Baluns - I got this one wrong (with schematic)
HTML-код
- Опубликовано: 7 сен 2024
- I got this completely wrong. I expected it to be for galvanic separation and impedance matching between the 75 ohm CCTV output and the 100 ohm CAT5/6 cable.
In reality it's just for impedance matching.
In my defense, it is called a balun (balanced to unbalanced) and I'd describe it more as an impedance matcher.
If you want to know what that's about then here are some Wikipedia links with enough formulas to satisfy the most geeky data-nerd. For the rest of us it will make some pretty tough reading.
en.wikipedia.o...
en.wikipedia.o...
en.wikipedia.o...
As an amateur radio operator, it’s comforting to hear even THE big clive has trouble with impedance and RF engineering.
Clive is scared of anything over 60Hz. ;-)
My understanding of RF engineering has been at the "It's black f&^king magic" level for anything more complicated than making sure there's no ground hum. 😁
I am also an amateur radio operator, if you put you mind to the job you will understand the operation of this subject. Just like your TV antenna to your TV set the balun. No black Magic At All.😀@@bonivuselderheart2716
@@bonivuselderheart2716nah, even that's black magic...
I’m pretty sure people who specialize in that field have trouble with it
The two basic types of balun are voltage - providing the galvanic isolation you were talking about - and current which is the type used here.
One of the problems with voltage baluns is that they are frequency sensitive so that not only is it difficult to build an off-the-shelf product but certain frequencies in signals such as video could be blocked or enhanced.
In current type baluns, at the transmission end the parallel windings cause the signal in the co-ax centre wire to be mirrored on the other winding giving a differential (balanced) signal on the output wires. At the receiving end the opposite happens and the signal on the earthed (shield) winding reinforces that in the centre wire.
As you stated, any interference induced in the link cable will be equal on both wires so the windings act as a common mode choke.
All that is needed is for the windings matched to the impedance of the co-ax cable so it can be built without reference to the frequencies being used (within operating limits of the cable).
The audio signal is already a balanced signal so doesn't need to be converted. Common mode chokes might be useful at the receiving end but that should already be a function of the audio circuit, if necessary.
Yes professional audio is balance but for example car audio is not. A balun can supress the majority of groundloop noises when there is such problem.
I used to install RF baluns on aerial and satv installs as part of a past life. Nobody clarified the science and it was always a matter of trial and error to find the best option. Channel 5 baluns were a thing in some places from memory. I may still have some amidst my collection of LNBs and associated gubbins. I first encountered them in my mother's 2mtr band equipment but only ever by name and the vague description of them being filters.
So how much are we talking about with interference? I mean, like milliwatts? I work with 1500 KW power supplies.
Great explanation!
@@wanjockey That will depend on what is close to the cable and how much shielding it has. If it is running past one of your 1500KW power supplies I would think it could be quite significant. Video signals typically run at about 1V so even an induced 10mV ripple could cause problems.
I was engineer for radio and TV stations in two adjacent buildings. Swapped audio between the two, including the TV SAP feeding audio to four AM transmitters (a cheap STL). Unfortunately, the buildings were fed AC power from seperate transformers. Nearly 30 volts difference in the ground potential between buildings. Had dozens of audio transformers isolating feeds between the two. Constant battle keeping hum out of the audio.
I spent an entire year in college learning all aspects of impedance matching in two different classes. It melted my brain a few times.
And bad impedance matching can melt transmission lines - quite literally. "Standing waves" on coax used for high-power RF transmission can produce local hot spots that will melt insulation.
My dad was an RF engineer working on early Radar systems in WW2. He told me that they tested power output with high Wattage lamps and to look for hot spots on 50 Ohm bus bars?
Passing signals through isolating transformers at each end to do the unbalanced-balanced-unbalanced conversion is the more familiar setup, called a "voltage balun". The setup here is an example of a "current balun". They really are doing unbalanced to balanced conversion and vice versa. The thing with video signals is that they have very low frequency components ideally going down to DC. Using the voltage balun setup, the transformers would be unfeasibly large if the lower frequency limit is to go down to a few hertz. The current balun will always pass DC.
That's very clear, thank you. Does transformer hysteresis ever become an issue?
If you make a transformer that big wouldn’t it cut off the upper end too?
Normally a balun would be designed so that it operates within the linear region, keeping away from saturation, hence hysteresis is not a problem.
It is perfectly possible to design large high frequency transformers, though special winding techniques and core materials are used. They can be found in high power transmitters for instance.
Data transmission over "RJ45" patch cables often uses isolation transformers built into network cards, with any DC power being 48V fed through center taps to use an entire pair as plus or minus, while still carrying balanced signals at 100MHz or more on the same pairs .
@@johndododoe1411 That is true, but for video transmission you need the signal bandwidth to go to DC. In the POE arrangement, the signal bandwidth won't extend below a few hundred Hertz even though DC power is being supplied over the same conductors.
Great to see the vice of knowledge make an appearance. I discovered that the wood chisel of interrogation makes quick work of glued and ultrasonically welded plastic cases. They spring right apart with a moderate tap from the hammer of persuasion.
One small example of data being sent over a power line is with model trains. In my day, you had a transformer and every train on the railway went at the same speed which you set. Nowadays you can get model engines with microprocessors in them, and the controller superimposes digital data on the power to the rails, so individual engines can detect commands sent to them and change their behaviour without affecting the others. In Hornby it's called DCC.
I have photocopies of Model Railroader magazine articles from the late '70s which gave the DIY version that evolved into the commercial product.
Same with slot cars
Earlier digital control systems struggled as manufacturers invented their own proprietary protocols, which weren't compatible with anyone else's (so if they stopped making it you were stuck chasing around for old stock or used parts to expand your layout). DCC has thrived as, after drawing it up, Lenz published all the details and encouraged everyone else to join in making hardware which followed their standards. So now you can buy a decoder from one of many sources and operate it with a controller of your choice. There have been changes and expansions to the protocols (particularly in relation to programming the decoders), but you can run a twenty year old decoder on a new controller and vice versa.
@@mattg437Märklin still has their own protocol, but since their stuff isn't compatible with everyone else's anyway, that doesn't really matter.
Sounds very similar to power line LAN networking :)
(and I think is a lot more similar to that than PoE, because it uses the same conductor.)
