solar setup? Starting a motor running off an inverter and solar panels, the startup current is too high and trips the inverter.
You should have used a mechanical relay to bypass the TRIAC after softstart. The would have allowed you to use a smaller heatsink. And at 3000W, the amount of current flowing through the TRIAC might heat it up a lot, so having the circuit inside a closed housing with no airflow might cause it to overheat.
Hi great scott, another amazing video !! Just a little question, at the middle of your paper schematic i have noticed u wrote 325V ... that's confusing me ... if we have 230-240 V AC isnt the max value +- 230 V ? why 325 ? could be a refuse thinking on a 3 phase system, where u have such voltrage between phases ?
Could this be used for simple ac loads like a non computerized refrigerator? My fridge has a terrible surge on it. 2200 watts, makes my inverter flip out and the lights dim.
Hi @GreatScott, I have two questions please. 1) Why is the optocoupler's anode connected via two parallel resistors (R6 and R2) both 220k. Would just one 110k not work? 2) Is the black filament the galaxy black from Devil Design :-) ?
I need one for my fridge. Every once in a while when the compressor turns on the amp draw pulls over 40a but only for a second but long enough to trip the breaker, there after it only pulls 2a. The fridge is only 2 years old.
I like your videos, and you admit to making a big mistake here. I think you should re-edit and re-upload it with a breakdown. Really not complaining, or putting you down, we all screw up, but your channels is very educational and this will confuse people. I've had similar problems and used a resistor to limit inrush and a contractor to bypass resistor and connect output at the same time when the transformers output voltage comes up.
The older types of circuit breakers had a time delay characteristic that allowed for a heavy load to start up. These new "politically correct" breakers apparently break in one cycle. This can be one hell of a nuisance requiring all kinds of "hacks" (like shown in the video) just to live. I'll stick with my 50 year old type breakers that can start a big air conditioner without tripping, yet a few seconds more reliably shut the current off.
Seems like I made a rather big mistake in the video. My transformer should not be switched on during the zero crossing point but at the voltage peak in order to decrease the current flow. While recording the current waveforms on the oscilloscope, everything appeared good to me and I am not sure why my circuit breaker stopped tripping when I switched at the zero crossing point. But many viewers pointed me in this direction and after doing research, it seems to be correct. Sorry about this. But if you want to build the softstarter you can still try it out whether it works for you and if not then you can always change the Arduino code in order to switch at a different time point. Sorry again for the inconvenience. I do make mistakes from time to time since this is a one man production. Stay creative :-)
Can you sir please explain me why to turn it on at the peak and not at zero crossing point?
@@aks8403 ELI the ICE man
@@aks8403 You have to change the timer value OCR1B. In my case I firstly used 590 which is around 9.44ms after the zero crossing point. 625 would be 10ms. So simply use half of 625 so around 313 in order to get a 5ms value at which point a voltage peak should appear after the zero crossing point.
it's ok you are a legend of electronics for me
Das war wieder ein sehr priffiges Video! Gruß aus Karlsruhe in Baden! :)
Tripping circuit breaker? Hah, I don't have such weakness with my 30-years-old house wiring. The circuit resistance is high enough to limit any inrush current.
Well.....that would also be a solution
@@403_Tuna aluminium wire and ceramic conectors and large swich board for a hause ?
My 80 year old cloth wiring has the same effect. Built in protection :)
@@403_Tuna the is thousands more over the world or even millions
@@alangolab6657 Same with our house
Easier and cheaper to change the breaker to type C time curve which is for high startup loads.
Also possible. But not always the best solution.
That was my first thought, but i guess it isn't a viable solution if transformer shares breaker with other devices. Type C is really slow and if wires aren't properly sized it can cause all sorts of problems.
@@cekpi7 true in some cases, but if IK is in the right range, it shoul‘d be ok
Might be true. But less fun😜
@@greatscottlab Yes, why do it simple when you can make it complicated 🤔
I had a similar issue with a 3000 watt transformer for reducing power from 220 to 120 volts. I later learned that the type of circuit breaker used makes a difference. There are some breakers rated for inductive loads and some for resistive loads. Most home breakers are of the resistive loads found in most homes. Replace breaker with an inductive rated breaker. Did this and no problems now.
