Fluke 8000A Help Needed

[Goatropercu]

Hello,

I recently inherited my dad's old fluke 8000A multimeter. It seems to work well enough measuring ohms and DC volts, but when I measure AC voltage it wanders fairly erratically. For example, when measuring mains voltage (115vac at my house) the reading jumps around from 110vac to 140vac, and anywhere in between. When connected to a known 6vac voltage source it bounces between 0vac and 5vac. With the leads shorted it bounces up to as much as 9vac.

This meter does have the mA/sec function that I believe was an option. The service manual/schema that I downloaded does not include the additional board that I believe to be associated with this function.

I have cleaned the controls (twice).

I replaced the following:
C6 or C10 hard to say which (.47uf/35v tantalum)
C20 and C9 (10uf/20v tantalums)
C17, C18, C25, and C31 (.47uf/20v tantalums)
C14 (.22uf/50v electrolytic)
I also replaced two 4.7uf/25v tantalums on the additional board that I assume is associated with the mA/sec functionality.

R45 was visibly cracked and spewed apart upon removal. It measured 60ohms both before/in circuit and after removal. I am no expert, but I assume this resistor may be more involved in current measurement rather than the AC volts function. It does look like the range switches are directly tied to these, but since AC volts is off in both the 200vac and 20vac range I do not think R45 is the root cause. Either way a replacement has been ordered.

Is there any way I can tell if the op-amp U1 is bad? It is stamped Motorola 352930 7098.

Any help is greatly appreciated. Thank you.

[wn1fju]

I would first check the output of the buffer (Q1 source) in the AC Converter section and make sure it follows the applied ACV input reliably.  If not, I would suspect the current limiter, CL1.  I had a similar Fluke 8600A DMM with erratic ACV readings and that was the problem.

As far as checking the op amp, U1, when operating properly the two inputs, pins 2 and 3, to the op amp should be very close to equal and the output, pin 6, should generally not be near the +15 or -15 supply voltages.  This goes for almost every op amp.

[Goatropercu]

Pardon my ignorance, but can you explain what you mean by the Q1 source should "follow the applied ACV input reliably"? I assume you mean that the voltage at Q1 source lead should "move" with respect to different ACV's being measured/applied to input, but specifically what am I comparing Q1 source voltage to and where?

Thank you.

[Dave Wise]

I know you said other ranges were working.  For what it's worth, the only thing that's ever failed in my 8000A is the power supply reservoir caps.  When approaching any test instrument for troubleshooting, the first thing to do is confirm that the power supply rails are good, because if they are not, the instrument can malfunction in every way imaginable and then some more.

[Goatropercu]

I am getting +19vdc and -19vdc from power supply, apologies for not mentioning this in my initial post.

Thank you.

[wn1fju]

Q1 acts as a "source follower" and so the output of the transistor should be a "copy" of the AC waveform you are supplying to the front panel jacks.  The amplitudes may be different because there are attenuators in the signal path, depending on the selected range, before reaching Q1.  If you have an oscilloscope, you would be better served using that.  You get a more complete picture of what's going on that by using a voltmeter.  Usually when probing a DMM, your "ground" reference is the negative input terminal. 

I only suggested Q1 because when you are experiencing erratic behavior, you want to logically start at the input and work your way through the signal flow until the observed signal suddenly becomes erratic.  Q1 is a logical place to start because it is before the active rectifier circuitry.  Sometimes this can be difficult if there is feedback in the circuitry you are analyzing - then the question is who is at fault.  At that point, you usually have to take out the soldering iron and lift a component or two to break the feedback before further debugging.

But, as the previous poster suggested, this all assumes the power supplies are working.  Again, a scope is more useful than a voltmeter because you can spot simple things like 120 Hz ripple indicative of a failed electrolytic capacitor or 60 Hz ripple indicative of a failed rectifier. 

[Goatropercu]

This is incredibly helpful. I’m out of my element here, but I do have a scope. Let me see what I can find. Thank you very much for your time and assistance.

[Goatropercu]

Still trying to put together some useful information. I replaced R45 and U1, but the issue persists. More to come.

[Goatropercu]

Since I seem to be able to trace a sine wave "further" in the low voltage range settings I input a 1.5vac signal into the test lead jacks and moved through the AC converter section with the range set to 2.

Good 1.5v sine wave at U1 input pins 2 and 3. Same applies at the junction of R6 and C7.

At U1 output pin 6 the wave is no longer a smooth sine wave; vertically oblong is the best way to describe it. It measures about 3.5v rms.

I replaced U1, and hope/assume (neither of which i rely on) the part installed was good. Lesson learned on establishing a proper baseline for comparison, again. I will see if replacing C8 has any effect.

[bdunham7]

I replaced U1, and hope/assume (neither of which i rely on) the part installed was good. Lesson learned on establishing a proper baseline for comparison, again. I will see if replacing C8 has any effect.

