Author Topic: Fluke 8050a True RMS U32  (Read 13049 times)

0 Members and 1 Guest are viewing this topic.

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Fluke 8050a True RMS U32
« on: March 06, 2015, 08:36:02 pm »
I'm trying to repair a Fluke 8050a.

I've already fixed a problem with the power supply, a bad selector switch (the Voltage selector 4PDT switch, which fortunately only used 3 of the 4 poles, allowing me to reconfigure it to use the unused 4th pole), some intermittent segments on the LCD display and a bad solder joint.

Everything is now working except for AC functions. I've tracked a good AC signal to the input of the true rms board -  the manual labels that board as a single component U32. It's a ceramic thick film board mounted vertically. It has 4 surface mount ICs and a few SMD diodes and capacitors.  It's marked on the back "Fluke 53 - 3045" and has approximately 30 pins soldered to the main PCB. There is no circuit diagram for that TRMS board, but its operation is described as comprising an absolute value circuit, a 2x log converter circuit, an antilog converter circuit and a log converter circuit.

It's difficult to track the circuit on the thick film board, but the input to that board is close to an LF351N single op amp, which I suspect is the absolute value circuit. A second IC is a CA3140e, which I suspect is the 2x log converter. A third IC is marked 473777. I've seen another post in which that IC failed. That post identified it as a dual op amp. With that in mind I suspect that op amp pair performs the log and antilog functions. The last IC is marked CA3046. It includes 5 transistors. I believe it works with the other components to perform the mathematical operations.

An AC input signal gets through the LF351N and arrives at an SMD diode pair marked A7-. The output of the LF 351 n  is a distorted sine wave from the true sine wave input. It has very vertical rise times and fall times with smooth almost square wave tops. It is both negative and positive and it's certainly not an absolute value. I did see a circuit from the 1980s using this chip in a precision absolute value full wave rectifier. There is no AC signal at any of the other chips or at the output. I suspect the LF 351N (JFET) inputs is bad but it is very difficult to get access to.

any suggestions for obtaining a replacement TRMS bored or finding a replacement LF 351 n? I'm going to try inserting a full-wave rectified ac signal after the LM 351 n and see if I get any AC output indication. At the present time I get only 0. Thanks for any suggestions.
« Last Edit: March 09, 2015, 07:10:14 pm by tango17 »
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #1 on: March 07, 2015, 02:50:47 pm »
If there is anyone here that has a Fluke 8050a and a scope, it would be very helpful if they could check the signal on the A7- diode of the U32 board (with a 60 Hz sine wave input on the meter)and tell me what they see. That diode has easily acessible test points nearby and is connected to the output of the op amp in the  LF351N. If a full wave rectified output appears there, I would know that part is bad on mine. I could supply a photo of the board and TP location?
 

Offline wiss

  • Frequent Contributor
  • **
  • Posts: 485
  • Country: ch
Re: Fluke 8050a True RMS U32
« Reply #2 on: March 07, 2015, 04:14:04 pm »
I have 2, can check tomorrow.
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #3 on: March 07, 2015, 08:58:36 pm »
Thank you for the offer. I don't want to put you to any effort until I'm certain I need help. I've made some progress.  The LF351N at the input is connected in the classical way to the dual diodes for a precision absolute value circuit. The output from that op amp is close to what it should be for the absolute value circuit. There should be another op amp performing the second half of the job, and I suspect it's the CA3140E. If it's connected as required for that circuit,then it would be the culprit, as it is not producing the full wave precision rectification I should be seeing on pin6 of the CA3140E.

I'm temporarily away from the bench, but as soon as I get back, I will check the circuit around those two opamps and confirm that it is designed as a precision absolute value circuit. Then I may be able to verify that the CA 3104 E is bad or that there is a bad component between the two pants in the absolute value circuit.

I really appreciate the offer of the help.I just do not want you to spend any time until I am 100% certain I need the help and, right now, I am still making progress.
« Last Edit: March 07, 2015, 09:15:51 pm by tango17 »
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #4 on: March 07, 2015, 09:13:19 pm »
A photo of the TRMS board. The chip behind the orange cap is the CA3140E. The chip to its  left behind the transformer is the LF351N at the input.
« Last Edit: March 08, 2015, 06:25:15 pm by tango17 »
 

Offline wiss

  • Frequent Contributor
  • **
  • Posts: 485
  • Country: ch
Re: Fluke 8050a True RMS U32
« Reply #5 on: March 08, 2015, 05:54:50 pm »
On the two pads towards the edge of the board there are nice looking halfwave rectified voltages, one positive and one negative, on the pad facing the LF351 there are a sine plus a diode-drop in the direction of the sine.

