How accurate needs an ESR meter be?

[Th3_uN1Qu3]

I've gotten myself a better movement for my ESR meter (well, a bigger VU meter). I used the schematics here. The scale on the original design starts at 20 ohms. I noticed that mine was reading down to 50 ohms which made it quite hit and miss in caps that had just started to go bad, so i needed to change that.

I paralleled a 1.8 ohm resistor with the 10 ohm one that was in series with the cap. Now the bottom of the scale is just 5 ohms, and with this new setup i could easily spot one 470uF 10v cap that had 0.3 ohms ESR. 0.1 ohm differences are now very obvious, with 0.05 ohms visible if i watch closely. I'm thinking 0.01 ohm resolution would be nice.

The movement doesn't deflect full scale anymore but it does get to the "0" of the VU meter so it's still handy. If i raise the gain on the opamp and change the measurement resistor with an even lower value, i can probably get an even finer range. So, how low do you need to go? Edit: Methinks this really needs a range switch. On a side note, it can also measure the ESR of batteries - now i understand why my old NiMHs won't run the digital camera anymore, even though they do fine in lower current stuff.

[devid21]

Re: How accurate needs an ESR meter be?

[devid21]

Re: How accurate needs an ESR meter be?

Hi Th3 uN1Qu3
My experiments with this scheme showed that it is possible to reach the limit of measurement below 0,01 Om.
Then the upper limit of measurement will be about 3-5 Ohms.
For capacitors with LOW ESR meter can be done with such limits of the scale of 0-5 Ohms,given that the capacitors with LOW capacitance is not very often refuse to work.
Good luck

[kc8adu]

Re: How accurate needs an ESR meter be?

to really test correctly you need to go down to milliohms.

[Th3_uN1Qu3]

Re: How accurate needs an ESR meter be?

Well, thanks for the challenge then. It should be doable with two 1.8 ohm resistors in parallel for the series resistor of the cap and an extra gain stage before the meter. But in this case it'll definitely need a range switch, as the highest resistance that will move the meter will be something like 0.3 ohm, making it unsuitable for testing small caps.

[devid21]

Re: How accurate needs an ESR meter be?

Transformer you did yourself or ready to use?
He´s here is of great importance.

[Th3_uN1Qu3]

Re: How accurate needs an ESR meter be?

Transformer is the base drive transformer from an ATX supply.

[devid21]

Re: How accurate needs an ESR meter be?

For example,like this
http://www.py2bbs.qsl.br/esr_meter.php

[PCBONEZ]

Re: How accurate needs an ESR meter be?

Precision and accuracy are completely different things.

You guys seem to be ignoring the effects of stacked [tolerance, deviation, accuracy] on overall accuracy.
That's way more math than I feel like doing.
Just saying it's there, it's real, and it adds up.
.

[PCBONEZ]

Re: How accurate needs an ESR meter be?

If the actual value is 2.002

1.993 is more Precise.
but
2.0 is more Accurate.

.

[devid21]

Re: How accurate needs an ESR meter be?

And who need your mathematical precision?
The device should look bad capacitors,which the successfully does.
But you have made at least one ESR meter own hands?

[Th3_uN1Qu3]

Re: How accurate needs an ESR meter be?

You guys seem to be ignoring the effects of stacked [tolerance, deviation, accuracy] on overall accuracy.

The instrument is not extremely linear as it is, and adding a gain stage will make it even more non linear, a VU meter makes a crappy display, and bla bla bla. But it DOESN'T MATTER as long as the resolution is satisfactory. Boards don't care about math you know. It's like saying that a 7805 doesn't produce 5.00 volts but 4.95 volts. So? Btw, this ESR meter circuit is the only circuit that i had LTSpice choke on - it doesn't seem to understand that the 1N4148 diodes that rectify the voltage going to the meter let quite a bit more voltage thru than it thinks they do. Anyway i'll be building that gain stage in a few days, got an exam on Wednesday so i can't be fooling around too much atm, but after that i will most definitely build it and see how it goes.

Here's another one - most cap datasheets say that tan delta (hence ESR) is allowed to vary up to 200% during the life of the cap. If i remember right, you were the one that said that a capacitor with 0.004 ESR in a motherboard is failed if it has 0.008. That isn't the case. Maybe if it had 0.009.

[PCBONEZ]

Re: How accurate needs an ESR meter be?

0.00800000000000000000000000000001

[mathog]

Re: How accurate needs an ESR meter be?

0.00800000000000000000000000000001

Is there any physical measurement that has that many significant digits?

[Th3_uN1Qu3]

Re: How accurate needs an ESR meter be?

I think he was joking.

[PCBONEZ]

Re: How accurate needs an ESR meter be?

I think he was joking.

You're a good thinker.

[PCBONEZ]

Re: How accurate needs an ESR meter be?

Is there any physical measurement that has that many significant digits?

Yes, in Nuclear Engineering.

[shovenose]

Re: How accurate needs an ESR meter be?

Yes, in Nuclear Engineering.

lol. and in cpu manufacturing i mean 32nm DAYUM THATS TINY

[Th3_uN1Qu3]

Re: How accurate needs an ESR meter be?

You're a good thinker.

I'm better at building tho.

[devid21]

Re: How accurate needs an ESR meter be?

I want to explain to some nuclear physicists why we need a transformer in this circuit.
It compensates for the error introduced by masuring probes therefore it is possible to use longer probes.
Transformer provides more current measurement,so it is possible to measure small values of ESR
This can not be achieved without transformer circuits.