Thanks to a couple subscribers on here, I now have the link to the actual schematic this project was 'inspired by' 😁 You can see my rough incomplete reverse engineering did identify the main functions of the Ring Tester and I correctly determined how the circuit operates, but I also made a few mistakes as well. In particular i got the 4015 IC pinout wrong which is why the order of the LEDS seemed strange bobparker.net.au/lopt_tester/k7205.pdf
I have to say, again, thank you for spending so much time on us! And always with a smile and positive attitude. You are always fun to watch. And again -The very best channel for beginners on electronics- 😀👍 Keep it up.
For anyone with an LCR meter: D(issipation) and Q(uality factor) provide this information too - no need for a ring-down meter. Transformers with shorted turns have miserable Q/high D.
the way to test them is with an LCR meter measure primary then secondary take the square root of the ratio Lp/Ls it should be qual within 1% to the turns ratio Np/Ns or voltage ratio Vp/Vs if it`s more than 1-2% something is wrong if it`s way off you have a short 100%
That may work moderately well for true transformers but it probably won't for a flyback inductor. The flyback inductor, commonly but inappropriately called a transformer, is not designed with a turns ratio based on input/output voltage, but on inductances as such.
There's a more convenient way of finding out all connections, by taking 2 pictures. In your case it's very easy because it's perfectly flat (otherwise you should take a picture) 1) Scan both sides of the PCB (place it precisely so you won't have to rotate the image afterwards) 2) import them in a image software (like GIMP) 3) Copy one of the image (the one containing no component IDs) and add it to the other image as a new layer 4) Make a mirror vertical and horizontal with that new layer 5) Play with the layer transparency and move it so it matches the other => you'll see all connections directly ! I've done it with a power supply and comes in very handy !
I made an inductance meter using an Arduino, an LCR resonant circuit (you connect the L - inductor) and the LM393. But instead of counting the rings, it measured the frequency of the rings (oscillations) which allows you to calculate the inductance value. These two things could easily be combined into one device, but I don't know how would you measure pulse length without a microcontroller.
Great job of reverse engineering, it's a pleasure to watch you work! It would have been interesting to see the oscilloscope with the shorted "winding" as added proof of the concept. Regards, David
@@LearnElectronicsRepair and also short the turn whilst the ring tester is powered up (if you have spare hands, I did notice you powered it off when messing with the 'shorted turn' wire), it should go down to the single red led and back up to them all being on when the short is removed... if you see the scope screen at the same time you'll see the ring substantially damped when shorted :)
Quite a nice little circuit, just to give you an idea of the state of the transformer. If i did more mains work i might be tempted to make one. I did a video on your first transistor logic circuit on the last video, don't know if it popped up in your feed or not. Could do with a bit of your sunlight over here if you have some to spare ;-)
@@LearnElectronicsRepair Well thanks for popping on by, please comment if I've made any errors. And thanks for the sub, slowly working to the magic 1000.
hey - thanks for your videos. I'm a mechanic and always wanted to learn about electronics and I've learned more by watching a couple of your videos than what I picked up during the last 50 years. since you seem to listen to user feedback -- please loose that stationshop biro. i saw you once writing with a black felt-tip and immediately felt better :D or try one of those gel-pens. anyway - thx!