The current into an inductor equals the integral of the voltage accross it, in formula: i = 1/L.integral(Vdt). If you switch it on at the (rising) zero crossing the voltage remains positive for 180 degrees. This causes the magnetizing current to rise to approx. 2x its normal value. However the inductor core is not made for this 2x magnetizing current and thus the core will saturate. This causes the coil inductance to drop, which increases the current even (a lot) more. If you switch at the maximum voltage, the current also starts to rise, following the integral relation, but after 90 degrees the voltage polarity reverses and the current starts to go down again. In this way the current starts of nicely lagging the voltage by 90 degrees, as it should be in the stationary situation, with peaking and tripping the breaker.
Best regards, Ben.
"If you switch it on at the (rising) zero crossing the voltage remains positive for 180 degrees." .......
But in normal operation, the voltage remains positive for 180 degrees and then becomes negative again for 180 degrees. What's the difference now?
@@Thomas72B The difference is exactly what you say: in normal operation a positive periode is always preceded by a negative period, and visa versa. An inductor stores current in a similar way as a capacitor stores voltage. Therefore the history is important. But at startup there is no history (current) stored in the inductor. When you then start with a full positive (or negative) period, the current becomes too high. This is only a temporary problem, after a number of periods the current normalizes, due to the series resistance of the inductor.
Regards, Ben.
@@beneindhoven3424 Not always ! If you put the plug in the socket you cannot know where the sine curve is. 0 degrees, 130 degrees or 174 degrees you don't know .... That's why I still don't quite understand where the problem is supposed to be.
@@Thomas72B Yes, that’s correct, you never know where you start when you connect the plug. That’s why the circuit breaker not always trips, but only sometimes. Switching in near the zero crossing is the worst case situation. Having said that, I have to admit that I don’t understand why Great Scott’s tests at zero cross switching didn’t trip the breaker. I tried this same experiment a few months ago, and could clearly see the current peaks are maximum at the zero crossing. And minimum when switching at the voltage peaks. Regards, Ben.
Ohhhh.... Brilliant comment! I think it's also just answered a question I've had about adding energy into a LC tank. Thanks for sharing the knowledge.
Edit: Yes! I've been trying to get my head around this for days. You're a catylist for understanding good sir. Double thanks!
Another much easier and cheaper way which I used succesfuly in the past was a 1s timed relay in parallel with a high power resistor with just enough ohms to limit the current to under 16A.
Both of that in series to the load will limit the current for 1s and then the relay shorts the resistor.
Worked like a charm :-)
good idea, but using an ntc instead of a power resistor is better, since it can prevent from burning down your house in case of failure of the relay/connection..
@@nicolaswannen1743 can you get ntc with lots of watt dissipating ? Better to use onetime thermal fuse on the resistors.. If relay is dead resistors heats and kills it series connection to the load by melting thermofuse once...
Before the relay and resistor an inline glas tube as slow 16A is needed as well...
No ntc/ptc disc that ages and sets it self on fire, just metal case grounded and fuse will do better job maybe?
In most cases you can also use a normal relay and connect it's coil parallel to your load. Once the current draw drops the voltage rises and the relay closes. You may have to match your resistor and relay a little. But that's probably the easiest way to limit inrush current fairly efficiently (the only power loss being the holding current of the relay).
not gonna lie but this is exactly something I need to make for about 10-15 past days and here you go, guess a trip to DIY electronic store is required!
I hope it works for you :-) But be careful when working with mains voltage. Thanks :-)
don't forget to check that you have adequate mains voltage by doing the good old tongue test
Your explanation of the inrush current on an inductive load isn't quite correct; what you describe is more like how a capacitive load acts than anything. For motors, the inrush current is due to the mechanical inertia; when first turned on, the motor isn't spinning and so there's no back EMF to limit current flow, but once it gets up to speed the back EMF lowers the effective voltage across the coils. For transformers, it's actually caused by connecting *at the zero crossing*, which can even saturate the core and cause its inductance to fall dramatically, further spiking the current upward. I'm not sure how your zero-crossing switch fixed the problem, actually... The ideal moment to switch a purely inductive load on is actually at the peak of the AC supply voltage. Your transformer may have a significant capacitive component to it, perhaps? Could be a parallel capacitor added for power factor reasons; I doubt it's parasitic winding capacitance.
Hmmmm I will have to think about that. Not quite sure....
@@greatscottlab It's a real puzzler; I wonder what's inside that autotransformer.