I am a (retired) E.E. I really enjoy your videos, you cover topics that we never learned in engineering school at a very useful level. I worked in television design for many years. We used baluns for going from 300 ohm balanced to 75 ohm unbalanced inputs. The were amazing 50-1000 MHz with less than .5 dB of loss and cost about $.10 each. Video signals are not so high in frequency from 0 to 5 MHz (in NTSC/PAL) and perhaps 25 MHz in more modern equipment. Notice I include d.c. the d.c. level is important for establishing black level. Thanks for all your content.
One upon a time every cable TV company truck had a bunch of 300 ohm to 75 ohm matching transformers on them. These days, I have no idea what a cable TV installer would do if confronted with a 1970s Zenith floor console TV with 300 ohm VHF input.
@@brianleeper5737 I bought a balun a few years ago to use with an FM folded dipole to 75 ohm receiver. It was a few dollars from Amazon.
Yes you are right i complettly forgot, thats why a "normal" impedance conversion doesn't work. Thanks
@@brianleeper5737 The installer my family had in 2015 had a hard time understanding that our older TV was a non-remote model. If they needed to modify the presets on the varactor tuner I'm sure they'd be totally lost. Today I don't think the latest cable boxes even have analog RF out.
We don't use much analogue video in the CCTV world these days but when we did the balun units we used were often PVD - Power, Video, Data. That was one pair for positive 12 or 24 volts DC and one pair for 0 volts DC. Then one pair for thebvideo as you described. Finally, the last pair in the four-pair cat 5 cable would be for camera telemetry data for pan-tilt-zoom (PTZ) cameras. This is RS485 data and works remarkably well down unshielded twisted pair cable over really long distances. Individually screened and earthed pairs are best giving about 1km (or 1 click as the military guys would say). However, 485 dataover unshielded cat5 is will go way further than the accompanying video signal and even over 1308 telephone cable, the data is good up to the point where you'd consider running fibre instead.
These days it's all IP video and PoE (Power over Ethernet) cameras with 48V DC straight from the network switch. Alternatively, we can run just data over ethernet and maybe also fibre optic, then either use a PoE injector powered locally from near the the camera or some cameras can run off 12 or 24 volts.
Good to see the Vise of Knowledge. Seems like we don't get to see it used that often these days.
Right up there with the conflagration containment tin.
Not to mention the hammer of truth and the chisel of revelation. 😀
Vice is spelled with a "c"!
I was waiting for the Hack Saw of Unyielding Truth.
Eff', I love the people around here!
Former IT cabler, electronics engineer, and radar guy for the Navy. If I was to take a stab at it I would say that the transformers take the signal(a/c) and ground (relative dc) and turn them into 2 singnal wires with an a/c signal(each wire having a equal and opposite balanced phases). The 2 signal wires then go across a twisted pair and join on the other side and the transformer brings the ground and signal back. I believe having an equally balanced (equal and opposite phases) signal wires will help reject any noise that would be absorbed by the signal or ground wires, causing interference.
Radio hams use baluns (and ununs) all the time at HF, but only about 10% of us can explain the operation and construction!
I used to be able to explain it back when I got my Extra but I doubt I could now.
Smart hams also use an optically coupled USB gizmo when connecting radios to computers
The construction is the easy part....😂
I think that the point of these is to send the video signals down a balanced feeder (one of the twisted pairs), which is much lower loss than coaxial cable, and will permit a longer run with lower loss. Also very much cheaper than coax!
They are 1:1 current, (or sometimes called choke) BALUNs, causing no impedance change at all, but they convert the unbalanced signal to balanced at one end and permit the connection of an unbalanced load at the other.
Nothing to do with filtering at all.
Thanks for the videos. I love your sense of humour!
Ellis
I've got several variations of these baluns, in a box of goodies, from the Vintage Computer Festival Midwest, in the States. Some use a coaxial interface, while others use a twisted pair. Very interesting little gadgets.
In the Telecom Industry, they deal with induction noise on copper lines all the time, especially on long runs. The are various filters they use, to suppress inductive noises. The biggest offenders are the power utilities, which usually run parallel to the telephones lines, & induce all sorts of fun sounds on the line. Farmer's electric fences are another source of problems.
Yes especially rural lines with the need to have those 88mH loading coils every few miles, though the worst has traditionally been railways, as the railways had, until they went to using phone company cables, used open pair wiring next to the overhead lines, and, because of archaic rail regulations, they never used those crossover insulator units to make a twisted pair every dozen poles, as they "needed to preserve polarity on the signalling" running on the same set of bare lines. Just meant the poor rail operators had to deal with massive noise and hum induced on the voice circuits, while the signalling and trackside equipment, running on AC, were unaffected. when they replaced all that with a single multipair bundle voice improved considerably, plus they never used loading coils at all, even on 100 mile long lines. funny the same cable still runs trackside, and the signals still work, as each gets it's own pair in the bundle.
Interesting thing here on a Ballin or matching Transformer.
Actually a few.
First of all I remember back in the day when I lived in Canada cable TV.
Which was actually closer to community antenna television if anyone remembers this stuff you don't hear that much anymore.
To put it in a nutshell essentially before the Advent of digital cable if people will remember that generally the First channels of the cable were actually on the VHF band so you can get channels to through whatever it was almost forgetting nowadays since we don't really have VHF anymore pretty much analog broadcasting what the way of the dodo I know there might be a few outline things or other but mostly pretty much dead as a dodo.
I remember the old matching Transformers we're about twice the diameter of the newer ones and ran across that years later and wound up using it still years back.
And when it finally did get messed up and was curious opened her up and pretty much same construction is the newer ones but just a bit larger chassis primarily I'm not sure if there's a difference in the size of the matching Transformer itself internally the toroid Transformer that is.
Been a long time.
Also I know that it is also possible to do something similar with a group of capacitors I have seen this at least once as well in practice.
Also with the ground Loop isolation Transformers or isolating Transformers absolutely this does help with audio if you have a device that introduces noise into the audio have had an issue with this with an Alexa device once.
Also I'm going to be designing a custom quote unquote Black Box.
To be able to run 2 stereo one level send receive as well as auxiliary some sort of connection likely 3 wire including isolation on the serial data and also audio.
Apparently ethernet Transformers work well for audio isolation from what I understand.
Also these boxes will not be passive since need power for Optical oscillation of the serial data.
Using infrastructure or represent to link 2rooms for audio.