I had some PCBWay pens for a while (freebies) and they were great, but the ink didn't last long as I used them a lot. I have an eye on buying something else.... watch this space 😁
kalli71 You stating you're a mechanic with an interest in learning more about electronics. So much stuff you picked up over 50 years. My score is at 67 I had to retire a couple of years earlier than I was planning. I had experience as an industrial electrician for five years as an apprenticeship when electronics controls were assimilated into electromechanical control panels. All 415v three phase because the motors were the design brief. My college training where I was forced into crossing from industrial electrics to industrial electronics. That may sound ungrateful because I knew electronics was the future. The City and Guilds or UEI examination body was a failure for me. Without being ungrateful and too long winded my conversion to industrial electronics was less than perfect. I would have benefitted from staying with electrics and assimilating to electronics with the use of manufactured catalogues. Catalogues were so much my teacher Learning gas engineering working for a major industrial heating manufacturer who moved me into the R&D department. Working with the cleverest research engineers, each department, gas heating, oil heating, ventilation as different trades such as the UK engineering practice from the 1930's. I was the clever kid with electrical and electronics knowledge only as using built controllers. Landis and Gyr, Danfoss, Satronic, Honeywell, Fyreye to control the ignition and safety flame control. Stuff that prevented boilers, air heaters, process control came later from exploding. Interpreting the design of the circuitry using combustion controllers, the ancillary devices, damper controls came as second nature. Enough for now at 03.38hrs. So much of UK heating manufacturering using electronics didn't move with the times. Maybe harsh but looking back time has checked out my opinions. Mainly we didn't train kids, both sexes, the tradesman, time served, UK governments, tech college's, something didn't happen. Certainly abandoning apprenticeships for years when Europe, the US were training and investing in the next generations. I was observing and weeping inside! Best wishes All
Thanks for the explanation and reverse engineering. I was looking at a circuit with an 8-bit Shift Register in it, last night, and wasn't quite sure how it was working. Your explanation really helped fill in the blanks for me. ☺
Hi Rich, Have to ask a question about the ringing. Would it not be possible to put an oscilloscope on the secondary to see the ring? BTW, I'm absolutely hooked on your videos. Great work.
excellent i think i will order the boards and make one. at the moment i use the cal output on my scope and zoom in on the wave form and count the rings. On mains transformers i might get 2 rings on the primary and none on the seconderys
That kind of goes along with what I found when testing the secondary of the 50Hz mains transformer. Though I omitted to test the primary, the best I could get on the secondary was one ring, as I think the first LED lights up just from the 'trigger' pulse, even if the tester is not connected to a transformer
Really interesting to learn all the aspects of defining what a flyback tester is for, I've heard the term previously but no more. Working with gaps in the items associated with such a project, namely the schematic, then reverse engineering using the PCB was a gift to all the viewers. I've handled transformers since maybe my first day at work, October 1963, assisting my foreman to build a control panel for a processing machine in a leather tannery. In the UK there was inadequate sources of electrical contactor's for industrial motor control, MEM, Midlands Electrical Manufacturing, based in Birmingham, founded in 1908 building switch and fusegear to British standards. RS Components based in London was a 16 page small paper of electronic components, the really bare essentials, resistors, wire and carbon potentiometers, capacitors, some small useful enclosures, not a lot else.
Use your search engine and look for "flyback tester PDF lowcost" to find the original project or "bob basic flyback tester" on youtube for an improved version from the same person.
The LM358 is not pin compatible with the LM393, the LM358 is a dual opamp while the LM393 is a dual comparator. Yes, with additional components an opamp can be converted to a comparator, however they certainly are not plug in compatible devices. Also, I believe the LM393 has an open collector output which would mean it can only sink current, not source it, which would in many situations make it incompatible with standard opamp output circuitry.
The circuit should use a LM393 and I didn't realise because in the video the author (who appropriated this circuit it would seem) showed an LM358 in the version he built, so it just so happened I didn't have any LM393 to hand. Google does say you can substitute LM393 with LM358 (apart from the higher current/voltage capabilities) but then again, when i built the PCBWay millivolt meter, Google told me that a 741 op amp is a suitable replacement for OP07, which turned out to be 50% correct when I tried it LOL
@@LearnElectronicsRepair They can SOMETIMES be compatible in SOME situations. If someone is using an OP07, that is one of the first "precision" opamps made in the electronics industry, so there is likely a reason they are using that part, over some other standard jellybean single opamp. If someone is using an LM393, they likely have a reason they want a specific comparator design opamp (for instance as mentioned above its output electronic design is completely different from that of a standard opamp) as opposed to a jellybean general one, again. Sometimes a jellybean part can do the job of a specific part but just worse, likely the case here (and for a course case, in the end that might not matter that much). Sometimes the design relies on a feature the jellybean doesn't have that would fail, like needing manual offset, or needing rail-to-rail input and/or output capability, super fast speed, etc. Whenever you see a non-standard part you can start asking if you can find a reason why they used that part instead of a LM358/LM324, etc.