Sorry @GreatScott!, your theory behind switching inductive loads at voltage zero crossing is wrong,
it should be switched at current zero crossing (that happens at peak voltage),
look around for documentations, the reason behind breaker tripping should be elsewhere.
(Maybe Arduino and/or SSR are taking up time and they are switching correctly? Or is it simply because of the soft-start that it works?)
Try to capture also the mains voltage with the oscilloscope in sync with the surge current and we will see. Thanks!
Thanks for the feedback. I just did lots of research and it seems you are right. I added a pinned comment that explains it. Sorry.
Greatscott " warning do not touch its high voltage "
Electroboom " Hmm this is high voltage wire 🎆 yeeeeaaawoooouch fu** fu** fu** I am an idiot "
I saw he made use of a FOOL BRIDGEE RECITFYERRRR >:U
Electroboom would also try to lick the transformer xD
Sorry to be the one to tell you that, but your calculations about the inrush current are not really correct.
The inrush current in transformers is caused by the magnetic remanence (magnetization of the core).
When you turn off a transformer it usually has a magnetization in the core afterwards. If you then connect the transformer again, this "offset" can drive the transformer into saturation, which creates losses in the core. This offset only lasts after a couple of cycles, until its gone.
This explains the unreliable tripping of the circuit breaker, since the stored magnatization is very much dependent on the timing when it was shut off during a cycle.
The DC resistance only limits the inrush current into the parasitic capacitance of the transformer (brings down the quality factor Q of the whole system).
So technically your not wrong, just not the whole story. It's probably rather hard to find a balance between entertainment and bone dry mathematical theory.
But all in all a great well explained video as usual. :) Keep it up.
Finally somebody who knows what they are talking about!
My simple solution is a resistor and a timer relay. The resistor needs to be small enough to drive current into the primary inductance but large enough to limit the current if it saturates. The time delay need only be a few cycles, so 100-200ms is sufficient.
That was very cool that the SSR has a zero crossing bult-in, however make sense when you think about what it does. Very cool solution.
Thanks for admitting the error. Thats news to me, since I don't normally work with inductors, so zero point made sense originally. Thanks to all the Commenters for the additional explanations!
On the large DC power supplies we design at work, we just use a 20 ohm NTC inrush current limiting resistor, and a contactor that bypasses it after the main bulk caps are charged. Which could be done with an RC timer + Comparator
You know what you need to make? An infinite-loop, programmable, loss-suppressing contest repeater tuned for an MSO ScopeNmultimeter that's compatible with a pre-century reserved shelf-gap filled with transient notta-waves that keeps inducing negative optical feedback?
I love it :-)
I am on it ;-)
Cool
Bazinga!
Brain hurty
Did a similar thing for the mystrey rating variable voltage transformer I got. Unknown what the ratings are, but it seems fine to handle 15A output current. Just used a high power resistor in the input line, 4 10R 10W in parallel, and with a 100C thermal fuse between them, with a relay fed from a simple resistor, bridge rectifier and capacitor smoothing, so the relay turns on around 5 cycles after switch on and shorts out the resistors.
Also used a winding tap that gave 60VAC (lowest one on it) that fed a half wave rectifier, smoothing capacitor, and then a 47V zener regulator, that in turn feeds a 33V zener regulator, providing power to a LM723 set to provide an output of 6V, that can sink 10mA if needed. This is the one side of a simple voltmeter, so that the scale starts at 60VAC, and goes all the way to 360VAC, which is the range of the output at maximum. All made using parts either repurposed or lying around, including the box, which was the case the variable voltage transformer came in, just with a coat of paint, and the assorted outlet sockets attached. 60VAC to 360VAC because the surplus meter I had came with a 1mA FSD coil, and was marked with 0 to 30 already, so quick to simply use a black drawing pen to relabel the numbers, after a painting over of the originals.
I rly wanna thank you for everything!!!
I learned so much from your Videos! I started watching your Videos like back in 2017 and learned almost everything about Electronics including English. Please never stop making Videos☺️☺️
Thanks for the feedback :-) I am glad that I could help you :-)
I’ve faced the same problem with my 10Kw Variac, I’ll try your solutions ASAP, Thanks 🙏
May I suggest that you add a relay to bypass the triac after completion of the load soft starting in order to limit the power loss in triac and saving it from overheating. And you might want to add a variable resistor to enable the user to set the soft starting phase duration.
Great job
Informative video.
I didn't realize that Solid State Relays have a zero crossing circuit.