I have an 8000A with a different problem (ADC?) but AFAIK the AC converter is OK.  Here is what I get at the output of U1 and at C10.  C10 is the large-ish capacitor rightmost and frontmost under the shield and the point to measure is on the right.  The schematic is confusing because the 'C10' designator is on the first half and the actual capacitor is on the second half.

The idea is that the output of U1 has to deal with the diodes in the feedback loop.  The stuff connected to the compensation pin appears to be feedback, but the actual feedback is through those diodes and that is why the output has those offsets at the crossover points.  That's normal.  What you should see is a fairly but not perfectly smooth output at C10 that more or less tracks the AC input.  I used 1.5V as you did, I got what you see.  As I change my AC input from 0 to 2V, that signal at C10 smoothly follows it.  That's what you want--that signal at C10 is what the meter sends to the ADC to be integrated.  Unless that is erratic, your problem isn't in the AC converter.

[Goatropercu]

Again, this is incredibly helpful. Thank you. I am learning quite a bit. My dad actually just loaned me his functioning 8000A for comparison.

With an 8vac input and range set to 20 the waveform at the output U1 (pin 6) is the same on both machines, except mine is jumping up and down on the screen where his is stable. More to come.

This is my first rodeo in this fora, and I am impressed. I expected crickets. Thank you very much for your time.

[bdunham7]

With an 8vac input and range set to 20 the waveform at the output U1 (pin 6) is the same on both machines, except mine is jumping up and down on the screen where his is stable. More to come.

1.5VAC on the 2V range is probably the most useful signal to use because there is no scaling or extra gain involved--it just goes (relatively) straight through.  So next you need to see what is on the ouput at C10.  You mentioned it jumps around even with no input, how about shorted?

[Goatropercu]

Yes it jumps around w inputs shorted.

[Goatropercu]

I am suspecting CL1 may be the culprit. It was mentioned earlier in this thread. At the + side it is much jumpier in comparison to the working meter with an 8v input (w/ a 1.5v input the jitteriness is much more subtle). Per the sm CL1 is a 1mA, fluke p/n 348482, mfg p/n 89536, and looks to be a TO-92 package (i could be wrong there). Would this be a suitable substitute? If not, any recommendations?

https://www.mouser.com/ProductDetail/Semitec/E-102?qs=wgO0AD0o1vuW%252BfX3FIcJ1g%3D%3D

I may get bold and swap in the one from the working meter.

Thank you.

[bdunham7]

What do/did you see at C10?  What is your reasoning for suspecting CL1? 

[Goatropercu]

I suspected CL1 because that is where the scope image starts swinging around. I swapped CL1's and the issue was not resolved...moving on.

At C10 i don't get a waveform on either machine, but just a jumbled up mess of a line that looks like noise. The image is swinging wildly at the C10 pin closest to S1. If you are referring to the "point to the right" being from an arial view of the component side, that would be this pin. I put my voltmeter on this point and it is all over the place (in mV).

Where does the negative lead of C10 go? I see a triangle with what appears to be a "1" inside of it.

Another observation is that with no inputs connected the incandescent lamp (DS5) in the problematic machine fluctuates in brightness like something is pulling down the current perhaps. The lamp in the working device stays well lit (and for the most part brighter) with no fluctuations in intensity.

[bdunham7]

I suspected CL1 because that is where the scope image starts swinging around. I swapped CL1's and the issue was not resolved...moving on.

This is where the voice in your head should tell you "Sir, put the soldering iron down and back away from the instrument!"   

At C10 i don't get a waveform on either machine, but just a jumbled up mess of a line that looks like noise. The image is swinging wildly at the C10 pin closest to S1. If you are referring to the "point to the right" being from an arial view of the component side, that would be this pin. I put my voltmeter on this point and it is all over the place (in mV).

Where does the negative lead of C10 go? I see a triangle with what appears to be a "1" inside of it.

What you should have at C10 is a relatively smooth, stable DC with a bit of ripple if your source is 60Hz.  If that is going crazy, that might be your issue--but if you see the same thing on the other instrument, perhaps there is an issue with your measurements?

The symbol represents the guarded ground or common.  That should tie in with nearby TP4 and that is where the ground lead of your oscilloscope should go.  Is your AC source ground-referenced?  If so, is the ground side connected to the COMMON input on the DMM?  Is your scope DC-coupled?  Once you are sure all that is good, check these points with your scope and a 1.5V_RMS input.

Q1 inputs pins 2 & 3, you should see a sine wave corresponding to the input but offset somewhat.  These should be quite stable and smooth.  The offset is the result of using CL1 and Q1 as the input buffer.  If it is not steady, try to describe accurately in what way it is unsteady.

Q1 output pin 6, you should see the waveform I posted, again quite stable and smooth.  If not, note how jumpy and irregular it is in terms of millivolts of deviance or instability.

