Evolution of the $1 Short Finder and Milli Ohm Meter Projects in chronological order: 1. Build this Amazing $1 Short Finder (original video) ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-P_2GGNr4q1s.html 2. The Amazing $1 Short Finder (testing on low voltage circuits GPU/CPU to prove it will not cause any damage ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-XEbWyQgX_Mc.html 3. The Amazing Short Finder Upgraded (Kelvin Prove Version) ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-0eixDdCpiO4.html 4. The MilliOhm Meter Project Part 1 ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-rDw3oVqJHHI.html 5. The MilliOhm Meter Project Part 2 ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Sz1CqJltWEY.html 6. The plus/minus 12V Linear PSU converted from a NiCad/NiMH Battery Charger (used to power the Milli Ohm Meter) ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-1pAOBPmC7Fw.html
@@paulc9139 Yes possibly - or the PSU voltage is dropping due to the 16V electrolytics not being good enough and they have clearly bulged, I'll replace them with some 25V instead.
Congratulations on 50.000 Subscribers! 15:57 - How to increase speed in measuring? If you need to take 5 measurements, each of 6-7 seconds its kinda slow or there is a chance that someone will stop measuring before real result, simply by mistake and then he will desolder wrong part and lose time. And another question, could it be that capacitor value can make difference in speed while measuring? Should be like that... 8:00 - What if we use lets say, power bank, 5V 2A, with large capacity, it should be possible? On field work, we cant bring bench PS, but we can power bank. 32:42 - I think you need to lift your feet and arms in air to stop it going to negative :) or it could be simply your skin charged.
Many many years ago I was involved in upgrading 16k memory boards to 64k. desolder all the 16k chips then insert 64k chips and a couple extra address line, job done. Results - short circuits all over the place. I used a micro ohm meter to locate the shorts, this tester had no display (yes none) but it would screech, as your probs neared the site of the short the pitch rose higher and higher till you were spot on the short. No need to keep looking at a meter all the time. :A large magnifying lens was used to physically see the short. A quick touch with a fine tipped soldering iron or vacuum desoldering iron removed the solder bridges. Best ever tool in the workshop!
dis-similar metals can give a voltage of several mV when they touch. Also, the current generated by the short tester could be heating the probe junction which would cause the mV reading to drift. So using as low a current as possible may give quicker, more stable results?
Yeah I have to say I was really impressed with that performance - it's amazing what you can do with mostly salvaged components and a few dollars worth if parts, Most expensive thing is the 5 Digit 3 Wires LED Digital Voltmeter at about $7 but you could just use an external multimeter with mV range instead. Other than that I had to buy the OP07 op-amps after failing to get it to work with UA741 but they were about $1 for a pack of 10, and the 4mm banana sockets were also cheap. I also think it looks pretty good as well for a home built tester. (affiliated links) www.aliexpress.com/item/4001056150137.html?af=ler2022 www.aliexpress.com/item/1005002957467146.html?af=ler2022 www.aliexpress.com/item/4000830210804.html?af=ler2022 www.aliexpress.com/item/1005004689666970.html?af=ler2022 Pretty much everything else was salvaged or I already had lying around. You have to hand select the 10K and 100K resistors - see earlier video.
@@LearnElectronicsRepair Thanks I am looking at building both of them, the original one with the multimeter you had (AN8008) and the other one with the affliaited links (I will order from them in a little bit). I was just working on a board today and suspected there was a BGA chip that had caused a short circuit because there were short circuits on two seperate voltage line (one being Vreg1.8V and Vsys) but after removing it the short is still present. And also I can't confidently go injecting hight voltages to try and heat up the shorted component because these two voltage lines that the shorts are on also go into other microprocessors that would more than likely break if I exceed 2V, given one accepts 1.8V and the other I don't know what voltage it uses, its just called Vsys. Its like in this case having this tool is just extremely useful for saving time and just not potentially breaking other voltage sensitive digital chips. The only other thing I could think of doing if I was really forced to would be to try and disconnect the chips where the voltages go into them and then start heating up capacitors and just solder some very fine wires to connect them back together. I am thinking it may still be possible to do at 1V if the resistance on the short is around 0.1ohm so I may try that but I think this meter is really the best option in general given I can't shrink my self down to the size of fine traces and cut them haha.
With another ‘hidden’ (or lower) feature too: it uses just 1.2V as open circuit voltage 😊! You have already addressed that (I saw your video), but this lower voltage might help other viewers too.
Thx for sharing the video. Seems like you have either hit an input protection or are fwd biasing part of the meter IC for the voltage to get pulled down like that. Since you've added a meter, short finder DMM, with an additional digit you'll also need to readjust the output offset adjustment. As far as the thermal voltages go, I believe they will cancel out as long as the same metals are in contact on both probes and the temperatures are also the same. Think of each of the thermal voltages as batteries or V sources. Their polarities should be in the same direction hence the cancelling. Again, Richard thx for sharing.
Hello Richard. How are you? I am trying to build one one this for myself as well. But I am encountering some problems I cant really explain. Somehow it is no working. I am getting near 0 (0.0049) reading on the volt meter with leads not touching. And one of the OP07 (U1) is heating up. I am afraid the problem can lie on the way I am powering it. I am using a 15V Switch mode power supply attached to a 12V Positive and Negative Dual Output Buck Board module. That Give me +12V 0 -12V. Maybe you know if the problem can be this and why? Any feed back is appreciated! Thanks a lot.
At last I found you sir you doings nice work without any revange for worlds,,,yes because of you I am a good repairing master in electronics I received alots income and respect from your channel thanks aloooots thanks so much you described each things extremely deep and clear love you sir...
Heya, finding shorts is already difficult and wen they are so close to each other it's hazardis but with this milli ohm meter it's not a problem any more and for my budget well inside it so very nice yes
I find the reading changes on the screen because the resistors heat up due to the current passing through them, changing the current. Recommend using high wattage resistors and also heat sinking the LM317 By the way, I saw those other short finder videos but not this one until I had already commented about making a milliohmmeter. Seems you were ahead of me!
Richard, first of all, thank you for this series of posts, quite interesting! I have some questions and would appreciate any comment from you or fellows from this channel: Q1 - READING SPEED: Although quite sensitive, it seems the time-constant of the overall micro-ohmmeter is quite long, about 1.0~2.0 second, resulting in a stable reading after 5~10 seconds. Both displays are much faster than this (for instance when measuring a 1.5V battery + voltage divider for mV). It means some smoothing capacitances might be too big. Where do you think is the RC responsible for this major tardiness of response? Q2 - OP AMP SELECTION: I see advantages of using those OP07, but outside a few countries (as USA, UK), obtaining non jelly-bean parts are prone to have them ‘chinese-rebranded’, or waiting weeks + dozens of dollars (post+customs). Specifically for the second OpAmp, using the Offset null feature and not requiring highest input impedance, maybe a uA741 would work there, the same might happens for the first one, where even an LM358 might do the constant current-source trick. Based on that, if you have a chance to de-socket and replace the OP07s by LM741s, could you share an updated jelly-bean version of this device regarding its performance?
You are in the domain of thermocouple effects at your probe tips and other connections of dissimilar metals. For reference, a chromel-alumel thermocouple has approximately 40 microvolts developed per degree Celsius. Your finger contact already can heat the probes for multiple degrees, although of course your probes are no real thermocouples, but if I recall, you soldered the two probe tips together, so you already have some dissimilar metals there. Then there are the offset drifts, even if you got the better amplifiers and got the offset null circuit functioning. But there is some unpleasant note on certain op-amp data sheets or application notes mentioning that the offset null operation actually worsens the thermal drift! I would also be a bit suspicious about the stability of your constant current source. That is another place, where you would probably need a Kelvin terminals equipped sense resistor.
Hi Richard... 14:32 to 15:27 in video, the LHS of 5capacitors you show are connected together as with RHS too so why are they later found to be at a different potential just upon the connected rail at the point of the shorted component? Tried to model idea with ladder of resistors which evades me🤔 thx in advance..anyone know please?
With the unit you built from PCB way try playing with the 1uF cap across the meter. You could possibly improve the slow response as the reading settles
What about when the short is partial? I've faced this problem, a cap was partially shorted and the rail was reading couple of ohms, I tried shifting the voltage with external power supply to zero the reading but the short finder was giving unstable readings and I was not able to find the cap using this method. did you face this problem before?
Thats the whole point of this device, a "partial short" is just high resistance short and this setup gives you the chance to look for the path of least resistance by giving it to you in ohms so can lead you to partial shorts as well as 'dead shorts' not sure if I made any sense there ! but hope it helps...cheers.
@@andymouse I get what you mean, but unfortunately it doesn't work with relatively high-resistance partial shorts (couple of ohms) because when you try to find the path of least resistance it's only couple of milliohms in difference which is lost in accuracy with high-resistance shorts, in other words it fluctuates with more milliohms than the tiny difference in the different paths.