This should fix my Siglent Bench PS from tripping the UPS on my router.
Great video, although using a microcontroller, a triac and a lot of other components seems like a massive overkill. A NTC thermistor with a relay that bypasses the thermistor after around 1-2 seconds would do the job as well, but with way less components. But of course, with the microcontroller route you learn a lot in the process.
Thanks Great Scott!, I’ve learned so much from your videos.
I’m having troubles with this one though, and as I’m building the circuit I’m hoping someone can help me with a few questions. I’ve researched using the timer registers to the point I feel I understand it, how clock cycles dependant on the prescaler settings correlates to actual time.
I believe I’ve confirmed that OCR1B = 590 (9.44ms prescaled 256) is based on 50 Hz mains power cycle. If I understand the sketch correctly, everytime mains crosses 0 VAC on a rising portion of the main’s cycle, the sketch checks to see if the button was depressed. If the button was depressed the value of OCR1B starts to decrease until i = 587.
What I can’t comprehend is why 9.44 ms is the chosen time, as by my math 1/50Hz * ½ a cycle = 10 ms. This would be the time for the mains to go from passing 0 VAC rising to a half cycle later when mains is at 0 VAC falling, and the triac is triggered for the first time.
So why 9.44 ms, and not 10 ms, and why holding OCR1B at 3 (590 - i of 587), and not 0?
If a person has 60 Hz mains and wants to trigger the triac at the peak of a mains cycle, as has been suggested, OCR1B would start at 260 (1/60Hz * ¼ = 4.17 ms) to trigger the triac for the first time, at the peak of the cycle after passing 0 VAC rising.
In the original sketch the triac is triggered over half a cycle from 0 VAC falling to 0 VAC rising. If triggering the triac at the peak of the cycle would OCR1B be adjusted to still equal half a cycle duration, but starting from the peak, this time peak to peak which would include passing O VAC falling?
I hope I’m correct in my explanations, and my questions are worthy.
Thanks!
I wish Great Scott would answer !!!
brillianT - just one question... why does the motor jump at the end before the shaft even starts turning?????
I think that was just the initial moment when the coils were magnetized and the rotor aligned with the field.
Newton's Third Law. The torque that turns the rotor has to push against something, which in this case was the frame of the motor, causing it to move as well.
@@thereynolds2725 Agree. But it looked like the same jolt as an across-the-line start of a motor that's not bolted down. Didn't expect it would be noticeable under soft start especially because the armature would have only had turned a few degrees to align itself. Cant make out and shaft movement from the video.
by using a primary to secodary winding in a toroidal transformer and make the output winding voltage divided chould prevent the spike all together making the output winding like a auto transformer. the primary winding would have nearly zero core loss in a toroidal transformer so there woudl be no initial overload by powerup.
Thanks for this, always wondered what soft starters were for. This was a great explanation!
Love the project with a practical example built right in
oooh JLC is back
Large transformers are always a problem for standard house breakers, if permanently installed the type of breaker can be changed to a C or even D, but then the tripping time has to be correlated with local disconnection times as dictated by local law. If the appliance is portable, then a soft start (cheaper than a full inverted) can be used to reduce the inrush, many are available commercially and are mainly used for motors. I am a little concerned that the motor jumps into life at the button press then returns to slowly accelerating, not something one would expect from an off the shelf solution!
As you mentioned, even a zero volt point initial turn-on is not always curing the problem. The reason is the remanent magnetism in the transformer core. If the magnetism left in the core at previous turn-off matches the direction of the start time voltage growth slope, there still is a big inrush. The theoretical analysis considers the fact that there is a 90 degree phase shift between the voltage and the current (on the primary). So, at the positive going zero crossing of the voltage, the usual transformer operation condition would have actually the negative peak current at this moment. You don't have that, if you make the primary connection at the voltage zero crossing moment. Your gradual increase of the voltage is indeed the way to prevent the breaker popping inrush. As funny as it may sound, if the transformer was turned OFF gradually, you would get near zero remanence and then the optimum time to turn it again on would be near the voltage peak (either positive or negative) !!! Too bad, you generally cannot know the state of the remanence, nor force it to any known state. I hope this explanation can serve other people who may suffer from the same issue and are wondering why.
An old TV mechanic I knew had a soft start circuit made with 200W light bulb and a switch 😀
ah, the old "dim bulb" great for slowly powering up something like an old radio or TV that has wax caps in them. light bulb lights up brightly you shut it the hell off.