R10/C9/CR5 junction, you can get this on the rear end of R10.  What I see on mine is a pretty good sine wave with a lot of offset.  It has a DC mean and an AC RMS value (set your scope to calculate those if you have that feature) that are both about 1.5V, same as the input. 

[Goatropercu]

This is where the voice in your head should tell you "Sir, put the soldering iron down and back away from the instrument!"   

Noted.

What you should have at C10 is a relatively smooth, stable DC with a bit of ripple if your source is 60Hz.  If that is going crazy, that might be your issue--but if you see the same thing on the other instrument, perhaps there is an issue with your measurements?

It is not the same on both instruments. The DC is incredibly unstable on the C10 pin closest to the switch (S1B), TP4 reference (triangle 1), opposite from R10.

The symbol represents the guarded ground or common.  That should tie in with nearby TP4 and that is where the ground lead of your oscilloscope should go.  Is your AC source ground-referenced?  If so, is the ground side connected to the COMMON input on the DMM?  Is your scope DC-coupled?  Once you are sure all that is good, check these points with your scope and a 1.5V_RMS input.

Understood on the TP4 designation/representation. I have had my scope ground and non-DUT multimeter's common connected to the DUT's common input, rather than TP4.

My AC source is a signal generator that I use for audio stuff, best I could come up with. I have the sig gen output ground going into the DUT's common input.

Q1 inputs pins 2 & 3, you should see a sine wave corresponding to the input but offset somewhat.  These should be quite stable and smooth.  The offset is the result of using CL1 and Q1 as the input buffer.  If it is not steady, try to describe accurately in what way it is unsteady.

With a 1.5vac input the sine wave at U1 input pins 2 and 3 indicates marginal up and down movement. The instrument display indicates anywhere from 1.3vac to 1.6vac. However, as the input voltage is increased the variation/deviation increases as well. At an 8vac input the wave on the display erratically moves up and down and the instrument display deviates further from the input.


Q: "Q1 output pin 6, you should see the waveform I posted, again quite stable and smooth.  If not, note how jumpy and irregular it is in terms of millivolts of deviance or instability."
A: Same waveforms on both machines, same as your post. The instrument with the issue jumps up and down from 3.77v to 3.59v. The other is stable around 3.52v.

[Goatropercu]

Clearly I need guidance on quoting partial posts.

I believe the issue is upstream of U1. That is where the instability begins to occur. Specifically at the anode of CR1. More to come. Thank you for your time.

[bdunham7]

It is not the same on both instruments. The DC is incredibly unstable on the C10 pin closest to the switch (S1B), TP4 reference (triangle 1), opposite from R10.

Understood on the TP4 designation/representation. I have had my scope ground and non-DUT multimeter's common connected to the DUT's common input, rather than TP4.

My AC source is a signal generator that I use for audio stuff, best I could come up with. I have the sig gen output ground going into the DUT's common input.

With a 1.5vac input the sine wave at U1 input pins 2 and 3 indicates marginal up and down movement. The instrument display indicates anywhere from 1.3vac to 1.6vac. However, as the input voltage is increased the variation/deviation increases as well. At an 8vac input the wave on the display erratically moves up and down and the instrument display deviates further from the input.

If you want to add your own comments into the middle of quoted text, you need an 'unquote' ( "/quote" in brackets) before your text and then "quote" in brackets after.

Your setup is correct and your sig-gen should be fine.  TP4 and the COMMON input are directly connected.

So the next step is to see how the signal goes through the buffer.  I'll go one step at a time so I can think about it as we go.  I may disappear for an hour or a day, but I'll be back.

At the back of the region under the shield there is a large resistor (R4) and a diode very near TP4 (CR1).  With your scope test both with a 1.5V input.  You should have a clean ~1.5V signal at the solder turret that is also the junction of R4 and CR1 (the left side) and then on the other side of CR1--right next to TP4--you should have the same clean AC signal but offset quite a bit.  If those aren't perfect, compare both sides of R4--they should be nearly identical in all ways, including amplitude.  The left side may be ~1% lower if you are using a 10X probe, 10% if 1X, but 1X is not so good for this.  100X/100M probes really do have uses.

[bdunham7]

This text will be shown as a quote.

This text will not be shown as a quote

The number of quotes and unquotes must match for the whole post

See the attached picture.  And lets not get started about attachments...

[Goatropercu]

At the back of the region under the shield there is a large resistor (R4) and a diode very near TP4 (CR1).  With your scope test both with a 1.5V input.  You should have a clean ~1.5V signal at the solder turret that is also the junction of R4 and CR1 (the left side) and then on the other side of CR1--right next to TP4--you should have the same clean AC signal but offset quite a bit.  If those aren't perfect, compare both sides of R4--they should be nearly identical in all ways, including amplitude.  The left side may be ~1% lower if you are using a 10X probe, 10% if 1X, but 1X is not so good for this.  100X/100M probes really do have uses.

