Fluke 332D DC Voltage Standard Repair

[classicTEK]

Good Morning All,

I just received my new-to-me Fluke 332D DC Voltage Standard.  Unfortunately it was dead on arrival with no fuses installed.  Opened up the top and slid it out of its case to find that the Pre-regulator PCB (A7A2) that has a "sticky" note attached to the K-1 Armature Relay.  There is also evidence of heat on R-4(a stud mounted, 1000 ohm, 8 watt resistor).  The "sticky" note that is slipped between the contacts on the relay says "BAD".

Any thoughts on how to go about getting this thing going?  I have worked on a lot of gear but most of it had other problems outside of the power supply.  I am concerned that if I am able to find a replacement K1 and R4, that when I do get the right fuses in there and try and fire it up, that I will just fry something else or the same parts all together.

Any help would be greatly appreciated.

Cheers,

David

[Dr. Frank]

Hi, The preregulator is the crucial PCB in these 33x instruments.
Download the manuals for 332D or 335D,  i.e. with the latest schematics.Google will lead you to an assets site of Fluke, vintage about 1972 / 1978.

The small transistors may have low betas, check for that. They deteoriate over time.
Then check power zener and both power switching transistors . They may be blown.
I wrecked one 332B/AF for that reason... power transistors were unobtainium.

Maybe the relais is damaged, but I doubt that. It's normal, that some power resistors show signs of heat.. After all these years.. Simply measure its resistance...

Please post some infos, photos about your unit, esp. which chopper, reference and sample string are inside.

Also take notice of the thread 332.. in the slaughterhouse

Frank

[classicTEK]

Hello Dr. Frank,

I found the Fluke 335D manual that you refer to and if you would be so kind as to call out the parts that are on the PCB layout labeled A8A2?  I assume that by "power zener" - you mean......CR6, 7, 8, and 9.  Correct?

Which "small transistors" and "power switching transistors" are you referring to?

I do hope I can get this fixed as it would otherwise end up as a possible parts unit for some folks.

My chopper board is labeled "REV-"H" - 251918" on the component side of the board and "333A-3004 REV-F" on the solder side.

I thought by getting the 332"D" version that I would automatically have the most up to date Chopper, is this not correct?

Cheers,

David

[Dr. Frank]

On A7A2, there'a heat sink. On it, there's CR14, Q1, Q2, power cases.
These were blown on my unit. Unobtainium. Maybe one may find alternatives in modern style.

The other transistors, Q3 ... Q9, small signal transistors, were either defect, or had low beta.

Can easily be replaced by equivalents.

I assume, that the damage was caused by this beta decrease after 40 years.. Found many other transistors in the 332B/AF which simply showed low beta..


Although D Version should have latest circuitry, it's often surprising, what's really used inside.

Pictures tell best... and the mentioned thread also shows  different PCBs.
The different version designators don't tell me anything..

Frank

[classicTEK]

Upon further inspection of said unit, it is apparent that it was someones parts mule as it is missing a shaft coupler on the 3rd digit from the left, the sample sting assy board is not attached to the frame, and who knows what else is broken, wrong, or missing.

DARN IT!!!!!!

This is worse than starting from scratch!!!!

ARGH!!!!

Anyone need any parts?  And NO................I am not giving them away for the price of shipping.

I will however entertain reasonable offers for the parts you may need.  Everything seems to be in order on the individual PCB's.

The 332D is Serial Number 38515 and is located in Akron, Ohio, USA

Cheers,

David

[doktor pyta]

I had a problem in my Fluke 343A. One of reed relays was magnetized so if the coil was energized once, the contact was remaining closed even after turning off the instrument. There were two problems: first with the "sticky- magnetized" reed relay and second - with the resistor in the circuitry that was driving coil of this relay. The resistor changed it's value by 30%...

[Vgkid]

That stinks about your 332, I will give you $3.50 for the front switches/ resistor assembly + plus shipping...
For those who know: Does the voltage divider use a KVD, I'm guessing it does as the switches are 2 tiered.

[classicTEK]

You intended $350.00 ..............correct?

[Vgkid]

$350.00... Its not April first yet
$35.00 with shipping...