Nice project I have a couple questions. Does this work for 120V? I've seen a couple other soft start projects which claim that a relay is a more reliable long term solution than a triac? Also would like to see a soft start which starts up automatically when the load device is powered on. Thanks.
You probably have a core saturation problem during the initial transient, this most likely to occur when you start close to a zero crossing, if you start close to a voltage maximum you flux peak will be lower.
Any more information on this? Would it work to have an SCR or something wait for a peak before turning on?
Wieso so kompliziert? Du kannst einfach ein NTC verwenden und ein RC Glied das ein Relais über den NTC nach paar Sekunden schaltet.
(NTC ist kalt beim einschalten wird selbst warm bei 230v und dadurch leitender). War schon so in den Audio Verstärkern in den 80er Jahren und die funktionieren heute noch😉.
or put one NTC in series and short it with relay contact with one transistor with RC constant in base, 5 parts... arduino in such a simple project ? what is gonna be next, how to turn on house bulb using arduino ?
@Tiago Ferreira Yes I know, I have already used that many times. Is super common because it's simple and bullet proof.
6:22 It's the opposite, it's better that the voltage is high at the beginnig, it's a complex fact but basicly because in inductance courrente are 90° late than voltage so maxmum voltage = minimum courrent.
Why not introduce a input resistor and bypass it after a set amount of time. That would not waste power all that much and prevent high inrush as well.
Alternatively you could use a MCB with a higher TCC curve.
Since you have a 3-phase induction motor lying around, why don't you make a soft-starter for that as well? Or do a DIY or BUY video for a 3-phase induction motor soft-starter?
Note that an autotransformer is just one with a single shared winding, yours is a variable autotransformer. Also, the input voltage is only across part of the winding, which is what gives it the ability to output a higher voltage than the input.
Great project! For softstart I used a "Finder DPDT Multi Function Timer Relay" set to 1 second with a thermistor over the "off" position. The first second you power up the thermistor heats up then the relay clicks to the "on" position so the attached device is powered directly. The switching time seems fast enough so far.
There are NTC thermistors that have an initial resistance then heat up quickly and lower the resistance, made just for this task of soft-starting.
@@gblargg yes but they will become very hot, not ideal for long-term use. Also when you switch off and on before the ntc had time to cool off the resistance is still too low. This is why I used a time relay in combination with an ntc.
@@o0arend0o Good points. They're probably used for lower-current devices like switching supplies.
@@o0arend0o this solution still has the lingering heat from previous on time. Unless chosing time delay longer allowing ntc to cool down enough to make a difference.
@@muthumanikya1847 the ntc only heats up for less then a second when power is switched on for first time. The rest of the time the ntc cools off while the device is on.
Isn't it better to simply connect a resistor in series with the inductive load that is later bypassed by a relay (exact same model as your DC softstarter, using a small smps to generate a low DC voltage)?
Both solutions have their application. Neither of them is the best solution for every problem.
@@greatscottlab mine is simpler, you don't have to program an arduino. It is also a lot cheaper, and doesn't dissipate anything in a triac. 230V 3000W through such a triac is probably going to blow up.
Which TRIAC did you use for 3kW x 250VAC?
You can see my used triac in the schematic. It was a BTB26 I think...
Upppps. No you can not see it in the schematic. Sorry. But it really was a BTB26
@EarrapeXD ?
You know the day is gonna be great when GreatScott uploads
Ohhh stop it you ;-)
you could get some C16 breakers instead of B16 ones
I like to call that thing a Variac :)
I agree with you but if I remember correctly Variac is a manufacturer's tradename.
May be - I just learned it from Photonicinduction who used it a lot to blow stuff up :))
That's why we research in be sure we are not being scammed at the worst or only misinformed. Good point though.
I've lost count of how many times I got scared with induction motors jumping when they are switched on in practical college classes 11:16 haha
Aha. So it's a motor characteristic, not something caused by the slow-start circuit?
@@RandomNullpointer yes, it is very common if they are not attached to a surface
11:16 ahahaha jumpscare
I have a 15A variac behind a 15A capable isolation transformer (isolated and adjustable voltage is sometimes handy). All in parallel with a second lower power isolation transformer.
I use a high power resistor and a timer relay. The inrush current is limited by a simple resistor in series on the mains side and after a second or two, the resistor is bypassed by a timer relay. Works a treat.
If you decide to use your small variac again you might want to use ring terminals and nuts instead of directly soldering wires. It should be safer and more durable.
This one of your best works. Kudos. I've learned a lot.
Well this circuit was much simpler: www.trafco.rs/softstart-en.php
And does not require an arduino!
Same problem. I just used a thermistor.
Great job! Only question is what is the motor jump at 11:17. Just at the button push there is a jump and than slow speed up. Looks like a short maximum peak at the initial. It shoudn't be. Am I right? Did you fix it?
When you pushed the button at 11:16, why did the motor jump before it started spinning? I ask, because I also designed a soft start some years ago (it's still in my table saw), and it does the same thing - not always, but frequently. Probably depends how far it is from zero crossing when I switch it on. I could investigate it, but it's a pig to get at, so I have taken the easy option and left it. It's mentioned at ruclips.net/video/9HDLc40Gmac/видео.html
could you just add a furnace blower motor capacitor in line to the autotransformer to resolve the quick power surge from pulling all the power from the breaker?
Only feedback I'd have for you would be to move those LEDs, put the red one near the back, as it indicates power in, move the other one to out outlet side and make it light up red when the button is pushed and go green when the sequence is done (those looked to be three or 4 legged LEDs, so either dual color or maybe RGB?). This way it is much easy to know what the state of the thing is by a quick visual glance, and therefore a bit safer to use.
May you do a "DIY or Buy" of an 12-to-230V-Converter?
I like your design.
I had similar problem for putting an AC capacitor in parallel with load to correct for power factor. When you want to connect the capacitor to AC line you might get a huge "inrush current", which is similar to “locked rotor current” in motors. -- By the way you can measure inrush current with many types good multimeters these days, and you do not need to use oscilloscope. --
I solved the problem by putting a small inductor in series with the capacitor, so the total impedance is still capacitive at the frequency of operation.
To find out how much inductor you need you should following laws:
Current in the inductor is continues and does not change instantaneously.
Voltage in a capacitor is continues and does not change instantaneously.
Rate of change of inductor current is d(I)/d(t) = V/L
So at worse time is when you connect at the peak of the cycle:
Vp/L is the maximum rate of change .
To find out what is maximum current you have to integrate d(I)/d(t) for 1/4 of frequency cycle. so I=Integral of Vp/L Cos(2 pi f t) d(t) where t from 0 to 1/(4f) where f is frequency of the AC line, Vp = SQRT(2) Vrms of the line.
In your case you could have put a huge AC capacitor in series with your auto transformer.
Other solutions I found out in the web:
Using NTC thermistors: When power is first applied, the Negative Temperature Coefficient thermistor is cold, so it has a high impedance and effectively limits the current. The NTC thermistor heats up very rapidly due to its own losses when it used in series with the load, and as result its resistance decreases. *** Problem ambient temperature dependent ***
Using fixed resistors After the electrolytic capacitors on the input of the power supply are charged, the resistor is bypassed.
Trailing edge phase dimming:
Why make it so complicated? I made softstarter to my 10kVA transformer by pre magnetizing the coil through power resistors for some milliseconds before directly powering the coils. Very simple and works.
Easier would be, (If possible) just to change the circuit breaker for a type C one.
For the shop yes, but now the whole setup is mobile.
I would like to see you create single sided PCBs (use wire links on the component side where necessary) - for all of your projects rather than use matrix perf board.
why?
@@flflflflflfl It makes better sense. Now I am not saying that I am an engineer or even a novice one, but I have been floating around the scene armed with an soldering iron, scope and other necessary tools for over 30-years now. Surely my opinion is not entirely worthless. ☆Magic☆
@@trentjackson4816 why does it make "better sense"?
@@flflflflflfl I owe you no explanation other that my opinion that reflects my efforts with academia and actual real-world experiences. Now if you want me to build you a rocket, then I am going to require a lot of money to keep my Wife satisfied. Thank you. Please come again and have a ☆Magical Day☆
@@trentjackson4816 ok how much do you need?
You could have just changed the MCB for a 'C' type, which has a characteristic that allows for inductive load inrush current.
A commercial soft start is about $300. It works automatically (eg for air conditioners), i have a bunch of inductive loads that I would love to put soft starts on. $300 each is a bit much. Is there a product opportunity for you here? Either a kit, or a PCB?