Just a funny FYI, Richard....at 12:29 my iPhone took "Two hundred and forty volts AC coming out", as "Hey Siri", then tried to come up with suggestions, not always the same ones. After rewinding a few times, to 12:19, it's pretty repeatable, maybe you could try to come up with words which follow "AC", to mess with us that have Siri sat waiting to 'help', in your future videos? You need to have Siri set to be available on the lock screen for this to work though. Thanks for the unintended laugh.
Good explanation, reverse engineering. Big Clive, eat your heart out, ho, ho. One of the associated videos shows 2 x 18650's powering it (so 7V or so). Yes, these testers work best with high Q transformers (line output etc), search Blue Ring tester for a similar (not the same) circuit.
Thanks Ralph that's a big complement being mentioned in the same sentence as Big Clive in a positive way 😊 I have seen that Blue Ring Tester before so probably I should build one too for comparison
I went and had a look at the guy's youtube vid (ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-58VCLHsdJ38.html) but I don't speak Indonesian !! I enjoyed your reverse engineering. There appears to be two positions for the transistor or am I seeing things please ? what we saw was a 'go-nogo' sort of test , what would we do if it only lit up to the amber LED's what would this tell us ? that would be tricky, is it good or bad please? is there a scale like full scale is X amount of rings ? ...cheers.
Great video about reverse engineering, but.... What about the strange connection of LEDs, when the last green is in the OUT4B... And DATA A is connected to OUT4A... It has no sense to me... Should you check it again or explain it little more? Thank
It didn't make much sense to me, but I was talking to Carlos from Retro Upgrade today about using AI (which is something he is working with a lot) to produce a complete schematic from a PCB layout - using this PCB, which I gave him, as an example, so I am sure you will be seeing more of it later
David O I now have the schematic (see the pinned comment) and it turns out my reverse engineering was OK, but I mixed up the pin order on the 4015 😅 So that's the answer to the paradox
From experience I can tell you that "component tester" from aliexpress can test coils quite OK, so if I have to have transformer out of the circuit, I will not bother to make this device really, I will just make chart of resistances and inductance then I will figure it out if its good or not. These days there is 2in1, component tester + oscilloscope in one device.
I WOULD LIKE TO CONTINUE WATCHING YOUR SHOW AND IM A LARGE SCALE PROMOTER. BUT I NEED YOU TO DO SOMETHING ABOUT THE BEEEEEP. ITS JUST WAYYYY TOOOO MUCH. IT NEEDS TO BE TONED WAY WAY WAY DOWN. I CAN'T TAKE IT THE AUDIENCE CANT TAKE IT AND IT TAKE YOU 5 MINUTES TO PUT A DROP OF SILICONE OVER THE LITTLE BERPER. ITS JUST WAYYYYY TOOOK LOUD. WE TURN OUR VOLUME ALL THE WAY UP TO HEAR YOU TALK AND THE BEEP IS DRILLING A HOLE IN OUR HEADS
This a project by Bob parker. Bob has made 2 improved versions, one digital logic based, and one microcontroller based. Bob made a good ESR meter too. Bob is on youtube, badcaps and eevblog. You can find the schematics of all on the net (also this one, with a complete explanation). I tried to post how to find them, but my comment got removed.
Its odd the guy who made these PCB's is Indonesian and his name is Fareed Read I checked it all out on the PCBWay site, sadly no schematics though, I will check out Bob Parker...cheers !