[kxenos]

If you came here to sell just write in the buy/sell category, don't pretend you need help to repair

[Dr. Frank]

That stinks about your 332, I will give you $3.50 for the front switches/ resistor assembly + plus shipping...
For those who know: Does the voltage divider use a KVD, I'm guessing it does as the switches are 2 tiered.

Maybe you should better read the manual/schematics of the 332/335, before you buy that sample string PCB..
Theory Of Operation is very instructive, btw.

No, there is no voltage divider, and no, there is no KVD, the 332/335 work totally different.. like an inverting OPAMP, with range resistors and sample string acting as the OpAmps feedback resistors

Frank

[classicTEK]

Hmmmmmm................some folks just love jumping to conclusions.

What about "I would love to get this going," do you not understand.  No pretending here, no false pretenses. I just wanted to fix it, but CLEARLY there is some hidden meaning in "I would love to get this going".  You should be ashamed of yourself for being such an ARSE!

If I do decide to actually part it out, I certainly would not do it here.  There are far more deserving people on earth than some of those who have shown their ass here.

Thank you very much Dr. Frank for you guidance and wisdom.  If there is any more communication, I would prefer private email to avoid the cattle.

Cheers

[Vgkid]

Frank

I was wondering why they(Fluke and I know in the EDC calibrators) they use these resistor strings as feedback resistors rather than using the reference to create a stable 10V reverence, then feed it through a voltage divider to either an output buffer/amplifier to get the desired voltage. This way you could easily account for voltage reference drift/ divide drift seperately.

[Dr. Frank]

Frank

I was wondering why they(Fluke and I know in the EDC calibrators) they use these resistor strings as feedback resistors rather than using the reference to create a stable 10V reverence, then feed it through a voltage divider to either an output buffer/amplifier to get the desired voltage. This way you could easily account for voltage reference drift/ divide drift seperately.

That's an interesting argument.
Here are some easy answers:

1. It makes no difference if you chose one topology, or the other.. due to error chain calculus, the error sum stays the same.. as all errors add up, the lesser complicated topology (only one resistor chain or divider) is more stable.
2. KV dividers are much more complicated, i.e. due to switching and resistor manufacturing, compared to simple linear resistor chains (decade box type)
3. A topology with fixed reference, divided by KV and again amplified by a resistive divider would a) imply an additional source of error which would add up more than the simpler topology, and b) would maybe require a non-inverting amplifier. Latter one is worse compared to the non-inverting type, because you have common-mode errors of the OpAmp, and null trimming is not so easy.



The range and sample string resistors in the 33xD are very interestingly matched:
The range resistor consists of two resistors, with exactly inverse T.C.: They are stamped 749.25k .. P1.0 and N1.0 or 74.925k .. P2.0 and N1.5.
So they have a maximum T.C. of 0.5ppm/K.
Same goes for the sample string, in the first three decades.. The resistors are combined in groups of 1, 2 or 4 resistors in series or in parallel.
The single resistors have 0.5ppm/K only, but the groups of 2 or 4 are all matched to virtually zero T.C. (P1.0 + P1.0 + N1.0 + N1.0)

Therefore, the combined inverting amplifier has a maximum T.C. of 1ppm/K, or even less, if range and sample string resistors are also matched to less than 0.5ppm/K.
That's still valid as per today.. I've briefly checked that just recently

Therefore, these resistors, plus the switches, are still much more worth than those lousy 35$..   

Frank

[kxenos]

Hmmmmmm................some folks just love jumping to conclusions.

What about "I would love to get this going," do you not understand.  No pretending here, no false pretenses. I just wanted to fix it, but CLEARLY there is some hidden meaning in "I would love to get this going".  You should be ashamed of yourself for being such an ARSE!

If I do decide to actually part it out, I certainly would not do it here.  There are far more deserving people on earth than some of those who have shown their ass here.

Thank you very much Dr. Frank for you guidance and wisdom.  If there is any more communication, I would prefer private email to avoid the cattle.

Cheers

Dave, I apologize. At that time I was thinking that you just wanted to see if there is an easy fix and that when you saw that there isn't any you decided after 2 hours to "entertain reasonable offers". I usually don't attack people like that but obviously your intentions where misinterpreted.
If you get to Athens I'll buy beers... 

Kostas

[Vgkid]

Frank
snip snip

That makes since, as  I was curious as to why they used 2 resistors in the loop. I figured it was either to cancel out the tempco, or to cancel out the tolerance.