At the junction of R4 and CR1 cathode I have a stable clean 1.5v sine wave. Probing the anode of CR1 produces a clean sine wave BUT it flutters/jumps up and down on a bit on the screen. This jumpiness is not present at this point on the working instrument. Seems like I’m getting some unwanted dc.

[bdunham7]

At the junction of R4 and CR1 cathode I have a stable clean 1.5v sine wave. Probing the anode of CR1 produces a clean sine wave BUT it flutters/jumps up and down on a bit on the screen. This jumpiness is not present at this point on the working instrument. Seems like I’m getting some unwanted dc.

OK, so I think you've narrowed it down to the input buffer 'stack'.  Try probing points 1-5 in the picture I've attached, both with and without an input signal.  If you can find an instability without a signal that will make things that much easier to figure out.

Points 1,2 and 5 should be fairly steady DC voltages regardless of the signal, although you might see a trace of it at higher amplitudes.  Points 3 and 4 will reflect the signal but they should move together.  What scope/probes do you have?  I could suggest some things with multiple channels and math depending on what you are using.  If you do see a noticeable, repeatable instability without a signal, use a DMM to measure across points 3 and 4--this should remain at a fairly constant ~2.2V (the value may be off 10% or so, but it shouldn't vary with time very much).  I think we're homing in on Q1, but it pays to make sure by eliminating everything else first.  R32 is a less likely but not unheard-of possibility. 

I'll test those points on mine if I get some time later so you can compare.

[Dave Wise]

Looks like should be: point 1 @ -15V, point 2 @ -8V, point 3 slightly > 0V, point 2 = point 3 - 2.2V.  Any leakage in CR1 or CR22 will disturb this.  If they are not painted black, any light falling on them will disturb this on the 0.2V and 2V ranges.

[bdunham7]

So here you go:

Points 1,2 and 5 were as expected at -14.7VDC, -7.5VDC and +15.1VDC (regulation on the power supplies is not fantastic on these...)  with no significant variance with or without the input signal.

Points 3 and 4 are shown here, CH1=point 3 and CH2=point 4, the F1 MATH is the difference, which as you can see reflects the constant ~1.0mA across R32.

[Goatropercu]

Apologies in advance if any of the following is difficult to follow. All observations below were made with the range set to "2".

Point 1 (CR3 Anode)
No signal: Slightly unstable scope image. Voltage bounces between -19vdc and -20vdc
W/ 1.5vac signal: Slightly unstable scope image. Voltage bounces between -19vdc to -19.7vdc

Point 2 (Q2 Base)
No signal: Voltage bounces between -12vdc and -13vdc
W/ 1.5vac signal: Voltage bounces between -11.9vdc and -12.7vdc

Point 3 (CR22 Cathode)
No Signal: Voltage bounces between -0.9vdc and -1.3vdc. The functional 8000A measures dead nuts -1.8vdc at this location; no deviation whatsoever.
W/ 1.5vac signal: Clean sine wave that slightly shifts up and down with voltage varying between -.9v and -1.1v

Point 4 (Q1 Source)
No Signal: Voltage bounces between +1vdc and +1.3vdc. The functional 8000A measures dead nuts -0.7vdc at this location; no deviation whatsoever
W/ 1.5vac signal: Clean sine wave that shifts up and down with voltage variation between -1.9 and -2v

Point 5 (Q1 Drain)
No Signal: Voltage bounces between +19vdc and +20vdc.

Voltage across points 3 and 4 is a rock solid 2.27vdc

I am using a Rigol DS1102E (2ch) and the probes that came with it.

I know my B+ and B- are about 4v off from the prescribed 15vdc, and do seem somewhat less stable than I'd expect. However, I'm not convinced that the PS is the culprit.

What's interesting is that with a 1.5v input and the meter range set to "2", the display steadily reads between 1.3-1.7vac. This is not accurate, but close. Now when I flip the range over to 20v, still with a 1.5v input, the meter spits out all sorts of readings in the range of 0.9v to 7v. The drift is compounded when the "20" range is selected. Perhaps this is expected and the drift at lower voltages is just being magnified via the circuitry associated with the 20 and 200 range functions.

Thank you for your time and assistance.

[Goatropercu]

Edited Point 4 above.

Image of points 3 and 4. As stated the waveforms shift up and down slightly.

[bdunham7]

   There's no way that this circuit can work with unstable power supplies.  Any shift in point 2 or point 5 is going to be reflected in the input to U1.  So best be having a look at those PS rails as there appears to be no regulation at all.  It is possible the DC and Ohms functions aren't as sensitive to this.

