ESR meter strange readings

[valveaudio]

I recently built the ESR meter to the VE71T design it works fine, but:

I set the FSD by shorting the connection terminals with a 4cm wire link all OK.

When I removed the shorting link and connected two 90cm test leads to the terminals and shorted out the test probes I got a higher deflection on the meter !!! equivalent to about 52 ua indicating of course that the leads were lower resistance than the shorting link. It is fitted with a 5v regulator.

Not really a problem since I have reset the FSD with the test leads in place, just wondered if any one else has seen this and knows what causes it.

Trevor

[GMB]

Well I would say that you just demonstrated that it doesn't work fine!!

I think a little investigation would be in order.
There are lots of possible reasons for this. Your original link could have been high resistance or not connected properly.
If the meter is a traditional one then it might have damaged bearings so keeps giving errors, or you changed something, like the meter was in a different orientation when you got the error.
Maybe the circuit is prone to giving erroneous readings when mains hum is picked up? If that is true then you really want to sort that out as it could have more dramatic effects sometimes.
I would try different length leads to see what effects you get, and see if the way the leads are arranged effects it, or if you touch the connections.

[jimmc101]

I suspect that the inductance of the test leads is partially resonating with the 100nF capacitor across the 100 ohm resistor.
If I'm on the right track then coiling the leads to raise their inductance will give an even higher reading.

Treating the source as a voltage generator with a 10 ohm series resistor then...
With zero inductance the loss between the generator and the 100 ohm // 100nF is about 6dB (at 100kHz),
With 2uH of inductance this reduces to about 5.6dB
With 20uH (resonance) the loss is about 2dB giving over 50% more output.

Jim

[valveaudio]

Thanks for the suggestions.

The test I carried out was out of range of any mains influence, the shorting link was screwed to the terminals, the meter has not been moved while testing, holding or coiling the leads has no effect.

If I extend the test leads by 25cm there is no difference if I shorten them by 25cm there is no difference.

If I short out the test probes (having reset the FSD) meter reads 50uA but if I leave these shorted and also short out the terminals the meter only reads about 48uA.
All my findings are readily repeatable.

Trevor

[David G4EBT]

I don’t know if you are referring to the ESR meter I wrote about in a thread I started way back on 20 May 2010.

It was so long ago, I struggled to find the thread, which I see has had 106 post and almost 41,000 views, so I guess along the way, scores of queries and answers have been given.

The thread is here:

https://www.vintage-radio.net/forum/...brew+esr+meter

If you are indeed referring the that project Trevor, originally I included a zero adjust control on the front panel to enable the meter to be zeroed (that is set to full scale deflection with the test leads shorted), just as one would with an analogue multi-meter on the Ohms range, such as an AVO8.

I subsequently removed the front panel zero pot, altered the PCB track layout and fitted a pre-set pot on the PCB. But once the meter has been zeroed on short test leads, if you then use longer leads, they will introduce stray capacitance and inductance not present in the shorter leads, so unless you have a front panel pot to set the meter to zero it's best to only use fixed short test leads. That's why - to minimise any stray capacitance in the test leads - I fitted two short leads directly wired to the PCB, protruding through the front panel about 3”, rather than to use plug in test leads.

My reason for removing the front panel zero pot that was that was outlined in post #36 of the thread, as follows:

In the original design a front-panel mounted pot was necessary as the designer used rechargeable cells to power the meter. The IC used (74HC14N) will work at voltages between 2 to 6V, but by using rechargeable cells, which – when fully charged will be 4.8V, and will self-discharge even when the meter isn’t used. This required that the meter be adjusted from time to time to get it to read FSD with the test clips shorted.

I preferred to use a 9V PP3 battery rather than rechargeable cells, so from the outset I altered the PCB layout to include a 0.1A 5V TS78L05CT regulator so as enable a PP3 to be used. The inclusion of the 5V regulators meant that once the meter adjustment pre-set pot has been set to give FSD on the meter with the test leads shorted, it ought not to need adjustment again, and should continue read FSD until the battery expires. (The meter only draws a few mA, and is used only briefly, so a PP3 should last a long time). The modified the PCB layout with the vertical pre-set pot, mounted on the PCB removed any complications in wiring the pot and meter movement, which – in the later layout, needed just two connections from the PCB.

It’s worth reiterating that when measuring ESR, really it’s a go/no go test – we aren’t looking for a precise reading. That's why the meter dial is marked 'Good, 'Compare' Bad'. (See pic below). If the ESR of an electrolytic is much more than one Ohm, it’s best to compare the suspect one with a new one. If the new one is a fraction of an Ohm, and the suspect one is several Ohms, then clearly it needs replacement.

It does depend on the value of the cap and its working Voltage, and in this regard, Bob Parker is I think the acknowledged expert on ESR drew up a table of approximate worst-case (highest) ESR values for new electrolytic capacitors at 20 degrees C (68'F). ESR generally decreases with increasing temperature. I posted somewhere or another on the forum but have appended it below.

It's nice to know that this little project is still attracting spasmodic interest, despite the fact that for some time now, digital 'multi-test' meters that test capacitance including ESR, resistance, inductance, transistors, FETs and gives a direct readout have been available on ebay for about a tenner. The little cheapo multi-testers do rather knock the idea of a home-brew ESR meter in the head. They come just as a 'bare bones' panel and can be used that way, or fitted in a box of some sort. Some come with a ZIF socket which is a bit of a nuisance so I removed that from mine and fitted three front panel sockets. A couple of pics below which show the readout, and how I housed the digital meter with a front panel push button and test socket.

Hope that helps a bit.

[valveaudio]

Thanks for the info,

I have reset the FSD and calibration using the test leads and this seems to be reliable so I can now get on with some testing. Already found a faulty capacitor in a Leak TL12!

Trevor