ESR meter upgrades

[eccerr0r]

would expect digitally used transistors to be fairly easy to check for operation... it's either on or off, if it doesn't turn it on completely, it's bad....

[stj]

not so easy though when they are fired with short pulses and drive triac gates.
(solenoid drivers)

[eccerr0r]

but still, you have a bank of solenoids, and one solenoid isn't working...well, something be a bad one on that circuit... (and then the TRIAC the AVRTransistortest pukes on...)

[stj]

but you dont have a wiring / loom schematic if the customer just brings the board.
i dont want slot machines delivered just to fix a control pcb

besides, you often spot parts that are "almost" failed so you can pre-emptivly prevent the board coming back a week later

[eccerr0r]

More fun with avr transistortester:

Plugging in a 4N35 ... detects as a transistor.

That's fine, now test 3 pins on a side at a time...

one side shows up as a transistor. Good.
other side shows up as a diode. Excellent.

Too bad ...can't tell if it's working... or can it... I wonder what it would do if I connected the diode's cathode to the emitter, and connected those to pin 1. Connect the diode's anode to pin 2, and the collector of the transistor to pin 3... what would transistortester report this as...

[stj]

actually - you can test opto's with an adapter

[eccerr0r]

I figured as much (need to lose one of the pins), but by doing this (shorting pins), isolation is lost

So does it actually detect an artificially "broken" "optoisolator" in this way? Tell a triac or scr or bjt or photoresistor or photovoltaic or photodiode or other type of optoisolator?

Incidentally, can it detect autotransformers? heh, probably shows up as a 3 pin short, but then again an inductor shows up as a 2 pin short.

[eccerr0r]

Fail. Autotransformers show up as two resistors...

[stj]

if you have the newer firmware for the opto-test it gives you the transfer value

[eccerr0r]

hmm... thinking about it, trying to test an autotransformer (tapped transformer) might not be a great idea... but depends on how well the circuit is designed....

[stj]

shouldnt it show as an inductor?

[eccerr0r]

It didn't for me... said two resistors, but that's besides the prob..l..em...
hmm. I think I figured out why it reported two resistors. aaah... circuit design.

[momaka]

Originally Posted by eccerr0r

Just wondering... Ever since I got this AVRTransistorTester it's pretty much stayed in my "toy" pile and have yet to actually find a bad transistor or other device that I already knew of its status of good or bad from other tools or debugging techniques... Has anyone felt the same way?
...
Yes this sorta sounds like parade raining but just wonder how other people feel... do they use it a lot? Has it saved you effort? How so?

I use mine almost exclusively as an ESR meter / electrolytic cap tester (for testing caps out of circuit.) I know it probably sounds silly to pull caps out just to test them. At that point, many people will probably opt to just replace the cap(s) in question. But that's not always possible when testing the primary caps in a SMPS, which tend to be expensive. I do like to test them, though, because if bad, and if the PSU has APFC, chances are things will get destroyed on the primary at some point. On the other hand, if the cap(s) is good, I put it back in.

I also use my GM328 once in a while for measuring inductance on mostly larger inductors / non-SMPS transformers.

I rarely use if for anything else, though.

Originally Posted by eccerr0r

BTW seems that having big caps inserted on power up really slows down the power on auto self-cal it seems haha.

Yeah, I've noticed that too.
Not only that, but sometimes my GM328 will complain it hasn't been calibrated if I insert a relatively large cap (typically 4700 uF and up)... though not always. Drives me nuts when it does that, especially if it does it a few times in a row. It still works if I don't calibrate it, but the constant nagging message is annoying, so I end up calibrating it right away... until it does it again. I do wonder if the MCU memory is going bad, though. It never used to complain about the calibration probably for the first 3-4 years after I bought it and calibrated it (if not more.) Actually, it really only started doing it very often this past year.

Originally Posted by eccerr0r

Fail. Autotransformers show up as two resistors...

Probably the inductance is too low to test??
I know at least my GM328 can't go lower than about 20 uH, IIRC. And sometimes even slightly higher than that will still report as a short-circuit / low value resistor, but no inductance (case in point: most output inductors on PSUs.) VRM inductors from motherboards are completely out of the question here, with most nowadays being under 1 uH. I actually performed a curious test: I gathered several such VRM inductors from scrap Xbox 360 motherboards and wired them in series. I don't know what their inductance is, but based on the circuit design's age, I'd guesstimate 2.2 uH or less. Wiring 4 of them in series did not give any inductance. I think even 6 of them didn't. And at 8x series, it reported 0.1 mH (10 uH) only half the time.

So yeah, at least the GM328 is not very useful for testing low-value inductors.

[eccerr0r]

Originally Posted by momaka

Probably the inductance is too low to test??

I thought if I hook up any two wires to the tester, it will give me the expected reading; if I connect all three to the tester, it gives me two resistors...

[stj]

it works a lot better if you use my firmware instead of the factory one.
it's had the esr/inductor code re-written since the chinese compiled it.
also, a 20MHz crystal increases the accuracy.
(but you need to reprogram it to recognise the crystal)

[eccerr0r]

So what does your firmware report when you have an autotransformer attached?

Two inductors in series doesn't quite work for an autotransformer...

[stj]

i dont have one.
what the newer firmware does is take several averaged readings at different frequencies to get a better result.

[eccerr0r]

Originally Posted by stj

i dont have one.
what the newer firmware does is take several averaged readings at different frequencies to get a better result.

If it's just that... it's still wrong

[stj]

the esr is more accurate

[eccerr0r]

esr of inductors is supposed to be frequency dependent, but telling inductors from resistors is different...