Fluke 332D repair..still sorting issues

[valley001]

Dr. Frank, thank you for your input.

I have observed noise with the dials set to 0.000000 in the 10v range.  The device will behave for a time moving 3-4uV and then spike 100uV or more.  It also settles to different readings at times, sometimes displaying around -88uV and sometimes -104uV or so.  It is not consistent.  I have not observed the other ranges as much, but I have switched through them with the dials at 0000000 and see similar behavior.  If I set the dials to 1.000000 on the 10v range I will see similar behavior as above. 

This unit has the solid state chopper board installed.

This unit was missing Q4 and Q8 on the differential amp board, I replaced them with the following modern components (Q6 replaced along with Q4):

Q4 and 6:  KSC1008YBU

Q8: KSD1616AYTA

I have also replaced many caps, some were obviously bad (externally swollen) and some tested bad. 

I plan to probe voltages outlined on the schmatic of the chopper board next and possibly the differential amp board.

Hello,

I made some stability characterization, also on the reference circuit a few years back, on a similar unit:
https://www.eevblog.com/forum/testgear/fluke-332baf-in-the-slaughterhouse/msg399283/#msg399283

These old calibrators are not that stable, and quite noisy.

Therefore, you first have to make clear for yourself, on what scale you see these variations of 10..100µV, i.e. in which range, and at which output voltage.
If you see 10µV variation on 10V output, that's 1ppm, and that might be ok, 100µV would be too much, but on 100V, it would be ok.
Your DMM might also create this noise.
Please compare that to my measurements.

The contacts usually do not create such noise.

Please measure the 15V master reference voltage, accessible on top of the shield.
What is the relative noise / stability?
Seems to me, that it is rock stable. Therefore, the whole reference is ok, including the oven circuit, which is fully separate from the reference circuit.

The oven consists of a precise and stable 24V  supply by the 723, which to me seems fully functional as intended, also these 47V raw supply.
I can't find the hint of a slow rise of the 47V, these should also be available instantly.
Oven set point of about 72°C is achieved by this KLIXXON heater element itself, which has a resistance extremum at the oven set point.
Very crude technique, but working.

The reference amplifier itself, that is the zener diode combined with the transistor, 'outputs' the printed reference voltage at the base of the transistor, which you can measure at the wiper of the trimmer.

The arrow labeld 6.6V should indicate there, instead of the collector of this transistor, which indeed has these 8.6V.
That's the fault in the schematic.

Everything seems to be fine there, also.

If you see excessive noise, the root cause seems to be elsewhere, maybe in the chopper amplifier.. please indicate, which version you have in your instrument, the mechanical chopper, or the MOSFET chopper.

But again, please analyse the relative stability first.

Frank

[wictor]

One reason for unstable behaviour can be bad connections between boards.

Wictor

[Dr. Frank]

If you observe instability at 000000 setting, this is very probably not caused by the reference.
Static offsets might be caused by bad connections/contact between the PCBs, but these are gold plated, or in the several intermediate switches.
Sporadic spikes  probably have different root cause, like oscillation or bad components.
Maybe you take an oscillogram of these spikes at the output jacks, how they look like.
Another source of joy might be the pre regulator board, which might be mis-firing, and then disturb the feedback regulation. Check the waveform inside that circuit, if you see frequent, regular pulses, or erratic/ sporadic bursts.
The filter electrolytic caps after the rectifier bridge should also be under observation.
The preregulator should provide an estimate of 35V overhead supply compared to the chosen output voltage at these camps. Also check the waveform at the unijunction transistor circuit.
Always pay attention with ground of circuits and oscilloscope ground.. 'how not to blow up your scope'

Pay attention there, high voltages may be present.

Frank

[valley001]

Here are waveforms from Q1 on the pre regulator board.  The scope is un grounded per the manual.  Scope "common" on emitter and "input" on base. 

The first is with the device on stanby, the seconf on operate. 

These do not look the same as the manual, what may this indicate?

[Dr. Frank]

Doesn't look bad.. I don't know what output voltage you have selected, but the Duty Cycle depends on that. small voltage => very small D.C. .
D.C. should go from about 100% at 1kV down to maybe 1..2% at near zero out.

Maybe you get a picture of the output spikes, if you output 10V and chose AC coupling in a small mV range. Osci ground should be tied to + output jack, as this is the 'real' internal ground of the 335D.

Frank

[valley001]

On the above wave forms the output was set to 50.00000.  The manual is contradictory, it says to put range to 1000 but output to 50.0000 (1000 range and 5000000 dials would net 500v, 500.0000 no?).  I tried boith and did not see a noticable difference in the waveforms. 