Example: www.gonewiththewynns.com/product/air-conditioning-off-grid
Did you discover photoninduction RUclips channel before? You can see monster version of it
This reminded me of photon's videos.
Thanks so much for posting this! I've needed an AC soft starter for quite some time but was hesitant to spend the $300+ that most places want for a decent one. Thought for sure the circuit would be much more complicated than it is! I'll be using it for an air conditioner instead of an autotransformer.
This is the worst decision. «The severity of the Switching Transients in Transformer is related to the instant when the voltage wave is switched on; the worst conditions being when the applied voltage has zero value at the instant of switching», link: www.eeeguide.com/switching-transients-in-transformer/
What you did with your zero-crossing schematics is just adding triac finite resistance in serial with transformer. If you want to avoid current inrush in inductance, you must: 1) turn it on at peak voltage; 2) turn it off at zero current to avoid residual core saturation.
The worst scenario nonetheless not only switching transformer at zero voltage moment but with the flux reversal of the beforehand saturated core which can result in 10-100 numbers of inrush current. That possibly describes your time-to-time circuit breaker tripping.
The oldschool way to solve this for toroidal transformers in particular, is to just put an NTC resistor in series.
GREAT video!!! I think for a normal house.. all circuit breaker are of for normal devices..!!! you had this problem but... not every person has a autotransformer at home!!! ajaja
Why such a complicated circuit? You could've done something simple, like adding a resistor in series with the transformer and adding a relay which shorts the resistor when it closes, if you connect the coil parallel to the transformer (so it only closes once the current draw goes below a certain level and the voltage rises). There you go, same thing achieved with two components... Of course you can also do it with one component by just using an NTC. But with a higher power loss and a cool down time.
Did you ever hear about C type circuit breaker? Guess what - they are designed for this type of loads
Dude.....I am an electrician.
@@greatscottlab So why not to use so simple and efficient solution?
I am not allowed to change the breaker in my apartment and such AC softstarters have other advantages like with a motor starter.
7:11 that explanation makes no sense to me. The transformer should be a an impedance which is an inductor (Lm) in series with another leakage inductance (Lr) finally in series with a resistive component being the resistance of the wires. The current waveform at steady state is of course going to lag the voltage, the amplitude of which (apparent current) easily found with phasor analysis.
But with a step voltage in to an inductor should result in i=0 initially and build up from there, an inductor current is proportional to volt-seconds, the integration of volts with respect to time. So why is there inrush to an unloaded transformer? Is the inter-winding capacitance playing a significant role or something? Motors are quite complex loads so I can understand inrush for them
I had the same question/concern. The inductance will resist the change in current as long as the core isn't saturated. Interwinding capacitance would allow some initial current spike. ..or maybe there is an input filter cap at the autotransformer? Or maybe it depends on the load attached to the autotransformer? This might be an idea for a follow-up video.
@@SkyhawkSteve My guess is there is a sizable parallel resonant capacitor meant to improve the power factor by resonating with the magnetizing inductance. Such a capacitor would certainly draw a massive inrush when power is applied at the peak of mains, limited by series inductance and resistance of the house wiring and impedance of the grid.
A transformer is modelled as an inductor Lm *parallel* to a series combination of Lr and coil resistance r.
The point you raised regarding current being proportional to volt second is true only as long as the inductor is in its linear range of operation(not saturated). When a transformer is switched on at the zero volt point, the flux set up theoretically rises upto 2 times the peak value of normal operating sinusoidal flux. Most transformers operate near to its saturation to optimise iron usage and hence when flux can go upto 2 times the peak, the core undergoes heavy saturation and depending on the saturation curve of the core, the transformer will draw a very large current to generate that flux.
@@manjunathvishnumoorthy7479 Oh, I see. Because of the "integral windup" from a previous cycle the definite integral from t=0 to the first zero crossing would indeed see a current that is in fact double, that makes sense! Edit: Oh and I get your point about why you mention saturation, that's something I'll have to keep in mind thanks for sharing!
--On the transformer model I was assuming open-circuit secondary and thus the 3 series impedances (really an open circuit transformer is just an inductor or ballast) there is also some distortion and core saturation + hysteresis losses causing a what I recall was 3rd and 5th order harmonics and higher than expected peak currents, sure.--
Not real sure of how to fix it, but from what I understand is that the core material has memory. When turned off all the lines get lined up in one direction, and if you turn the power back on, and the phase is 90 deg. from the lines of force, it acts like a short circuit. It is kind of hard to explain, so I hope it is a little bit clearer.