Who knows what nastiest could happen on that cable I know for a fact that exits the building but we'll have this connected by service provider with a coupler if indeed it does go outside not even sure it is cat 5 cat 6
You know currently terminated as a Telco connection.
Have confirmed through service providers this is the case.
Using some infrastructure to make it easier to do whole home audio unfortunately not able to run cabling For custom install.
Logistics ugg.
Make the best of what you got available.
Also in addition to isolation full surge protection including audio lines and the data connection.
Also of course the Firefox would be a device to provide a breakout connection for everything and possibly provide some other functions as well potentially essentially those audio alliance with service and receive to be able to link both parts of the system having all the expensive stuff at one end in the main cabinet with all the audio gear since all equipments in the same cabinet makes things easier!
Having lived on a farm and also having relatives that lives on farms this is a known issue with the electric fence Energizer.
Anytime you would have a 120 volt and sometimes the other radios you would get that tick or click at a regular interval a lot of people will be confused by this and didn't realize that it was the electric fence controller especially if they didn't usually be on a farm or even realize this not that one very often confuse me at first when I was young.
Although my mother did teach me how to test if a fence was electrified or if an electric fence was active you basically grabbed a long strand of grass or other piece of some foliage and it would kind of Wilt if it was a active electric fence.
Still remember vacation at night sing in Ark jump across to overgrowth along the fence lines sometimes.
And yes there are cases where lightning and start the fence and had injured someone nearby scary thoughts.
What time was showering son thunderstorm I felt something through the floor best of Whitney Houston truck nearby and sure enough it had.
Also remember something somewhere happening at camp what's myself but campers for showering and the same thing let's check out the intercom system too did manage to do some repairs for them at least amplifier was not completely toast.
I don't know how they say they survived the lightning strikes but they do sometimes
The main method by which telcos, at least in the USA, deal with induction noise on copper lines is by using shielded cable and grounding it well, especially to the power company's multi-grounded neutral. Federal RUS (Rural Utilities Service) specifications call for telephone cable to be grounded at least every 1/4 mile. In some places where the telephone lines and power lines run on opposite sides of the road, you'll see a wire run across the road to ground the telephone cable shield to the multigrounded neutral on the other side. The effectiveness of this approach can be measured by using a clamp-on ammeter to check how much AC current is flowing in the cable shield; if there's little or none the cable isn't grounded well enough. Unfortunately, cable locators often disconnect cable shield grounds and sometimes don't bother to reconnect them. I've seen that quite a bit around here.
@@SeanBZA There's a video on RUclips from a retired Union Pacific employee showing the equipment they use which sends all the signal controls right down the rails (an AC track circuit), and I thought that's how most railroads do it now. Think the equipment was Genrakode.
Baluns are a good choice for simple, straightforward extension applications however they are limited to SHORTER DISTANCES VS. "EXTENDERS". Extenders are powered ACTIVE devices (not just a transformer) that convert the signal to another standard for longer transmission (like converting HDMI to HDBaseT, and then back to HDMI on the other end).
I run an auditorium where we have five extenders running between the stage and projection booth, each on a dedicated Cat6 cable...two HDMI going up to feed the projector, TWO VGA going down to feed computer monitors on stage, and a USB extender so we can control the computer in the projection booth via mouse and keyboard from the stage as well as transfer files via thumb drive (so we don't have to run upstairs to the projection booth to directly connect the thumb drive to the computer). We do have a sixth Cat6 cable for running 4xXLR, but that is a "cable snake" since it's not active and doesn't do conversion (although there may be a transformer inside to facilitate the different cable type, never opened it).
Extenders AND the balun you showed do NOT use "Ethernet". Yes you use the same Cat5, 5e, 6, etc. cables, connectors, and connector wiring pattern that Ethernet uses, BUT extenders and baluns require DEDICATED Cat cables...they can NOT carry the extended signal along with Ethernet signals...so it's best to call them Cat5e (or whatever cable you are using) to differentiate them from Ethernet cables which carry computer data. There are a few extenders that do work over Ethernet, but it's best to avoid them as you want a clean dedicated signal with as little latency as possible (unless your use requires converting to Ethernet). Also you CAN'T use Ethernet hubs or switches with extenders/baluns.
When we first tried this WE DID TRY an HDMI balun setup, but the distance of 75 feet of Cat6 cable we had installed was too far for it to work. The active extenders haven't failed us.
p.s. When we connect audio from laptops or MP3 players we use what called a 2-channel "DI BOX" (DI=Direct) which converts stereo unbalanced audio to 2x XLR. Yes, it basically is a dual "balun", but calling it by it's name of DI-Box differentiates it from other baluns and it's the industry standard term for this specific device. While we use a passive DI-box since what we connect to it is already amplified (like a laptop headphone output), there are also active ones with amplifiers that allow you to connect electric guitars.
The need for an active extender vs passive balun really depends on the characteristics of the signal, the cable, and the receiver equipment. HDMI is high bandwidth and receivers are not built to receive degraded signals, hence the need to an active extender. For analog composite video you only need 4-5MHz so you can use much longer cables without significant signal loss. Audio can go even further, in the old-school analog telephone network _everything_ was only passively matched (except for certain trunks between offices).
You can get video 'humbuckers' that do provide galvanic isolation. Useful in larger systems. You can also get baluns that put multiple analog video signals over CAT5, one per twisted pair - in my experience the passive ones of these are fairly terrible and unless all your sources are genlocked the crosstalk between pairs is too high and you see artefacts of the control signals (like the H sync and colour burst) of one video line appearing in another.
The single channel ones are great, and work really well for remote cameras as they remove the need to drag power to the far end - I use them a lot for CCTV cameras on stage sets, and also in the other direction for monitors in orchestra pits - removing the need for power and a PSU at each music stand.
There's also equipment available that goes the other way around allowing Power over Ethernet and Ethernet traffic to go over a 75ohm coaxial cable to for example cameras, which is useful when you plan to replace analog cameras with modern IP cameras.