I am not sure how to hook up the oscilloscope with the ground on + sense.  Would I then hook up the input to - sense?  Would I disconnect the shorting bars?

I did hook the scope with the ground on - and input on + with shorting bars present between sense and output.  I am not sure I see anything meaningful.  I do see a 30-50Khz ish (fluctuating) spike with the scope sweep speed set high, (don’t recall how fast exactly).

On the HP 3456A I do note that as I increase the output the noise seems to travel with the increase in output.  That is, if I set the output of the 332D to 1.000000 I see noise in the uV range, if I set the output to 100.0000 I see noise in the mV range.  (noise is amplified?)

There is also a noticeable buzzing coming from the rear of the unit (HV mother board, pre reg, or series pass) that sometimes becomes louder for a brief period.  I do not notice any correlation to the increase in this buzzing and noise on the output.  I am wondering if other units exhibit this buzzing sound.   

[Dr. Frank]

On the above wave forms the output was set to 50.00000.  The manual is contradictory, it says to put range to 1000 but output to 50.0000 (1000 range and 5000000 dials would net 500v, 500.0000 no?).  I tried boith and did not see a noticable difference in the waveforms. 

They have to be identical, independant from range chosen.

I am not sure how to hook up the oscilloscope with the ground on + sense.  Would I then hook up the input to - sense?  Would I disconnect the shorting bars?
 

Simply have a look on the principle schematic. The 33x are simply big inverting opAmps, and + jack is the (virtual ) zero. Input can be tied to -. leave the shorting bars where they are.

I did hook the scope with the ground on - and input on + with shorting bars present between sense and output.  I am not sure I see anything meaningful.  I do see a 30-50Khz ish (fluctuating) spike with the scope sweep speed set high, (don’t recall how fast exactly).

On the HP 3456A I do note that as I increase the output the noise seems to travel with the increase in output.  That is, if I set the output of the 332D to 1.000000 I see noise in the uV range, if I set the output to 100.0000 I see noise in the mV range.  (noise is amplified?)

There is also a noticeable buzzing coming from the rear of the unit (HV mother board, pre reg, or series pass) that sometimes becomes louder for a brief period.  I do not notice any correlation to the increase in this buzzing and noise on the output.  I am wondering if other units exhibit this buzzing sound.   

That buzzing sound is probably the pre-regulator,which changes its duty cycle proportional to the output voltage.

noise is amplified, as the amplification factor is determined by the range switch (input resistor of opamp), and by the sample string (output/feedback resistor of the opamp)

[Fred_47]

On the above wave forms the output was set to 50.00000.  The manual is contradictory, it says to put range to 1000 but output to 50.0000.  I tried boith and did not see a noticable difference in the waveforms. 

I am not sure how to hook up the oscilloscope with the ground on + sense.  Would I then hook up the input to - sense?  Would I disconnect the shorting bars?

I did hook the scope with the ground on - and input on + with shorting bars present between sense and output.  I am not sure I see anything meaningful.  I do see a 30-50Khz ish (fluctuating) spike with the scope sweep speed set high, (don’t recall how fast exactly).

On the HP 3456A I do note that as I increase the output the noise seems to travel with the increase in output.  That is, if I set the output of the 332D to 1.000000 I see noise in the uV range, if I set the output to 100.0000 I see noise in the mV range.  (Nosie is amplified?)

There is also a noticeable buzzing coming from the rear of the unit (HV mother board, pre reg, or series pass) that sometimes becomes louder for a brief period.  I do not notice any correlation to the increase in this buzzing and noise on the output.  I am wondering if other units exhibit this buzzing sound.   

My 335D hums/buzzes when feeding high(ish) currents especially when in current limit. I very rarely have it above 100V but it probably makes noise when outputting high voltages.

Don't disconnect the sense terminals from the output terminals except to regulate voltages at the input terminals of a load.

Be very aware of the scope grounding; the ground prong on the plug at the wall socket is connected to the test lead 'ground/common'. You can get into very serious trouble if you isolate the scope case/chassis (assumed to be metal). Touching it or any other exposed metal (USB socket, BNC sockets, common of any other channel, etc.) while the scope common is connected to a voltage source can be fatal. If the scope is connected to anything (computer, scope cart, terminal whatever) the scope chassis common voltage is coupled thru to that device. There are many opportunities to let the magic smoke escape or worse.