The reason you see more of an effect on the 20V range is that it is using much more gain in the U1 circuit.  The 2V range appears to be approximately unitary gain, but the 20X range uses 100X gain with the 1000X divider if I've interpreted things correctly.  This is a fairly goofy way to make a meter, but the 8000A is an early, thrifted version that really hasn't stood the test of time TBH. 

Edit: this isn't a battery-equipped model, is it?

[Goatropercu]

Then back to the power supply I go.

Both of the 8000A’s here are connected to mains, no batteries.

I have been using an old 8060A for the past 5 years. My dad has had this one lying around and with the cost of 9v batteries, combined w/ my forgetful nature I finagled this one from him. I’m already on the hunt for a more modern/accurate/better bench meter, but this was my Dad’s and I want it to function as well as I can get it to.

Really appreciate the help.

[Dave Wise]

I suspect CR8 is open.

[bdunham7]

Can you post a photo of the whole board?

[Goatropercu]

Pic

[bdunham7]

I was afraid of that....can you post a photo of the front face as well?  Your PSU issues are not going to be simple.  They use a crude inverter and that is why there are the little fine ripples on your waveform.  The AC circuit can probably work OK with some noise and lack of precision, but stability issues will wreck it, as you are seeing.

Are the batteries any good or are they completly toast?

[bdunham7]

I suspect CR8 is open.

Oh he wishes it were that simple!  He has this mess, not the simple PSU on the line-only version:

[Dave Wise]

Oops sorry.  After I posted, I began to wonder if the OP's instrument is the Option 001 battery version.

The Option 001 version "regulates" only as well as the battery regulates.  Yuck.  If the battery pack - which also acts as main reservoir capacitor - is crapped out, we can't expect +/-15 to be anything reasonable.  I would disconnect the AC power cord and the battery pack, and temporarily power the instrument from a bench supply connected in the battery's place.  Add a big capacitor at the battery terminals, since the inverter draws heavy spikes of current and the external supply may not keep up.

[bdunham7]

Both of the 8000A’s here are connected to mains, no batteries.

I have been using an old 8060A for the past 5 years. My dad has had this one lying around and with the cost of 9v batteries, combined w/ my forgetful nature I finagled this one from him. I’m already on the hunt for a more modern/accurate/better bench meter, but this was my Dad’s and I want it to function as well as I can get it to.

OK, so there's a big difference between a non-battery model and a battery model (which yours is) without batteries.  I don't know how well they work with the batteries not installed but I would suspect not well.  Does your good meter have the same battery holders on the board as yours?  And if so, they are empty, no batteries installed? 

IMO, it is worth the experience of fixing this even if the end product isn't something all that helpful.  But I'll warn you , it can become a lifelong addiction!

[Goatropercu]

Does your good meter have the same battery holders on the board as yours?  And if so, they are empty, no batteries installed?

The working meter also has the battery option w/ no batteries installed.

IMO, it is worth the experience of fixing this even if the end product isn't something all that helpful.  But I'll warn you , it can become a lifelong addiction!

That is what I like to hear and I really appreciate the positive attitude. If you can't tell I have no formal electronics training. I have been repairing stereo equipment for the last few years, but that's about the extent of it. I know this meter isn't the holy grail fluke by any stretch of the imagination, but it belonged to my dad. I've already learned quite a bit up to this point, and have had a few refresher courses in things I already thought I knew as well. If you have the patience I am all ears.

[Goatropercu]

Picture of the working meter attached. The only difference in terms of battery is that this meter has the battery terminal connections still in place. It’s interesting to me to see the different vr’s, resistors, etc that have been replaced assumingely when they were sent in for cal.

[Dave Wise]

Looking at the power supply schematic, I don't expect an 8000A Option 001 to do anything reasonable without a good battery.  There's no regulation or filtering ahead of the inverter except the battery itself.  Disconnect mains power and hook up a bench power supply in place of the battery.  Any troubleshooting done without this (or a good battery) is meaningless.  "Ain't power supply happy, ain't nobody happy."

There are instruments that run on AC without the battery (e.g. Fluke 8100B), but the 8000A Option 001 is not one of them.  It must have a good battery to work.

[bdunham7]

Picture of the working meter attached. The only difference in terms of battery is that this meter has the battery terminal connections still in place.

Well that's interesting.  Apparently at least that meter works sufficiently well without batteries in place.  Can you post scope pictures or screenshots of the power supply rails from both the working and the unstable unit?  It would be worth it to confirm that there is some signficant difference and perhaps what might be causing it.  Are you averse to 'improving' the original design a bit?

[joeqsmith]

Looks like someone told you to start swapping caps and you replaced tants for alum elec.   

8000A was my first DMM.  Like your dads having some sentimental value, that's about the only reason I still have mine.   

Those custom ICs are not well protected and those stupid trim resistors  that came with the kits being plugged into the IC sockets.  Not that those mechanical joints would ever get flaky.  Then there was the cost for the IC repair kits.   