The impedance of a twisted pair cable is about 100 ohms and a coaxial cable is usually around 75 ohms, but that's a small difference and the transformers in the end units are basically there to ensure that the signal through the twisted pair is balanced to avoid causing interference.
power over coax is easy, you just need a capacitor and inductor at each end. the cap to remove the DC from the signal at the ends, and an inductor to allow the DC into and out of the coax. remember the capacitor blocks DC and the inductor blocks high frequencies.
most common I see do POE over twisted pair cable, 2 pairs for power, and 1 pair video, and the last one audio/PTZ control if camera has it. They will do 10M data, as they isolate the power pairs from input and output, and then you connect the camera to the converter. 48V on the cable, and the output side has a switchable buck converter, providing 12V or 5V, at up to 1A for the camera. The 5V is generally for things like non POE capable wireless AP's, as many will run off 5V, and old ones max out around 10M anyway, and another common use was to power a DSL modem, as often phone lines were not convenient, and DSL here never really got past 10M in any case, even if you literally were 30m from the DSLAM cable wise, and it was outside your door. As a bonus now DSL is kicking it's last legs, Starlink has more (illegal) customers than DSL has, as the state owned monopoly is killing off copper service, and no longer repairing it either, unless the fault can be resolved with a single service call and stuff on the van only. Stolen cable you get a GSM phone that looks like a regular handset, too bad about data unless you also got it from them, then you get a dongle as well, and your subscription price and data available change for the worse.
Many years back, there were DirecTV "DECA" adapters going for something like $30 for a bulk pack of 20. Quite good for projects that involve running IP over coax.
The maximum distance from the datacomm distribution frame to a wall outlet is up to 90 meters, so a video camera could have a 90m cable from the camera to the frame, and another cable from the frame to the security system. That's up to 200 meters with patch cords, etc. So the signal quality depends greatly on the lengths of the cables. Some video cameras may lose a lot of quality on long runs.
You have to have opposing windings on the video signal because one of the gnomes in the camera has to sit on his head in order to fit, so they need to flip it upside down so he can read it.
I think the video ins/outs are separated like that to cause any induced noise to be canceled at the outs. It ensures the signal out perfectly matches the signal in.
For isolation, have you seen the long HDMI cables that work by having an optical convertor at each end, and optical fibre as the body of the cable? A neat solution to avoid signal loss and interference, as well as providing electrical isolation.
The long cables it is more due to the very limited drive ability of the HDMI chipset. Driving a long cable, even a nicely made twisted pair one, will require a lot of current on the edges, to charge and discharge the capacitance of the cable, both the common capacitance to the shield, and the capacitance of the twisted pairs. So a short cable, then a HDMI input chip, will mean only a low drive needed, well within the limited capacity of the sending side. Then you can convert to fibre, and have the length be moot, only the amount of degradation such that the receiving end is able to reconstruct the data well enough to send out the far side HDMI connection, for the receiver error correction to reconstruct well enough. Yes often they provide electrical isolation, as a side effect, but some of them also are only powered from the one end, so have a pair of cores with the fibre optical link to provide power to the far side, so actually do not have isolation.
Just like networking, longer cables result in signal degradation, as the long length attenuates the signal, though that length limit is not absolutely fixed, toy can be 10% either side and still have perfect operation, with no lost packets, or have used cheap cable, and a few couplers in the line, and have a 30m run not be capable of giving reliable data, unless you force the link to 10M. I have run 10M connections over an existing phone cable, because I could not pull a new cable, limited time, and 10M was well within the data rate, plus had both a pair of those BNC to twisted pair baluns, and a old pair of 10M hubs (not switches, still got one 10M hub around with a RJ45 port set and a BNC port, useful to use as a tap in for network data analysis, though it limits speed down badly) for the ends. Worked for the few months needed, before new cable went in.
Any digital signal that can plug into a media converter can be transported on fibre. With the right support, you can plug a media converter onto your PC and access fibre directly, this was how we would test large arrays of fibre outlets.
@@SeanBZA If you have to run LAN over old phone lines that cannot be replaced (e.g. if you're renting), hack a cheap pair of Homeplug adapters and get at least 100Mbps, a better pair I modded for that purpose would go up to 600Mbps.
@@NiHaoMike64 Yes but was limited to what I had around. Cost no object, just do it as fast as possible, and at zero cost. Plus had the hubs, as recently had done the upgrade from coax to cat3 cable, and replaced all the old long coax with cat3, along with putting in some actual 10/100 switches, replacing the old hubs that connected all the PC's per floor onto the coax. New network cards, as new computers only had PCI riser, so the old ISA cards, with coax connector, no longer usable, and cheap PCI cards no longer supported coax, though the PCB had the spaces there for the components still, just non populated, and no cut out in the slot cover. Plus the upgrade from Novell Netware 3.0 running on DOS to Win 95. Never had a computer there before either, but plenty of phone lines, which I was able to grab 2 pairs out of the trunk cable, that looped through, to use.
Naughty, as those wires actually were phone company property, and was not supposed to touch them at all. But at least it was cables from the 1980's, when the building had been upgraded phone wise, and not the old 1960's era wire, which had as plan one phone per area, and that was a 3 wire system, with a common cable. Who would need more than one phone in an office of 20 people?
A lot of AV companies got these in the noughties. The fibre didn't last long for many of them.
I honestly enjoy reading about Transmission line theory and impedance matching.
You just need to get your head around that the signals don't travel infinity fast.
My background is data a telecoms, we used baluns that would take an ethernet signal and convert it to a 2meg carrier on mini coax, these were actually pretty commonplace, but in recent times have been replaced because the PSTN is being replaced by IP telephony.
I've seen all manner of baluns, but i've never seen one like you opened.
Stay well Sir.
Or the curse of any line tech, the 0+16 line expander, which would cook the cables when it got wet, and gave you a nice zap with the 300VDC it used. Or ISDN, which did the same, plus plugging in a butt would make it smoke nicely. Always made sure any interconnects with those pairs, or ADSL, were a different colour in the junction boxes, so as to warn of the peril. not fun for the line tech to open a street junction, pull the cover off the cable joint, and see that there are a few black wires in there, leading down the paper insulation into the cable, saying time to cut back a few metres to find good paper. Or the 1970's era 2M punch units, where you punch one wire down, and 3 pop loose randomly elsewhere in the frame.