The ground prong on the 332 is connected to the outer case*/front panel BUT NOT TO THE INNER CHASSIS WHICH IS CONNECTED TO THE + OUTPUT/SENSE! The + output is the common terminal for the various control power supplies. You need to think of 'ground' and 'common' (and 'Guard') as different nodes that can  sometimes have very high voltages between them. It is possible to ground (to earth) either the + output terminal or the neg output terminal or neither of them.

To do what Dr Frank asks you need to make sure that both of the 332 outputs are isolated from ground or that the + terminal is grounded. That way you don't have to float the scope.**

If you examine the mechanical attachment of the front panel to the inner chassis, you will see that the guard terminal is connected to a metal strip between them and there is a rather complex stack of metal and insulation. It's all arranged so that the front panel (and the screws attaching it) is grounded and the guard intercepts any leakage current between the inner chassis to the front panel. I think the guard is also connected to shield terminals on the transformers. For most usage, you can put a shorting bar between the guard and ground terminals on the front panel. I think the guard may be connected to the shield inside of the case but haven't checked to see if that is true.

My 335D had a shorting bar between the ground terminal and - output/sense terminals.

*The outer case is only grounded when the inner chassis is fully inserted. Otherwise it is floating. Same is true for the inner shield which may be connected to the guard terminal )or ground, I don't know which) when the chassis is fully inserted.

**Edit: You DO have to float the scope to examine anything in the pre-regulator circuit. It has direct ohmic connections to the power plug and is not isolated from the power outlet <by a transformer>. In that case power distribution wires labeled 'line (hot)', 'neutral' and 'ground' must be considered wrt the scope connections.

Edit again to add <by a transformer>

[valley001]

Hello Fred,

I need to study this more deeply (the + side used as common) before I hook up the scope again.  I am not fully grasping this concept.  I am very careful around this instrument.   The manual actually states that you must isolate the scope from ground when testing the pre regulator. 

Mine make a harsher buzzing sound all the time, even with in standby.  I gather this is not normal. 

[Fred_47]

Hello Fred,

I need to study this more deeply (the + side used as common) before I hook up the scope again.  I am not fully grasping this concept.  I am very careful around this instrument.   The manual actually states that you must isolate the scope from ground when testing the pre regulator. 

Mine make a harsher buzzing sound all the time, even with in standby.  I gather this is not normal.

Absolutely correct about the preregulator. It is on the line side of T2 with direct connections to the power plug. I will edit my post.

[Fred_47]

Hello Fred,

I need to study this more deeply (the + side used as common) before I hook up the scope again.  I am not fully grasping this concept.  I am very careful around this instrument.   The manual actually states that you must isolate the scope from ground when testing the pre regulator. 

Mine make a harsher buzzing sound all the time, even with in standby.  I gather this is not normal.

There is a circuit that drives a small current thru the series pass element even in standby. It is the source of the 700 odd volts that you measured early in this thread. The source is the big T1 transformer that supplies all of the control power (I found that there are 2 T1 transformers in this thing). The rectifiers and filter caps are on the HV MB. That could be the sound you hear.

As to the concept of + side being common; string the 3 sheets of the 1001 dwg together and follow the plus and minus leads back from the front panel. All of the controlling devices are within +/- 35V around the plus lead (except for the series pass element bias ckts @ 150V). The minus lead simply goes thru a relay contact to the neg side of the main rectifier (and positive side of the minimum current rectifier).

It's like any positive regulator except on a much larger scale.

[Fred_47]

Mine make a harsher buzzing sound all the time, even with in standby.  I gather this is not normal.

Thinking about your noise.

Maybe examine the HV MB in a dark room with the lights out.

I remember doing that with a noisy 230KV switchyard on the roof of a certain pumped storage hydro plant one night about 50 years ago. Quiet the light show with arcs jumping 2-3 sheds (about 6 inches) on 6 foot insulator stacks. The other startup engineers and I agreed that it was pretty spectacular, then we called the dispatcher and him shut down the switchyard before it had a chance to do an ultimate light show

[valley001]

Hello Everyone,

I thought I would update this.  I would like to thank member RandallMcRee, who very generously offered to send off a set of boards from his non working 332D.  Installing his chopper board has resulted in my 332D becoming stable.   It is still "warming up" but only dithers between a uV or so, no more radical spikes. 

I suppose I should look at waveforms on the chopper board and see what might be going on. 

[RandallMcRee]

I extricated the hermetic resistors out of my 332D parts mule to send to another member. The first deck of the KVD divider is completely isolated using teflon grommets and teflon wires. I have never seen this before. Thought it would be of interest to this group.