[Dave Wise]

Add a reservoir cap and a three-terminal regulator that outputs the battery voltage.  Of course then you can't put a battery in it.

[m k]

I know my B+ and B- are about 4v off from the prescribed 15vdc, and do seem somewhat less stable than I'd expect. However, I'm not convinced that the PS is the culprit.

Better believe what you see.

Your #21 is already showing that the issue starts appearing at the R32 level of the direction of input.
Then points 4 and 5 here are showing that the issue is present both sides of Q1.

Two possibilities, either something is overloading +15V or it's unstable.
The other machine seems to be stable, so my guess is the latter.

[Goatropercu]

Can you post scope pictures or screenshots of the power supply rails from both the working and the unstable unit?  It would be worth it to confirm that there is some signficant difference and perhaps what might be causing it.  Are you averse to 'improving' the original design a bit?

Images attached with leads at C17(+) and C18(-) of both instruments. Please let me know if you need me to adjust anything and try again. The voltage isn't 100% rock solid at these points on the working meter, but clearly it is closer to spec.

I am not opposed to modifications in the effort to get it functioning properly. I have no intention of ever powering this meter via batteries.

Disconnect mains power and hook up a bench power supply in place of the battery.

I do not have a bench power supply at my disposal (yet), but I do have a tub of wall warts of various voltages.

I have a stable 8vdc at the cathodes of CR13 and CR14. I am getting between 5.3vdc and 5.5vdc at R22.

[bdunham7]

Images attached with leads at C17(+) and C18(-) of both instruments. Please let me know if you need me to adjust anything and try again...

...I have no intention of ever powering this meter via batteries.

Could you change the timebase to 10ms/div (100ms total across the screen to correspond with the U3 measurement period) and try again on both meters? 

As for batteries, one of my suggestions was going to be to put 4 alkaline D cells in it and see what it does!

However, a bit further down the road you might put a bit of circuitry in place of the batteries to regulate your 5 V supply.   The +/-15V supplies can probably be fixed well enough by simply crowbarring them with 15V 5W zeners across C17 and C18.  But I'm interested to see what the differences are between the good and bad meters first--you don't want to re-engineer the whole PSU and still have an issue.

[Goatropercu]

Updated images in previous post. Will rustle up some D’s.

Edit: It states to use nicad batteries only. It charges them when plugged in. I don’t have any rechargeable D’s here and don’t plan to. I’ll use alkaline for testing porpoises and just not plug the thing in.

Will have to rig up some sort of battery terminal doo hickies. All for robbing parts, but i’m leaving those in the working meter intact. Might just see if I can rig up a voltage supply in place of batteries.

[bdunham7]

Updated images in previous post. Will rustle up some D’s.


Edit: It states to use nicad batteries only. It charges them when plugged in. I don’t have any rechargeable D’s here and don’t plan to. I’ll use alkaline for testing porpoises and just not plug the thing in.

Will have to rig up some sort of battery terminal doo hickies. All for robbing parts, but i’m leaving those in the working meter intact. Might just see if I can rig up a voltage supply in place of batteries.

ms = milliseconds, 1000X slower than what you have posted which is 10µs (microseconds)!  The sampling period of U3 is 100ms (1/10 of a second) so that's the timeframe we need.

Can you look at the 5V supplies as well?

[Goatropercu]

Not sure what I was thinking. Long day yesterday.

[Goatropercu]

Images updated again, and attached here as well.

I have 8vdc at the cathode of CR14 and 5.3-5.5vdc at the positive lead of C25.

Thank you.

[Goatropercu]

What’s the purpose of the incandescent lamp DS5?

[m k]

Charge indicator?

Seems that the level is more critical.
Try 0.1s/div.

My 1st bet is R21.

[bdunham7]

What’s the purpose of the incandescent lamp DS5?

It limits battery charging current if they are discharged.  High-tech stuff, hey?

I'm not sure what to advise you to do next.  Those PSU rails look pretty similar in terms of ripple at both 60Hz and the ~4kHz of the inverter.  Perhaps we shouldn't ignore the very obvious fact that the voltages on the bad meter are much higher, although I don't know what that would cause in terms of the instability.  Can you do the same scope capture using points 1 and 5 from the diagram I posted earlier?  And then do a second capture zooming all the way out to 100ms/div (1 second total capture)?  Earlier you said those points were unstable, I'd like to see what that looks like.  If there's any instability at points 1,2 or 5 that isn't reflected in the power rails, then there's an issue elsewhere.

I can't say that this is a PSU problem at this point.  If I had it I would connect a bench PSU to it for further testing, but since that's not currently an option you'll have to keep poking.

[Goatropercu]

A bench power supply is also now on my must get list.

The incandescent DS5 bulb is faintly pulsing/flickering like something is drawing down the current. Just an observation.