@@LearnElectronicsRepair I'm not sure what caused it but it was fluctuating in the 10's of MilliOhms which made it impossible to find the shorted cap using this method, maybe the short resistance was unstable because of the heat from the passing current I'm not sure.
Here's my two penn'orth. It's possible that because you have increased the sensitivity of the meter, it's now sensitive to thermals as in heat is changing the resistance or heat is actually generating very small currents that are biasing the reading. Another reason could be induced noise. You would need some pretty tasty test gear to measure such voodoo, though. Maybe someone in the comments section has a Hewlett Packard 3458A Multimeter that is surplus to requirements, and they will donate it to you. If you didn't want it, I'd gladly have it and give you £100 for your trouble. 😁
Over thinking it , switch the op-amp gain . Micro Ohm Meter = x10 Short finder = x100 You may need a internal trim-pot for the 0 adjust & add a voltage limiter maybe 4 x IN400X ?
Yeah fair point - but my method just required the addition of a couple of 4mm sockets to attach the meter - yours requires more hand picked resistors and a switch to select them, another trim-pot which also may require more switching and some more diodes (maybe) So run that 'overthinking it' thing past me again to convince me I didn't just do it the easy way 😋
@@LearnElectronicsRepair It could be just a extra socket & resistor for the short finder ! ( 2 > 3 pin PCB Molex update + a resistor or a link on PCB & resistors on the sockets ) . I don't mind a small zero offset with the Short finder function as it will drift a bit with temp ..... You are just looking for the lowest value ! PS If it followed good design it should not require hand picked resistors & you could drop 6 maybe 7 original parts ! As for me adding 4 diodes a 5W 3V3 zena will work . But we all have IN400x diodes in the junk box !
As a newbie without a big pile of salvage to rely on, I'm wondering if the project could be further enhanced. Make the changes to your short finder (100mA CC with x10 amplification), but use a single supply opamp or some other way to get rid of the negative rail. Then people like me could just power it from any old USB charger which are really cheap and easy to get.
The problem with that is what is called the offset voltage - a single rail op-amp can not zero accurately, you need a dual voltage supply. I am sure others can tell you more about this.
38:25 i guess that the difference in voltage 12V in volt range, and arround 8V in millivolt range, is due the lower internal impedance of the ANENG meter in the millivolt range.
I like to go up in current to 1A when getting down to the low mOhms, plus i just use a power supply on the contstant current mode. You have not taken in to account the +/N % of reading +x digits that also needs to be addded to what the meter is displaying as a result. So you may lose 1 or more digits of the displayed result.
I have found that the exact temperature of the Kelvin clips can affect the reading on a micro-ohm meter. Just touching one clip more than the other can give different temperatures which will give a big error. I wonder what the best material for the Kelvin clips would be, to reduce this effect?
Same applies to DUT, have a high resolution commercial one and you can see the measured value skyrocket just by placing a finger on the resistor. Copper has very high specific heat value and nearly the lowest resistivity, so hefty copper probes with thermally insulated manipulation surfaces are hard to beat.
Always love your videos. But sometimes the constant continuity beeping makes it distracting when streaming while working in the lab. Or when we’re relaxing and working on my own projects.. Probably nothing that could be done about it but just thought I would mention it FYI
Since your using the Fluke could you use it in the 'low ohm's ' range (40 ohm) please as I'm sure that those of us that use this meter, and there's quite a few ! would like to see a real world comparison of it against the other devices as well, because this feature is often forgotten about by Fluke 79 series 2 owners and I have found it handy in some cases...cheers.
Construí algo muy similar pero no me ha servido 😮.... De alguna manera algunos condensadores marcan unas décimas de ohms más bajos que un fusible. Saco y saco componentes que podrían ser el culpable y nada.. el corto se esconde terriblemente
Well done, very useful test equipment (on my to-do list). Switch on is up!, I expect you to transition from northerner to an American accent fairly soon. Ho, ho.
A much better multimeter can be used. Brymen produces very good (and cheap and easily beat 5times expensive Fluke in same range) multimeter, with 6-digit DC millivolt range resolutio. Something 53xxxx is the best reading. I just tested with a 100mA constant current source and used a Isabellenhütte Manganine Kelvin type calibration resistor with a value of 5.6mill-ohm and with a deviation of up to +-0.01% and I got a stable reading 000560mv i.e 5.60 milli-ohms, so, the resolution was 10 micro-ohm. And this is possible only thanks to LM317 Cconstant current source (you need to trim as close as possible to 100mA) and: NO AMPLIFIER! but if gain 10 amplifier added then 1 micro-ohm resolution seems very possible.