Michael from Washington State, near Seattle.
All these JLCPCB shills basically selling out their own local PCB manufacturers. Shame.
I put an NTC in my Variac to prevent the circuit breaker issue. But this is cool for other reasons!
Nice video! One recommendation, make sure your diy pcb can handle up to 3000 watt or preferred 3600 watt.
The small screw connectors are not designed for this loads.
I like u'r video.. i'm from INDONESIA.. give me subtitle indonesia pls.....
You also could change your old B16 Brecker to C16 Breaker
Or even c20 or C25. I want to make one wall socket for C25 in my garage so my angle grinder and welder wouldn't Trip the breaker.
Tripping circuit breakers?
Ah I just remembered when I was 10yo I plugged in a copper wire in socket assuming it was nichrome wire to get heated
Suddenly I heard a trip and tv was off
Conclusion:-great scott teaches very good
Similar story about 10 year old me: I figured that if I shorted out an AC/AC adapter, it would short out the mains and blow the fuse. When I finally felt brave enough to try out my new found super powers, it just blew the fuse in the adapter without even making a noise or smoke.
I'm curious why you didn't just use a thermistor? I'm assuming it's because they wast energy as heat, but Im curious if that's the reason or not
Will this soft starter work for a solar setup? Starting a motor running off an inverter and solar panels, the startup current is too high and trips the inverter.
You should have used a mechanical relay to bypass the TRIAC after softstart. The would have allowed you to use a smaller heatsink. And at 3000W, the amount of current flowing through the TRIAC might heat it up a lot, so having the circuit inside a closed housing with no airflow might cause it to overheat.
True, adding a relay would have been a nice addition.
Hi great scott, another amazing video !!
Just a little question, at the middle of your paper schematic i have noticed u wrote 325V ... that's confusing me ... if we have 230-240 V AC isnt the max value +- 230 V ? why 325 ? could be a refuse thinking on a 3 phase system, where u have such voltrage between phases ?
Could this be used for simple ac loads like a non computerized refrigerator? My fridge has a terrible surge on it. 2200 watts, makes my inverter flip out and the lights dim.
DiodeGoneWild had simpler solution: use an additional inductor in series. When it saturates, just short the inductor.
I am curious if you tried a startup capacitor just like AC motors use ?
Always add extra bypass relay for post soft start bypassing to prevent lose while system is running at full Walt.
Soldering wires on to screws? What the hell is wrong with you? Were you dropped on your head as a child?
So its an AC voltage divider basically.
I think that using a relay when done with a softstart would be a great idea, because of the loss on the triac.
Can you make a video on register level interfacing for tft displays?
Thanks
Great project but really over complicated...one resistor in series and one timer relay close resistor
Why not use a resistor in series with the load that gets bypassed by a relay after f.e. a second?
Also possible. I actually did a video about such a softstarter a while ago.
Hi @GreatScott, I have two questions please. 1) Why is the optocoupler's anode connected via two parallel resistors (R6 and R2) both 220k. Would just one 110k not work? 2) Is the black filament the galaxy black from Devil Design :-) ?
I need one for my fridge. Every once in a while when the compressor turns on the amp draw pulls over 40a but only for a second but long enough to trip the breaker, there after it only pulls 2a. The fridge is only 2 years old.
"Just 500W" i just found that super funny for some reason haha
good
I like your videos, and you admit to making a big mistake here. I think you should re-edit and re-upload it with a breakdown. Really not complaining, or putting you down, we all screw up, but your channels is very educational and this will confuse people. I've had similar problems and used a resistor to limit inrush and a contractor to bypass resistor and connect output at the same time when the transformers output voltage comes up.
think I'll stick to the lazy thermistor and bypass relay
I found an even bigger problem than that zero crossing point thingie......
Why have you mounted your socket upside down ⬇️........
The older types of circuit breakers had a time delay characteristic that allowed for a heavy load to start up. These new "politically correct" breakers apparently break in one cycle. This can be one hell of a nuisance requiring all kinds of "hacks" (like shown in the video) just to live. I'll stick with my 50 year old type breakers that can start a big air conditioner without tripping, yet a few seconds more reliably shut the current off.
Is UPS a Sine wave inverter..? Plz explain about UPS...