[Goatropercu]

Can you do the same scope capture using points 1 and 5 from the diagram I posted earlier?  And then do a second capture zooming all the way out to 100ms/div (1 second total capture)?  Earlier you said those points were unstable, I'd like to see what that looks like.  If there's any instability at points 1,2 or 5 that isn't reflected in the power rails, then there's an issue elsewhere.

Images attached. The + and - rail seem to behave the same from the CR15-CR18 rectifier onward. They both bounce around between 19 and 20vdc (+ and - of course).

Voltages at Q22 and Q23 below. These are not too far off from the working meter, other than the base of both measure -.2vdc.

Q22:
E: 8mVdc
B: -.4vdc
C: 5.3vdc (+/- 0.2v)

Q23:
E: 8mVdc
B: -.4vdc
C: 5.3vdc (+/- 0.2v)

Thank you.

[bdunham7]

Alright, I think we're onto something here, although it still isn't clear at those scales.

If you can make this work, set your scope scale to 2V/div and position your CH1 off the bottom of the screen (-14V if you can) and the CH2 off the top (+14V) so that the traces appear on the screen.  This should work if you scope has sufficient offset capability.  If not use 5V/div and just put CH1 at the second line from the bottom and CH2 second from the bottom.  Now set the scope for 500ms/div and post the results.  Do the same for the working meter and post those results too.  If I'm seeing what I think, the difference should be obvious.  Unfortunately the fix won't be...yet!

The incandescent DS5 bulb is faintly pulsing/flickering like something is drawing down the current. Just an observation.

Yeah, it may turn out that as crude as it is, missing battery and all, the PSU may still be good enough for the meter to work more or less properly.  And the fluctuations may be due to some other part of the meter drawing it down erratically.

Edit:  Just out of curiosity, what happens if you take the light bulb out entirely?

[m k]

The incandescent DS5 bulb is faintly pulsing/flickering like something is drawing down the current. Just an observation.

Dang, I didn't check the manual and thought the switch is either or.
Surely the lamp is then the major part compared to R21, maybe it's too bright.

[Goatropercu]

…set your scope scale to 2V/div and position your CH1 off the bottom of the screen (-14V if you can) and the CH2 off the top (+14V) so that the traces appear on the screen.  This should work if you scope has sufficient offset capability….Now set the scope for 500ms/div and post the results.  Do the same for the working meter and post those results too.

Images attached. Not exactly sure what the others indicated and/or if these are useful in making a determination. That being said if I need to change something up let me know.

Just out of curiosity, what happens if you take the light bulb out entirely?

With the incandescent bulb DS5 removed the first thing I noticed is that the display is significantly dimmer. Rail voltages have slouched to positive and negative 11vdc.

[m k]

So to operate correctly the machine needs a correct lamp, it seems.

I had a car once that had an ashtray lamp as a headlight level motor fuse.

[bdunham7]

Images attached. Not exactly sure what the others indicated and/or if these are useful in making a determination. That being said if I need to change something up let me know.

OK, even though there is quite a ripple envelope, you can clearly see the whole envelope going up and down over the timeframe of ~100-200ms.  In addition, the two power rails are moving with each other in magnitude, or opposite each other in absolute voltage if you like.  That indicates that the two likely causes are either the 5V input to the inverter is varying or there is some variable but matching load on the power rails.  Since the AC input buffer stack can't be it (the CL1 keeps the current to 1mA), I'm thinking it is probably the 5V.  One possible cause is the digital display which almost certainly draws some varying current.  The light bulb flickering seems to indicate the same thing--varying current draw on the 5V supply.  None of this would be surprising without batteries, the true mystery for me is how the other meter manages to work properly!  Are you sure that no modification of any kind has been done to the good meter or that there aren't some little batteries hiding in there? 

A couple of random things to look at:  What happens to the power supply voltages, the light bulb flickering, etc, if you put the meter in the DCV range?  Is either U3 or U4 getting excessively warm?  Are all of the display segments working?

It also has occurred to me that the milliamp-second function is a signficant revision of the meter overall, including the AC input circuitry (which seems fine at this point).  Without a diagram of that board and how it connects to the rest of the meter, it is pretty hard to know if that might be involved.  However, right now I'd try to force the PSU to be stable and quiet even if it doesn't want to and even if the other meter functions with fairly noisy power rails.

[Goatropercu]

At this point I like the crowbar via zeners idea.

Now that you mention it the connector at the mA/sec board is wired differently on both devices. EDIT: and now that I look even more closely the big 5uf/50v white cap is not present on the working device. The additional fuse tacked on to the one above the shield has stood out, but i just assumed this makes no difference in what I am experiencing since I have continuity there.

[Goatropercu]

For some reason I couldn’t upload this pic in the previous post.