Basically, imagine what Dolby did for noise reduction on audio compact cassettes and apply that same principle to noise reduction in an unbalanced cable: It encodes and decodes passively using the two transformers as the 'check-sum', which add together the waveforms in the positive and negative sides, to reconstruct an approximation on what the original clean signal was prior to being sent over the CAT5(e)/6 cable, allowing for long runs of ethernet cable to be used for further distances instead of the more costly Coaxial cable over relatively shorter distances. The circuit is probably so simple, so as to allow quite a large scale of frequencies & wavelengths to be used, possibly even allowing coax audio signals to also travel longer distances with little to no signal degradation. You've got me intrigued to experiment with a few of these units in order to discover more of their uses 🙂
CONGRATULATIONS ON 1 MILLION SUBS CLIVE!!! 🎉
yay hope I'm not too late
This is a CCTV video and power balun.
It's for using Cat5/6 instead of COAX.
CCTV systems are designed to use 50/75ohm COAX, these baluns don't just covert cable types nicely, they also effectively make the camera and the recordings equipment think its using standard 50 or 75 ohm cables.
Within the baluns are usually a couple of discreet components these effectively cancel out the reflection that gets introduced by using twisted pair cables.
You can also get 4 video feeds down 1 cat5/6 as each pair can be used as a separate cable.
Video signals though must only flow one way. You cannot oppose them otherwise harmonics will appear on the video images.
The use of a balun for interface removal has a lot to do with the phase of the interference and by using a balun to change the phase of the signal in relationship with each other nulls the interference (usually). It doesn't always work, it can be affected by the number of turns on the balun, the length of cable and the frequencies involved, however it's more reliable more often than it isn't, and baluns are often used even when only using coax because of how they work they still offer benefit (depending upon the type of data or signal being sent).
In this day in age, if your CCTV is still using coax (analogue or digital) it's probably of an era where is significantly more insecure and of such poor quality it's would pay to get it replaced (unless it's used entirely offline, and in a well lit environment.
I tripped up with this myself.
In my case, the baluns I was trying to use were for datacomms. They were real 1:1 isolation transformers. I lost the DC and LF components in the video signals, so messed them up.
I really should have known better !
Oh, I am actually relieved to hear that the audio doesn't pass through a coil. That would make for a wonderful low pass filter and screw up the audio signal. A cheap USB DAC with optical out is absolutely the best solution to get rid of a ground loop and mains humming.
And where would you get the power and optical fiber for that monstrosity in a building already wired with Cat5 structured cabling?
@@johndododoe1411 Why do you assume that your particular situation applies to everyone? Also, you only need the galvanic separation just before the amplifier. The optical cable only needs to bridge the short distance from the DAC to the amplifier. Where there is an amplifier, there is a power outlet to power your DAC, be it mains voltage or a USB port. The "monstrosity" is a DAC as small as a pack of cigarettes, and 50 cm of optical cable!
When I saw balun I instantly thought “ooooo. Clive is also into his radio stuff” then chuckled at the “really geeky” comment. Yes that’s me.
Looking at what you have there it looks like an impedance matching transformer. It’s very similar to the 49:1 transformer that I use in my current HF antenna. It’s “49:1” is basically it turns an impedance in the realm of 2450 ohms and converts it down to the 50 ohms the radio likes to see. The antenna is what is known as an end fed half wave. So it’s just a straight wire which is half the wavelength of the minimum frequency I use. So mine is 20m long (for approx 7mhz).
As the current goes up so does the apparent impedance. The voltage therefore goes down. When you look at a sine wave and just take the bit above the line it’s basically the start of the wave. Voltage then goes up as current goes down as does impedance. So the cat cable in the middle acts in a similar way. To transmit the signal down the wire the transformer is used to “match” the impedance.
Transformer is there to match the 75 Ω impedance coax cable with the 120 Ω of twisted pair which is used to run for the video signal.
BALUN means: BALance ~ UNbalance.
This is the only way to use twisted pair (UTP) cables for video signal.
Not quite. It's not the galvanic separation that matters here. It's the DC blocking. A composite video signal relies on the voltage to represent the brightness (or sync pulse). If the transformers were connected "conventionally", the waveform would always be centered around zero, whereas for very dark video, the waveform should be significantly above zero, and it in no case should it go below zero (which would be necessary some of the time if you were to center it around zero). I love your videos Clive. Thank you for showing so many interesting things from the inside :)
Clive: “unless you’re really geeky and Into that”
Me: You called? 😊
Clive has just called me "geeky". Achievement unlocked.
1:15 That slight shock you get is because of a mandatory return path capacitor that connects the mains neutral to the output ground. But with non-polar mains connectors you can plug the charger in backwards and now there's a ~4.7nF capacitor from live to the laptop's ground. Which gives a slight shock when you're grounded and touch it.
It would be fun to see this demonstrated!
The RF filters are also a type of balun that looks to make the RF signals from the PSU look like they are 50%:50% (anti-phased) relative to the ground so no radiating. That needs a definite leakage current at the core 50Hz (or 60Hz US) frequency. If you have too many then the leakage detector circuit can trip!
@@philipoakley5498Just another reason why the old mains wiring regulations had special rules about computers with high ground leakage . Unfortunately this was lost in the push to get everything on 30mA or lower RCDs .
Excellent information Clive. I’ve used baluns on video systems for several years now. Always wondered about the circuitry in them. Very good tip to use baluns on audio systems for isolation from ground loops that create noise / buzz etc.
I love that i can thank my trade course and my old instructors as i did learn the theory as well as practical applications for set-up as well as repair/maintenance of the equipment. I do miss it at times tuning cavity filters and making or testing antennas etc, i work in fiber optics now
Just realized what was in the picture. This is going to be a good video and I haven't even watched it yet I can't wait.
Oh dear. You've brought back to memory all the 'joys' of Smith Charts, and the cruel theories of waveform propagation on twisted pairs (and waveguides). Truly a study of penance for anyone reaching for understanding of the higher sciences of radio and electronics. So_o_o long ago, and far away now. So glad I'm retired and don't have to think about such things. 😂
😂🤣 A dark art indeed..... RF Isolation is not for the faint at heart, and, more often than not, beginners tend to land in hot or melted soup when it comes to RF and Balun....... 😁I remember when going from CB and ham to RF equipment at airports..... You go from VHF to UUHF equipment in one room, along with tuning cavities the works...... And ALL need to be temperature compensated...... 😂🤣 A dark art indeed, I was a hot mess when I started my training...... I enjoyed it a lot but oh me oh my......... What a learning curve it was..... And don't forget about the temperature bit......🤭 Just a tip with the audio bit, when connecting to desks which have balanced inputs you can use a set of active DI's or now-a-days you even get dual DI's specifically for use with laptops which can give you a balanced output, thereby also isolating the laptop and any noise it might carry.... I love audio electronics, right up my ally, out here most of the guy's I get to train have been laying cables and setting up without even knowing WHY they use the cables they do, they say, this is what we were told to use, and they do, but, have no idea why they have to use it...... Sad really that people have become so secretive with their trade that they would choose to not teach the trade but let it die with them.....
7:15 I believe that circuit is being used as a common mode choke. The transformer is wound such that differential mode signals (i.e a proper balanced signal), with the same (but opposite polarity) current in each leg, cancel out; they produce no magnetic field through the transformer core and it essentially has no effect on them. Common mode signals, however, induce a net flux in the core; essentially, it looks like an inductive choke to common mode signals, but is completely transparent to differential mode ones.
This video helped me figure out a issue i was having with a ADS1115 16-Bit ADC.
I did not expect to see these. I have only watched a bit of the video but I really worked very closely with a LOT of Baluns due to CCTV installs. The classic Analog cameras use the BNC connectors and get the power through the Baluns. SO you cable up Cat5 or Cat6 and use Ethernet cable for the analog cameras. In Miami we installers always referred to them as 'Balloons'. Also referred to any Ethernet-cable extender (RCA, HDMI, ETC) as a balun.
-former installer.
6:53: Ppl pls correct me if mistaken - to my understanding, the windings on the video wire literally work as balancers/unbalancers. An unbalanced signal comes in, then the opposing windings in the transformer make the unbalanced component of the signal balance out across into the respective other wire, then the balanced signal runs down the cable, then the process is reversed by the matching (identical?) balun at the far end. So it's from UNbalanced to BALanced, then from BALanced to UNbalanced once again. If I understand correctly, the balanced signal is effectively like differential signalling. Corrections welcome.
I'm glad that you admit when it's beyond your expertise. I have with that, so you're a good role model. 😮
They are balun transformers. matching an unbalanced input to a balanced output suitable for transmission along a twisted pair (within the ethernet cable) and then back to unbalanced again, at the remote end. Much used in amateur radio / RF engineering.
Seems like a current balun, similar to what you were talking about with a voltage balun. You are basically right, just a different type of balun. The audio signal already looks differential so no need for it there.
So this is what my Panasonic toughbook cf-31 has noisy audio out .........., thanks Clive I really appreciate your teachings I'm in your debt.
"Reassuring squeeze with the vise of knowledge..." That one's going into the quote bank!
*_CLIVE:_* We really need to know exactly how the transformer is wound. In the image at 5:44 It looks to me like there is a something wound around the toroid outside (i.e. going around the outside circumference like a belt) between layers of normal turns, but I can't really tell for sure. That may be a tap in the windings wound in that way for convenience.
I'm also curious what sort of core is being used. My guess is that it is a high-permeability ferrite and will just be plain dark grey with no markings. Things can be learned with the right instruments but if they aren't at hand it can be time consuming.
A lot of people are suggesting it is just a common mode choke but I don't buy it. I think it is a proper so-called "current balun" which preserves the DC component of the composite video while matching the impedances.
Years back we had ground loop issues when running composite video coax cables from one building to another due to the building grounding system differences. I remember purchasing industrial video ground loop isolators. Each coax needed one and they were quite expensive at the time.
I use Baluns in my security cam setup but, for video and PTZ data only. Those you have are the cheap ones that I don't recommend unless you're using small, very low current cameras. I spent a little money on some beefier ones that can handle more current but, I still have issues running power with them. I have very large PTZ's that require 3-5A each. I get the best results when they use their own dedicated power ran to the closest outlet. There are more "professional" ones that I'm just not willing to dish out the money for.
Thanks clive. Im a security tech and install/replace/remove balans in by day to day work. I now know that I dont want to know what these baluns actually do to the signals.
The transformer is a common mode filter like mentioned as noise canceling.
I can't believe I had to read this far down into the comments before someone recognized those common-mode chokes.
Hi Clive, Love your channel! Been watching for several years now. The transformers appear to configured as nothing more than common mode chokes - a clever way of removing interference picked up in the cable between camera and monitor. The camera signal is differential (not balanced) interference will likely be induced into both conductors and hence common mode. I think no impedance matching is going on here.
"Yep."
"Of course."
"Well, how absolutely silly of them, indeed!"
- Me, who knows nothing about it but had a blast watching it. 😎🤘
I very rarely have any idea what you're talking about, though I am learning, but your channel is great!
Well, today I learned that "balun" wasn't just...a word! Only taken me 40+ years of life, and 20+ years of using the damned things!
Remember as a kid looking at some TV antenna magical thing, converting "the round cable" to "the flat cable". It had a binocular core with some wires going back and forth, and it didn't make sense to a kid used to batteries, tiny bulbs and motors. Now I design and make this magic myself 😸
Those are balanced to unbalanced antenna balluns. The antenna is a ballanced, 300 ohm dipole, the coax drop cable is unbalanced, 75 ohm coax, and the TV input is another 300 ohm, ballanced connection. On the other hand, other people just used balanced, 300 ohm twin lead, without baluns, for the drop. Eventually, they started making TVs with the unbalanced input, mainly because people were connecting them to cable systems instead of antennae.
A few years ago, I watched some of the lectures by Greg Durgin, eng prof at Georgia Tech, on Time Domain Transmission Lines. This is digital signaling which really helped me get my head around transmission lines. I had the Time Harmonic lectures back when I was in school but you get lost in all the waveform interactions. They didn't even teach Time Domain back when I was in school. And now that mode rules the world with digital comms.
I find it amusing that we ended up with our "modern" world being controlled by the Telegrapher's Equations dots and dashes on lines instead of waveforms.
I use laptops for audio often. Usually if i have a buzz or other hum, option 1 my audio jack is bad. Option 2, which happens more often, noise is being introduced into the laptop because the power supply is beginning to fail. Then it picks up noise either from its own switching, or off of ground pin. When I'm feeling cheap, i use one of those dirty and nasty adapters between wall receptacle and adapter plug. I do eventually end up replacing the power supply cause they have a habit of destruction upon disassembly.
Why they do it this way is simple, Video's frequency range basically starts at DC and extends up to many megahertz, you cannot pass video directly through a regular transformer winding.
The low frequency/DC bits just pass almost directly down the twisted pair, 1 volts DC (for example) will pass straight through, however as the frequency increases the balun's more and more convert it into a fully balanced signal on the twisted pair, it's at these frequencies that all the transmission line effects really come into play.
Actually transmission line effects covers all frequencies, it's just that at the lower frequencies the line length needed to see them becomes huge. Old school telephone lines of many many kilometres see this is audio, however no one will try and run video over this kind of distance. Most of these video baluns are rated for use over say 50 meters, so the transformer designers only have to worry about signal components over say 20KHz, saving lots on what goes into those coils.
To go any further you need to convert the video into AC (a modulator/demodulator), or go optical.
Interesting to watch your contents, Great service to technical society
Interesting thing about ethernet is that it's normal to use tiny transformers at each end of each wire pair, mostly to reject common mode noise. Usually they resemble an unusually bulky 16-pin chip but it's not that uncommon for them to be integrated inside the ethernet jack. Might be that they're achieving galvanic isolation that way, although I kind of doubt it. A transformer that small probably wouldn't play very nice with such a low-frequency signal.
Hi Clive as I was an alarm , CCTV Engineer ,the old Composite signals between cameras were always a problem as if the grounds were shared at the control unit with out filtering you got crossfeed from all the cameras giving ghosting ,so all BNC's had to be filtered by baluns.Take care buddy
I saw the word 'balun' and was thinking BC had got into amateur radio, then I saw the opening shot and started scratching my head real hard. ;)
It never ceases to amaze me what signals people pass over ethernet cable without actually, yanno, using ethernet. o_O
Yeah, well it's just a bit of cheap wire, innit?
You mean like HDMI, USB and such........
The baluns are ununs (or un-un) strictly for common mode interference suppression.
No impedance alteration. Ron W4BIN
Oh yes the video signal is getting impedance shifted going from 75Ω to 110Ω and then back again. The crazy thing is if you get really creative with cutting up coaxial cables (maths) you can actually use different coaxial cable pieces to shift between 50Ω and 75Ω.
And you can get your video and audio isolation transformers to get rid of ground loops. And that's a similar dark magic.
I'm not an expert but I have done a little bit of reading in the ham radio and things and finally got my license. And far much more reading in the Electrical audio video end of things. I did a relatively small experiment one year one day where I took 2x 10bass2 50Ω coaxial Nettwerk "hubs" and connected them over traditional television 75Ω Cable for about 50 feet, I had to use a variable resistor to match the impedance expected for the termination resistors but it worked and was stable surprisingly. Normally they are terminated with a 50Ω resistor on each end of the chain.
I think they are wired as Common Mode Chokes. Any noise picked up on the cat5 video signals will be common to both video conductors, and this noise cannot propagate through either of the chokes.
Ham over here, I think in this case the "baluns" are 1 : 1 and they are choking baluns, not really a balun but, if the ferrite is for the correct frecuency and if its correctly calculated and constructed can do the work.
The best option in my opinion would be the Ruthroff configuration that is it a real BalUn (Balanced to Unbalanced) system.
Balanced is the UTP Pair and nothing have to do with the fact that they are twisted and unbalanced are both the camera output and the monitor or DVR input .
They wont be nedeed if you run an unbalanced transmition line within them that have the same impedance (Coaxial cable, 75 ohm, could be RG6, RG59, etc).
They are NOT matching any impedance.
Aside from impedance matching using transformers with a centre tap can be used for "phantom power". Traditionally this was how microphones were powered with 48V DC or in some rare cases 110VDC. Even without the need fo DC power balanced circuits run over shielded twisted pairs provide excellent common mode noise rejection.
The standard way today of feeding HD video from cameras etc is using HD-SDI over coax or fibre.
I didn't realize they're so complicated. At school, we had to use a different thing called a balun to connect rj45 to the older token ring infrastructure.
Token Ring Baluns deal with the difference between the old square plug/DB9 signalling and the more updated Cat5/8P8C signalling if the token ring protocol, with one of the two being electrically unbalanced . I haven't used this for years .
The true use of those baluns is in the cctv industry, camera on one end and monitoring on the other end, the cat cable between each is normally the structured cabling with in a building ie the network cables, the use saves the need and time to run km of new coax cable
In this application there's no point galvanically isolating the video signal when it's power isn't isolated as you'd just end up re-grounding the isolated signal.
Old technology now for CCTV but these were such a better option compared to running coax and power. Especially when they later wanted to upgrade to IP. No recabling needed!
Also for interference being cancelled out through balanced transmission lines, a graph showing how the induced signal gets cancelled out is a really good way to visualise it.
The way those transformers are connected look they are configured as common mode chokes. Any noise induced on both conductors on the cable run will see a high impedance at each end.
I have a pair of these for USB. It was for a USB web cam we used to keep an eye on the output of a giant inkjet printer.
It does the video equivalent of an audio DI box, mainly allowing for longer cable runs by converting high impedance signals on a coaxial line into a low impedance signal over a twisted pair cable. There are other things you can do with them but this is the main reason such things exist. As a musician and electrician Ive used both for that purpose. Without taking one apart they should use an isolation transformer and some resistors to present a high impedance input for the incoming signal just like a guitar amp or pedal will use high ohm resistors at the input for the same purpose.
The coax impedance is 75 ohms. The twisted pair is 120 ohms.
if I can give a 2cent break down, video signals, such as NTSC or TVI (an extended type similar to NTSC on a scope) will not pass through a transformer (wired for isolation), its like sending a square wave through a transformer, it will not come out square. So the DC is fed through these inline so the DC NTSC signal passes through, but higher frequencies on the lines will cancel out as they will induce on the coils. - very simple over simplified 2 cent version
I think cat 5 twisted pair impedance ranges from 100 to about 130 ohms when you get into the HF range and above, so they are probably trying to match it to the 75 ohm impedance of an RG 59 coax.
The electrodynamics shows us that the twisted pair (or an old ladder pair of wires) doesnt have symmetry of concentric but has the same electrodynamic properties... in the ideal world - the Maxwell equations with "a little" simplification bring us to the same conclusion of a transmission - in the ideal world :) but the world is not ideal and the power adapters have faults :).
I think its just a radio balun - like these being added to the antennas and similar - it matches the impedance.
There are two ways that a signal can flow down a pair of wires. Either Differential or Single ended.
Telephone and Ethernet cables are Symmetrical and Balanced. The Differential signal flows in push-pull mode down both conductors.
But then there are Microphone and Video cables which are coaxial, and always have one side earthed. These are called Unbalanced.
Often it is necessary to convert from one type to the another, and that is what a Balun or Unun is designed to do.
The classic example is where an old style TV ribbon (eg balanced) needs to connect to a Coax type cable (eg unbalanced).
Incidentally the function of a Balun or Unun is not primarily for impedance matching. It may do that, but that is a secondary part of the design.
A related device is the "Common Mode Choke", or "Braid Breaker". These act to force the required mode, either Differential or Single Ended.
And there is also the "Isolation Transformer" which allows Differential signals to pass, while blocking Common mode (eg single ended) signals.
Each device has it's own unique function and is used to cure a specific problem.
I never did like a single-ended audio connection to a P.C. or a laptop.
A differential signal or an isolated connection with a Toslink and a DAC is what I prefer the most.
Thinking we're smart comes with refusing to admit we don't know something.
Being intelligent is the ability to recognize what we don't know.
Thank you for being intelligent.
HAM radio operators can open a whole new world with antennas different types of impedance matching inductances and slang like balun, unun, etc. These ones are just balanced to unbalanced matches.
Baluns are quite a common thing in RF power amplifiers.
I don't know if they can be found on the interwebs, but Motorola produced a bunch of ap nots on design of RF power amps and many of them empolyed baluns between stages.
@7:00 The two windings are wound opposite each other. Since the signal is relatively inverted in polarity, this reinforces each signal while cancelling out any common mode signal before it even reaches the twisted pair. It also significantly reduces any unwanted signals on only one of the conductors relative to the main signal.
I had this thought too - ie the rejection of common mode interference.
I've been doing a deep dive into RF baluns (for amateur radio) for more than a year now and still don't feel even one day closer to understanding them in all their complexity. They hurt my brain, but it's a good kind of hurt... I think.
I am still doing it - studying for my Full exam here and yeah it makes my head spin at times. "..Impedance matching which is a horrific science in its own right..." never a truer word spoken.
I could be wrong but it looks like the "transformers" are being used to remove common mode noise. Basically this can be a "cheat" to get round slight impedance mis-matches. EG using a 51R resistor to terminate a 50R line. I cant remember the specifics but network cables hace an impedance of roughly 100R with CAT 6 being tighter tolerance than CAT5 and CAT5e. I think CAT7 also has a tighter specification for DC resistance, and CAT8 works up to data rates of 40Bn per second. That said CAT8 is only specified for shortish runs for the higher data rates. So in short if you substituted a CAT8 cable you might not even need the common mode rejection inductors in some applications. That said those Baluns are not designed how I would do it.
I thought they were going to be "Bias Tee's" for supplying power over a single pair without upsetting the RF. Normally used on coax cables for powering switches or pre-amps etc. As others have said the transformers are being used as common mode chokes for noise suppression and possibly impedance matching.
I think you just helped me troubleshoot the issue with my stereo lol! I have a little 3.5mm Bluetooth battery powered device hooked into a diy line-in input on an old Sony boom box and it sounds pretty good but I've always had the feeling it could sound even better. After watching this I will be buying a "3.5mm isolation" device off ebay for a few bucks and see if that helps!
I had no idea these existed, very interesting I knew ethernet cables had the ability to use POE but never realised these existed. Come to think about it CAT 5/6 cables are kinda neat.
I have installed hundreds of those baluns in cctv environment in combination with UTP
super easy to connect and have never had any problems with them
with the use with outdoor cameras you could sometimes get some corrosion but with a well-sealed junction box that was still survivable 🙂
With the way the schematic is drawn, the windings are acting as a common mode choke. if the windings were isolated between the device and the cable, then it would be acting as impedance matching.
Another advantage of signal isolation transformers is protection against electrical faults and surges. If you buy dodgy eBay/Amazon amplifiers, I'd recommend using an isolator in case live shorts to the low-voltage side to block AC from getting into anything you may have plugged into it.
As for "galvanic isolation" causing problems with video signals, it isn't a problem for 10GbE passing 300+MHz per pair, it shouldn't be a problem for 6MHz analog video signals. NTSC/PAL video signals do have DC bias though, which would require active circuitry to compensate for.
I found putting a wide wood chisel along the seam and giving the chisel a tap or two with a hammer splits glued seams.
Opposed windings. Think Common Mode Chokes. At each end means a composite video signal can be sent (1V PtP) over the cable in either direction and the interference is blocked.
Frequency sensitive perhaps too. Distance is still an issue due to losses, but it is better to amplify a weak clean signal, than a dirty one.!
David G1ZQC Radio Amateur since 1986.
Ethernet cables makes the world go round
No wonder why the world is so "twisted" today. Things were just a little bit less "twisted" when we only had telephone lines.
I don't see any comment that would summarize why those baluns are even used and why they are constructed this way:
1) twisted pair cable is cheaper than coaxial (at least when comparing cat. 5e UTP and RG6), though not much. It is though common sense to install twisted pair as it is more future-proof
2) Because RG6's cable impedance is 75 Ω and twisted pair's 100 Ω an impedance matching is required, otherwise the signal will reflect (see Wikipedia what that really means...)
3) in this case there could be no galvanic separation, because video composite signal (PAL at least) contains a DC component, which would be lost when transformers were used.
Quad aluminum shield, copper clad steel center conductor RG6 is cheaper than cat5e last time I checked.
I realize there is a really low bar. But I ended up down a rabbit hole when I went to replace the excessively long and mangled RCA cable used by the original 2600 to connect to the TV. when I started researching I found that to get the best picture I learned that I needed the right impedance cable AND the cable had to be a certain length such that the source and tuner were at a total cable distance at a specific multiple of the signals average wavelength.
I gave up and am still using the original cable.
Love the pink cable Big Clive!
For some of the cameras we have on poles that are a long way from the switch, we actually use ethernet to coax converters that take 1gb with poe and run it over an RG6 coax cable to maximize distance.
We also came across game changer cable that gets you a stable connection over 200m with poe.
I've heard of those "game changer" cables but I don't think they're worth it. IIRC you are limited to 100 megabit Ethernet at best and it still violates installation best practices. It's much more flexible to run an optical cable and a separate low-voltage 48v power cable. It's probably about the same cost yet it will let you upgrade to 10gig or even better in the future.