"if (sin(x) > 0) {
  reutrn sin(x) + 0.6
} else {
  return sin(x) - 0.6
}"

Or, more like 0.4 V, and it is not symmetrical.
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #6 on: March 08, 2015, 06:43:28 pm »
Hmmm. I've labeled the four pads symetrically positioned around the A7- diode pair in my photo above.  At T2 I see a weak half wave negative going rectified signal. It's inverted from the positive half. At T3 I see strong positive going half wave (inverted from the input negative half cycle).  If I look closely at T3, Ican tell that between the strong positive  peaks, there are weak positive peaks that match the negative going peaks on T2. The strong peaks are roughly four times the amplitude of the weak peaks, so on some scales and inputs it looks like half wave rectification.

On T1 and T4, which are connected, I see a symmetrical near square wave.  The tops and bottoms of the signal bulge out, up at the top, down below.

Do you see full wave rectification anywhere? I sort of expect it (absolute value) at a second op amp, and the logical suspect would be the output (pin 6) of CA3140E.

I'm seeing that pseudo full wave signal (strong and weak peaks) at the inverting input (pin 2) and sharp narrow pulses at the output (pin 6) of CA3140E.
« Last Edit: March 08, 2015, 06:58:57 pm by tango17 »
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #7 on: March 08, 2015, 08:46:48 pm »
"if (sin(x) > 0) {
  reutrn sin(x) + 0.6
} else {
  return sin(x) - 0.6
}"

Or, more like 0.4 V, and it is not symmetrical.
That sounds like what I see at T1, T4, and the other pins seem close to the positive and negative half wave rectification I see, except for scale.  I was beginning to wonder if I should be looking at the 473777 chip. There is another report of it failing, and I'm not seeing any output from it. It's grounded at the output (pin1) on the op amp connected to the output of the U32 TRMS, and both the inputs on that output op amp (pins 2, 3) are also always grounded.
 

Offline ModemHead

  • Frequent Contributor
  • **
  • Posts: 730
  • Country: us
  • No user-serviceable parts inside.
    • Mr. ModemHead
Re: Fluke 8050a True RMS U32
« Reply #8 on: March 08, 2015, 09:17:27 pm »
Here's what I see...

Meter input is 1.00Vrms sine at 200Hz, shown as bottom trace (2V/div) on all shots.

Top traces are T1, T2, T3, T4 (0.5V/div)

Sweep at 1ms/div.
 

Offline wiss

  • Frequent Contributor
  • **
  • Posts: 485
  • Country: ch
Re: Fluke 8050a True RMS U32
« Reply #9 on: March 08, 2015, 09:20:46 pm »
That look simular to what I saw on the diode.
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #10 on: March 09, 2015, 04:34:43 pm »
Wow! Thanks, Modemhead for those pics.  You must have done that before, as you posted everything I needed to interpret them, the baselines, the settings, the input signal, etc.  I'd always wanted a Fluke, and while reading on the web I came across your posts about converting an 8050a to LEDs.  It was that info that caused me to buy the (for parts) 8050a I found on EBay at a price I could afford.

So back to the repair.  The 8050a manual says the input of the U32 board goes to an absolute value circuit, and studying those circuits from the 1980s, many use a pair of diodes (A7- which my notes label "D1" and "D2" for convenience) between a pair of op amps (almost certainly LF351N, "U1") and possibly CA3150E ("U2") with multiple precision resistors- often laser trimmed (the thick film U32 board seems perfect for this).

I set my scope to match yours, except I used 60 Hz input and 5ms/div.  The same baselines, the same 1V RMS input, the same .5V/div on the upper trace and 1V/div on the lower trace .  My T1, T2 and T4 are the same as yours, so I won't post them, but my T3 differs. The signal between the positive going half wave is not zero, it's got some weaker positive  signal there.

