Fixing a Greenlee DM-820 with bad RMS readings

[bitseeker]

On my friend's DM-820 multimeter, the AC V and mV settings give incorrect results. When he originally encountered the problem, the value it would show would be too low and drift around. However, when I test it now, it's stable and the measured value changes proportional to the magnitude of the input, but the value is incorrect. For example, a 1V 100Hz sine wave will show 0.1 V instead of 0.7V, 10V shows 1V, etc.

What I've checked thus far:

1. Switching between auto-ranging and manual doesn't correct it.
2. Input protection resistors, spark gaps, and PTCs are OK.
3. Other measurement modes such as DCV and Ohms measure their respective values correctly.
4. I removed the 22uF 16V electrolytic capacitor that controls the time constant for the true RMS converter (AD737). It measures 23uF with 3.4 Ohms ESR, which is OK on most ESR charts.
5. Measuring a 1V, 100Hz sine without the cap installed, which makes the AD737 act like an average responding converter, still shows the same incorrect 0.1V.
6. Connected a capacitor decade to the pads where the elcap was. Different values (2 - 99 uF) make little to no difference to the measured value.
7. I tried comparing the output of the AD737 in my DM-820A (successor to the DM-820) when both meters are fed the same AC signal. The output is different, but my meter also uses different capacitor values. So, I don't know that it's a valid comparison. Same IC, but different configuration, which may also have different scaling.

I'm not sure that the RMS converter is faulty, but it seems that way. Recommendations? Other checks to make?

[kasone]

You are beyond my abilities with your troubleshooting.

You may want to read this post: https://www.eevblog.com/forum/testgear/greenlee-lifetime-warranty-an-answer-almost/

It may still be under warranty and replaced by Greenlee.

[Andy Watson]

4. I removed the 22uF 16V electrolytic capacitor that controls the time constant for the true RMS converter (AD737). It measures 23uF with 3.4 Ohms ESR, which is OK on most ESR charts.

Did you check it for leakage?

[bitseeker]

You are beyond my abilities with your troubleshooting.

You may want to read this post: https://www.eevblog.com/forum/testgear/greenlee-lifetime-warranty-an-answer-almost/

It may still be under warranty and replaced by Greenlee.

Thanks, kasone. All ideas are worthwhile. Unfortunately, since he bought it used Greenlee said it wasn't covered.

[bitseeker]

4. I removed the 22uF 16V electrolytic capacitor that controls the time constant for the true RMS converter (AD737). It measures 23uF with 3.4 Ohms ESR, which is OK on most ESR charts.

Did you check it for leakage?

It showed a little over 1% leakage on the tester. Nevertheless, I did test the meter with a capacitance decade box to remove the original capacitor from the equation. Even with no cap, the meter doesn't do basic averaging correctly.

One thing I didn't check yet is if the resistors that determine the scaling factor are OK. If one or more of them went bad, then that could explain the behavior I'm seeing.

[Ordinaryman1971]

I have one of those, would you like to stare and compare... or maybe I can do any measurements for you that would ballpark the problem.
I've attached the picture of the meter.. .is it the one you are talking about, this is the older version of Greenlee?

[bitseeker]

Hi Ordinaryman. Thanks for joining in. Yes, that is the same model that I'm troubleshooting. If you're comfortable probing inside it, the RMS converter chip is an SOIC8 under the lower-right corner of the LCD. Pin 1 is at the bottom-left corner of the IC when viewing the meter in the normal orientation.

Tonight, I'll take some specific measurements at the pins and post here for comparison. I should take some photos too.

[bitseeker]

The test setup is as follows:

1. COM of the measurement DMM connected to COM of the device under test (DUT).
2. AC signal from function generator set for 100 Hz sine wave with 1V peak (0.7 Vrms) into 10 MOhm load (DUT). Positive lead into V input of DUT. Negative lead into COM of DUT.
3. DUT set to V, auto-range mode.

The following values were measured with no input signal. All pin references are for the AD737 IC, labeled U3 on the PCB, located under the bottom-right corner of the LCD. Pin #1 is near C25.

Pin 1: 0.272VDC
Pin 2: 0.27VDC initially and constantly falling to 0
Pin 3: 0VDC
Pin 4: -3.21VDC
Pin 5: -2.73VDC, drops to -2.68V. Display on DUT goes from 0.002 to 0.068
Pin 6: 0VDC
Pin 7: 2.80VDC
Pin 8: 0VDC

Next with AC signal applied.

Pin 1: 0.137VDC (101mVAC @ 100Hz)
Pin 2: 0.1VDC dropping to 0V (42.8mVAC @ 100Hz)
Pin 3: 0VDC
Pin 4: -3.21VDC
Pin 5: -2.67VDC
Pin 6: 5.1mVDC
Pin 7: 2.8VDC
Pin 8: 0VDC

According to the datasheet, Pin 1 or 2 can be the input and Pin 6 is the output (see Figure 27 and Figure 28).

So, I get 5.1mV output on pin 6 for a 1V sine input. I'm unclear which pin (1 or 2) is being used as the input. They both have a 100Hz AC signal.