Can’t thank yall enough for your time. 95% of what I measure is dc and it’s ok if the acv function on this meter just wasn’t meant to be due to design, non documented modification, and/or remote troubleshooting w/ the likes of myself. I’m not throwing in the towel, just don’t want to be a burden. This has spurred me to really start looking into purchasing a dedicated bench meter and power supply.

All of the display segments are working.

I will compare readings w/ the dcv function selected and check U3 and U4 temps and report back.

Just out of curiosity, how would I go about connecting an external 5vdc power supply to serve as the unregulated ps? Would I just disconnect the cathodes of CR13 and CR14 and feed 5vdc to those two points? Ground?

Thank you.

[m k]

No, unreg. is not used, so not needed.
Connection is of course correct but voltage you would need is that earlier measured 8V, but then you also need a correct lamp, diodes can also stay.

For 5V you use the other side of the lamp.
Treat it as a normal battery operation.

[Goatropercu]

Ah, thank you.

[m k]

The real power route is supposed to go through the lamp, problem is that its resistance comes along, so no very easy way out.
Now R22/R24 is 330/82, so around 10mA and a quarter of 5V to Q22 and Q23 bases.
R24 voltage is what it is, so that the lower part of T2 can switch those transistors, and R22 is just its feeder.
So changing R22, or R24 to a potentiometer is not going to do anything to the actual feeder voltage.
Maybe the easiest permanent way is to remove R21 and replace the lamp with adjustable mini regulator board from China and add a LED to its output side, if needed.
Since C25 is the only reservoir and then missing you may need a new one, how big depends how low input your regulator accepts and how steady output is needed.

For a curiosity you can also put the regulator before the lamp and make the whole thing adjustable.
Lift CR13 and CR14 and connect them to regulator's input as a source, for output you can then use those new freed holes.
Then adjust the output so that the other side of the lamp is 5V, after that you can compare how the output compares to that earlier 8V.

For a curiosity you can also just lift R21 and see what it does to +/-15V.
The problem with a lamp as a voltage divider is that its resistance is temperature dependant.
So lifting R21 will change the resistance over the lamp as high as it can be and so drop the 8V as much as it can.

You can also find a low value potentiometer and try that in place of the lamp, but few Watts may not be enough, so better use a heavier one.
Then you can also test how the machine reacts when 5V changes.

BTW,
I'm still wondering how unregulated is changing to not unregulated over a lamp.

[Goatropercu]

I measured much of the same with the meter set to the DCV function and received the same results.

The mA/sec boards are different revisions; one is Rev C and the other is Rev D.

On a side note, early on in the discussion I noticed that the display on the "functioning" meter is not flashing in an over limit situation when the "2" and "20" ranges are selected. I didn't want to derail my own thread, but figure it worth mentioning at this point.

I may continue to fiddle around with it if anyone has any further suggestions (m k and I posted at the same time) or after I up my bench game to include a power supply

I suppose adding the zeners won't help if the pulsing stems from upstream.

Thank you for your time.

[bdunham7]

The problem with a lamp as a voltage divider is that its resistance is temperature dependant.

Maybe he should swap the lamps between the two meters!

[bdunham7]

I suppose adding the zeners won't help if the pulsing stems from upstream.

It probably would help anyway because you are well over the nominal voltages for the power rails.  The current through the zeners would be varying, but the voltage across them would vary less.  The dynamic impedance of a 15V 5W zener under those circumstances would be just a few ohms, so probably pretty decent control over a moderate range.

https://www.mouser.com/datasheet/2/68/CSEMS05385_1-2539170.pdf

But this all sounds like a great excuse to acquire a bench PSU.

[m k]

The problem with a lamp as a voltage divider is that its resistance is temperature dependant.

Maybe he should swap the lamps between the two meters!

Yes, and even more if they seem to be equals.

E,
'' <-> 's'

[Goatropercu]

I will swap lamps and see what happens.

If I were to add 15V 5W zeners across C17 and C18 would it go like this:

C17+ to zener 1 cathode
C17- to zener 1 anode

C18+ to zener 2 anode
C18- to zener 2 cathode

Don’t grill me if that is incorrect.

[bdunham7]

C17+ to zener 1 cathode
C17- to zener 1 anode

C18+ to zener 2 anode
C18- to zener 2 cathode

Not grilling, but why would you connect the zeners differently on the two rails?  For a diode, the pointy end (on a diagram) is the cathode and the "positrons" (current) flow in the direction of the arrow.  A zener has a regular diode function as well, but you want it not to conduct in that manner so you'd always put the cathode to the more positive side.  Then when the zener voltage is reached, the current (conventional current being the opposite of electron flow, thus the positron joke) flows against the arrow.  So cathodes go to the positive sides of both capacitors.

[Goatropercu]

Thank you for being easy in your explanation. I’m learning, everyday.

I don’t have any 15v 5w zeners in my stash, so will have to include them in my next order.

Thank you both for your time and patience.