I then traced to the CA3150E ("U2") and the inverting input pin 2 shows something similar - close to full wave rectification, but not symmetrical - see below with 0.2V/div on the upper trace.  The output (pin 6) shows pulses at the zero crossings - see below with 2V/div on the upper trace, still DC coupled.  (I've got some better traces AC coupled and showing the pulses more clearly, but you can see the basic signal below - alternating size positive going pulses from a negative DC baseline.) The offset null (pin 1) of that chip is near -5V and comes from the trimmer pot on the U32 board.  The NI input (pin 3) is at ground.

Any ideas?
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #11 on: March 09, 2015, 05:01:49 pm »
Just thinking out loud here:  The inverting input pin P2 of U2 looks like (T3-T2). Note that the peaks are highest when the input signal is going positive, right where your T3 is zero and my T3 has some positive signal.  I suspect the full wave  (absolute value) rectified signal will be on your P2.  If so, then U2 (CA3140E) is not part of the absolute value circuit, since its input is already absolute value (assuming the T3 signal is working and is zero when T2 is negative going). The output P6 of the CA3140E is what I'm not sure about.  Is it doing zero detection?  I can't think why. Is it just an inverting amplifier and the chip is bad?  And what is causing the nonzero portion of the T3 signal on mine?  Hmmmm?
 

Offline wiss

  • Frequent Contributor
  • **
  • Posts: 485
  • Country: ch
Re: Fluke 8050a True RMS U32
« Reply #12 on: March 09, 2015, 06:15:23 pm »
Leaky diode?
3140 is not good for much, except low input bias. I would not expect it to be part of any low level AC signal processing. High noise and high offset.
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #13 on: March 09, 2015, 06:39:58 pm »
I thought about the "leaky diode" scenario also.  However, that can't be my main problem. I should be getting some sort of signal out of U32, and I am not.  Using the flow diagram, I suspect that the 3140 is the 2x log converter and it's bad. The 473777 chip is a dual op amp and I suspect it performs the log and antilog functions. If the 3140 is the 2x log converter and if its input is the absolute value from the 351/diodes, then the 3140 is bad. I don't know why I'm getting the leaky diode problem, but once I've got output from U32, I can go back and try to figure out why the input to the 3140 is not exactly the absolute value.  Once someone confirms for me that the output pin 6 of the 3140 is roughly 2x log(absolute value ( input)), then I'll replace that chip, since I know that's not what I'm getting.
 

Offline wiss

  • Frequent Contributor
  • **
  • Posts: 485
  • Country: ch
Re: Fluke 8050a True RMS U32
« Reply #14 on: March 09, 2015, 07:04:25 pm »
Pin 6 of the 3140:
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #15 on: March 09, 2015, 07:14:51 pm »
Yes! That's what the log of the absolute value of a sine wave looks like. I'll order a CA3140E, if I can find one. I'll grab some diodes, too, if I dont have any with similar specs lying around.  I wonder if I can do this without desoldering the 29 pins of U32?
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #16 on: March 09, 2015, 07:40:32 pm »
Before I do,  however,  here is a typical log circuit using a BJT. Its in the feedback loop. If it was blown, I might be seeing something like I see now where the op amp goes to the negative rail. The CA3046 chip has 5 BJTs, and the circuit has 3 more log or antilog functions to perform. They probably use those transistors.  Some more checking is warranted.

Thanks, wiss and Modemhead.
 

Offline wiss

  • Frequent Contributor
  • **
  • Posts: 485
  • Country: ch
Re: Fluke 8050a True RMS U32
« Reply #17 on: March 09, 2015, 08:05:45 pm »
In a normal one quadrant square circuit you'd use a log-anti-log circuit, the I0 tempco will be cancelled out by other junctions and the Vbe tempco cancelled automatically in the log-anti-log conversion, well, beats me how they do it here :)
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #18 on: March 09, 2015, 08:46:29 pm »
beats me how they do it here :)
It's really tricky to work on this. There's no circuit diagram in the manual, like there is for the rrst of the 8050a. There's no visible resistors on the thick film substrate, and there aren't any visible traces to see the interconnections.  I have to buzz out every possible interconnect, and if there's a resistor, I get no buzz signal.
 