Assuming your reverse circuit is correct, the input stage to the LM393 comparator is a LOW Q because of the low value resistors after the input "hot" capacitor (it's funny that he used a high quality high Q cap and then destroys the Q on purpose???) , which will destroy any high inductance rings (like the transformer you tried it with), ...lower inductance will fair better because they cap will not be subject to the resistor current drain (Q leakage) for as long, but at low frequency nearly all the energy will be drained by the time even the second ring tries to start!. It is a VERY BAD design! For reference, anything above about 30-50 Meg Ohms will have a noticeable affect on the Q (ring count), he has values 3-4 orders of value TOO LOW. The circuit is intended for low cost, but he failed because with the same components (except resistor values and bias current compensation) , a much better design can be had. ....SORRY MAN, IT IS GARBAGE!
BTW, I suspect that is why he said to use the secondary of the transformer, ..it is the much lower inductance part (normally). He saw the issue with his circuit, but no WHY this happens!
@@stevenbliss989 I can't say for certain the reverse engineering was accurate, as I mentioned in the video I was showing that sometimes you just need to work out enough stuff to understand how the whole thing works, then you can test the circuit and if it doesn't work for example a circuit that should be oscillating but isn't, you can then go deeper into actually completely reverse engineering that part of the circuit to find the actual fault. Someone has sent me what he thinks is the actual schematic for the Ring Tester so I can play around more with this - maybe we can improve it too 😉
@@LearnElectronicsRepair Yep, looks like it. C3 is the "RING" cap of course, and it's a good value for a wide range of inductance, BUT, for the ring to persist it must have a high Q. A shorted turn is exactly what kills the Q by sapping energy from the rings. Sadly so do the input resistors, and their effect is time dependant, such that the lower the frequency of the ringing, the more time they have to DRAIN the energy away (in less than a single ring cycle for high inductances). Luckily the LM393 has VERY low input bias current, and ok bias offset currents. So if you half decently match the input impedances (5% R's should be fine), you can get away with using fairly high values for R11 and R12.. You should, but not required, reduce the input DC blocking cap C4 appropriately. Then to balance things out, you will also need to add an appropriate value resistor from th junction of R13 & R14 and the +ve input (i.e. cut track and insert). If you don't, you will get an unwanted offset that can fuck things up. As a starfting point I would suggest trying R11 & R12 at 220k & 1Meg , put a series 1.2Meg resistor from the junction of R13 & R14 as described. And lastly, since R10 is part of a voltage divider with R11 & R12, it will also need to become 31.3K, ...so 33k maybe. The LM393 is NOT rail to rail input, but it is ok with ground. Surprised it gets away with 4.7k (now 33k) because it really should be MUCH higher. I would try values much higher, but it will mess with the detection threshold, which MUST have the + input holding the output +eve with no rings, so watch your voltage dividers. Neither the + or - input should never go above Vcc - 1.5 (4.5V with a 6V battery) ...see datasheet common mode input range. I'd rather you work out by math and some trial and error all this, my brain hurts! :) Suggest using LTSpice XIII to try out voltages (LM393 will have to be imported as not part of standard library ...the ones I found online are fairly crap but good enough for this, just watch your input voltage to never exceed Vcc - 1.5V.. OK, there's something to start with :) ...have fun! :):):) BTW even 1 Meg is still somewhat too low, but a VAST improvement on what is there right now. Higher values may be possible, again, try it and have fun. Just remember, and to be repetitive, the goal is to stop energy being taken out of the C|L tank rings. Ideally a JFET input comparator, or JFET input buffer + LM393 would be better (with MUCH higher R's of course). But ideal may not be needed here, just better than the horror there is now. You will know you are doing better because the rings will decay slower at higher inductances ...more LEDS on :) You get 1/2 righ now, even 5 would be better. After all you want the shorted turns to steal the ring energy, NOT the input stage of the detector. ....HOPE THIS HELPS :)
@@LearnElectronicsRepair I have not looked at the pulse generation, but a BJT switch is low leak, so maybe this part might be good. And one other point; you cannot go too high on the R's because the leakage of diode D3 will become a factor (somewhere around 1-5Meg for R10 this will happen), It's just a discharge path for C5, and I would say not needed anyway. ...R10 / D3 is not really an issue, just saying, be aware. :)