Offline ModemHead

  • Frequent Contributor
  • **
  • Posts: 730
  • Country: us
  • No user-serviceable parts inside.
    • Mr. ModemHead
Re: Fluke 8050a True RMS U32
« Reply #19 on: March 09, 2015, 08:48:10 pm »
Just thinking out loud here:  The inverting input pin P2 of U2 looks like (T3-T2). Note that the peaks are highest when the input signal is going positive, right where your T3 is zero and my T3 has some positive signal.  I suspect the full wave  (absolute value) rectified signal will be on your P2.

I could not find a signal on the inverting or non-inverting inputs of the CA3140E (U2P2, P3).  Both are basically 0Vdc, plus maybe 20mV of noise.  (The meter is really working, the display shows "1.0037".)  U2 output, and its strobe pin 8, are shown here with the same scope setup, except the input reference trace moved to the top.

I really don't have a clue how this circuit works.  :-\  It sounds like you're getting it narrowed down though.  :-+  And I don't mind getting a little practice for making "old-style" scope shots.
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #20 on: March 09, 2015, 09:17:54 pm »
Not getting a signal on pin2 (inverting input) is very confusing to me. Wiss didn't measure that pin,  but I was sure it was going to be the absolute value full wave rectified. My input is close to that, and it makes sense that the 3140 is doing the 2x log function of the flow chart, working with one of the 5 BJTs in the 3046. At least your output is the same as his.
(Sorry I'm not faster on my testing. I'm not able to sit or stand for longer than a few minutes at a time.)
 

Offline wiss

  • Frequent Contributor
  • **
  • Posts: 485
  • Country: ch
Re: Fluke 8050a True RMS U32
« Reply #21 on: March 09, 2015, 09:19:00 pm »
If +in and -in both are zero, then it has to be an inverting amp, could it be a summing op? Other than that I would say output buffer, but then it would not output a waveform... ?
 

Offline ModemHead

  • Frequent Contributor
  • **
  • Posts: 730
  • Country: us
  • No user-serviceable parts inside.
    • Mr. ModemHead
Re: Fluke 8050a True RMS U32
« Reply #22 on: March 09, 2015, 09:37:26 pm »
Not getting a signal on pin2 (inverting input) is very confusing to me.
Seemed odd to me, so I double-, triple, and quadruple-checked both in+ and in- pins.  Nothing but a little fuzz.  I thought maybe I had broken something, but the meter is definitely working.  Also I have the -01 battery model, so it has a bigger power transformer and a battery pack that obscures the module, probing can be a bit difficult.
 

Offline tango17

  • Regular Contributor
  • *
  • Posts: 82
Re: Fluke 8050a True RMS U32
« Reply #23 on: March 10, 2015, 02:49:17 pm »
As far as I can tell, my AC signal is having trouble at the 3140. The signal at the input should be close to ground and the signal at the output should look like the logarithm of the absolute value of the sine wave. instead I see a signal at the input that looks like the absolute value of the sine wave with a leaky diode offset and at the output I see a signal that is usually at the negative rail but occasionally spikes towards ground.

One possibility is that the 3140 is bad. However,if it was bad I would expect the output to be either at ground or at one of the rails. I don't see that.Instead I am getting something that looks more like a comparator, where the signal goes to one rail except when the input goes to zero at which point I get a spike towards ground.

 So my best guess is that there is a problem in the feedback loop around the 3140, and I think there is a transistor from the 3046 involved in that feedback loop. I'm going to pull the 3046 transistor array and test it. It seems likely to me that the transistors are used in a 2x log circuit in combination with the 3140 and that the transistor feedback loop, which should produce a log signal at the output and a virtual ground at the input is just not working correctly. The transistor array may also be involved in the leaky diode problem I'm seeing.

After pulling the chip off, I will test it to at least see if the pn junctions on the transistors are working correctly. in-circuit testing didn't tell me much.

 Does anyone think I need to use special silver containing solder for connections on the thin film ceramic circuit board? Or will regular 60 40 work?
 

Offline krivx

  • Frequent Contributor
  • **
  • Posts: 763
  • Country: ie
Re: Fluke 8050a True RMS U32
« Reply #24 on: March 10, 2015, 03:30:33 pm »
Is the CA3140E hard to source? I think I have some old RCA parts in DIP8.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf