FLUKE 3330B no output but relays clicks - some life in it - Now FIXED

[kawal]

Got a nice piece of gear in great overall physical condition. This is the FLUKE 3330B from 1974.   I looked over all the boards and replaced a few dried out caps - so far found 2 that measured high ESR and complete lack of capacitance when removed. Both blue colored electrolytics but there could be more that i missed.  Also had  to replace burnt insulators and cables on two power resisters in the pre-regulator board.
I checked fuses and made sure links are installed in the outputs. So far the unit starts up and i hear relays clicking but there is zero output. No matter what the output setting , crowbar option, current limit option I get zero reading from the output.

I did measure most of the large power transistors and found a strange thing on the pre regulator  board . The regulator transistor Q1  (round case on the heatsink) measures like an PNP but is supposed to be NPN. I double checked it but will check again. I did that in circuit so maybe that is the issue.

Next plan of attack is to remove the power supply boards one by one and test outside the circuit. Also most bulbs need replacing. probably will install some Leds in place of the bulbs.  The bulbs are on the -25V supply so might need to move them to the -5V.  Two if the bulbs light up so most likely -25V is there.  When changing setting there seems to be clicking of relays on the ladder boards but not 100%  sure.

I am open to suggestions on the right way to proceed with troubleshooting this kit.

I also need advice on the lacquer used on the board.  When done repairing what should i use for protecting the board ?

Here are some pictures of the beauty:
https://photos.app.goo.gl/1eqCgKT6KxZD8Hzh9

Help appreciated

[Dr. Frank]

The 3330 shares several compatible PCBs with the 332/335 calibrators, like the reference, pre-regulator board, crowbar, chopper, and so on.

My 332B/AF was wrecked, because on this pre-regulator board all semi conductors had died, especially the fat zener and end transistor, all other had no beta any more.
See here: https://www.eevblog.com/forum/testgear/fluke-332baf-in-the-slaughterhouse/msg393609/#msg393609

I think, I had exactly the same pre-regulator PCB as yours.

Check that you get the correct schematics, because there were many basic changes.
Get the latest 332D manual with schematics also.

You may check the other voltage supplies, apart from the HV pre regulated one, which you may pull outside upfront to avoid dangerous voltages.
Check whether the reference module delivers its ref voltage of about 6.71838V.

You may temporarily provide 15V from an external PSU to the +/- supply points of the pre-regulator, simply to check if the instrument is working besides that.

I also replaced all the light bulbs by LEDs, simply by putting a resistor of a few kOhm (2.2k ??) in series to the 25V supply.

Good luck, that 3330 is a fine instrument, and yours is just looking crisp.

Frank

[kawal]

Frank

Good advice. I was planning to take out the low voltage supplies and bench test them for proper voltages.  I will also test the reference on the bench.
Looks like no one has messed with this unit so there is hope in fixing it.
I will get the 332D schematic to compare it to the 3330B.  I found 2 versions of the 3330B  manual online but both are the same vintage.  I do not see any differences apart from one part hand corrected in the part list.
I will do the Leds tonight to see if the correct functions are turning on.

Maciej

[kawal]

Frank

The boards might be compatible but they are different. The 3330B has a sub par voltage reference when compared to t the 332D. The newer (1981) 332D has a oven stabilized reference and the 3330B (1974) uses the same reference but without the oven.
The pre regulator is a little different but not a lot - the idea is the same. I am still going over the schematics.  Again thanks for the tip.
Maciej

[kawal]

Checked the voltages on Aux power and main power +25V, -25V -5V -15V all there . Should have load tested but forgot to do it.
Reference had a  blown output transistor (I might have caused that...ooops). The schematic for that board shows wrong type off output transistor as NPN when old part was PNP. I replaced it with one NPN an struggled for a while with a 13.5V output . Finally replaced that transistor with a general purpose PNP and the board started working.   Reference shows 10.000014V on my Keithly 192 at room temperature (19C) and goes to 9.99983 when heating the zener with my finger for a few min (lets assume 30C-19C=>11c  difference 31ppm/11C  or ~3ppm/C) . I guess ok for reference without oven.

At that point i tried my luck with inserting all boards but still no output.
I did notice the pre regulator  is getting really hot - specifically the 2 power  10 Ohm power resistors. The power transistor measures ok for conductivity but did not check Hfe.

[Dr. Frank]

Correct, Q2, the output transistor, is on the wrong side of Q1, to make an NPN Darlington pair, so it must be a PNP. That's also in the BOM of A4A1.
Although this version of reference has  no oven, the Reference Amplifier IC2 is usually trimmed to <1ppm/°C. See also 731A/B reference. This zero T.C. occurs at a certain environmental temperature, in this case, that should be the inner working temperature of the 3330B.

The currents through R4 and R15 determine the T.C. of the zener diode and of the transistor diode inside IC2, which have different sign.
The zener current through R4 is fixed to about 3mA, and R15 is determined at the factory to give about 38µA for a combined zero T.C. , as written on the label, a bit misleading: 'Icz  ... 38'
Calculation: 1.23 Volt across R15 (given by R4 over R3), R15 is 32.4k, and that gives 1.23V/32.4k = 38.0µA

So you may desolder R15 on one side and check, whether its value has changed (I assume, it drifted upwards). If that's not the case, you may later trim R15 for zero T.C.


My pre regulator board had sockets for the small transistors, and most of them had hfe < 10, or so.
Q1 and CR14 both were blown.

And pay attention.. if you open the case, there are some switches to disengage dangerous HV, which is present on the inner cage.


Frank

[kawal]

Went back to the Pre regulator and checked all the Socketed transistors . one of the SS7504 shows hfe =0  rest show 30...50 range , seems to conduct fine in diod test.   Only if i knew what the specs are for SS7504.

[wictor]

I have a question: why is beta decreasing with those transistors?

[Dr. Frank]

The SS7504 is obviously a PNP transistor, I have 4EA here, hfe between 40 ..100.
The case size looks like a 500mA Si-pnp type, but I assume any general purpose pnp @ 100mA will do the job.
Driving currents of all pnp seems to be < 30mA @ 10V, so maybe you better use a 500mW type,  hfe >= 100.

I also found that most of the smaller transistors on this board showed a decay of beta, all hfe being lower than 10.
Reason for that may be the regular, high inductive pulses on this board, which probably killed the power zener and the driving transistor.
Another might be heat and age, and the bad quality of the silicon crystals / wafer in the 1970ties.

 

[wictor]

Thanks for explanation!

[kawal]

Frank

Found some specs for SS7504 on www.csgparts.com
PNP  modest gain 40...100 per Frank   - modest CE voltage (85V) per spec


I used 2SA1220  as replacement. (Vce =120V...160V),  gain 100....300, current 1.2A.

Still no go after replacement and still a lot of heat from the 5ohm resistor. That worries me as this means the transistor Q2 is fully conducting from the 10V supply to ground.  There should be no reason for that.  I will have to setup some kind of external test for the board. Q2 must be conducting to ground but dont know why   



Got the LEDs installed  and wired for the decimal points and  working on the other control lights. Decimal points  are easy - just cut the trace for
the -25V common  added a 0.625W 2kohm resistor (made from 5x 10k 0.125W) in the cut.  I drilled the PCB near the cut trace to locate the resistor on the parts side.  I snipped off 3 of the 4 commons. All decimals connected as common Cathode  to the 2kOhm. The anodes are individually connected to the signal  wires- switched ground.  That circuit is at  ~20ma supply to  white diod with 2.6V drop and is plenty bright. The diod's are epoxied to the aluminum light pipes.

I plan to get the rest of the control indicators wired this week.  Maybe that will give me more information on the fault - maybe over current or over voltage fault ?

I think most parts can be replaced with more modern parts but perhaps  in different packages. 

I hoped for an easy fix (with some measuring and shotgun replacement of parts)  but it looks like this one will take a while to troubleshoot due to the dangerous nature of the High Voltage.

[wictor]

I also have that 3330B. It had quite many problems also. Some of the white reed switches were bad, contact problems in board connectors, some wire was broken near big trafo and if I recall right the overcurrent protection was tripping.. I got it working, but still need to fix the front panel indicator lights.

[Dr. Frank]

1.5A transistors might be a bit too beefy.

Have you checked all the diodes, especially CR15 between base of Q1 and GND?

Maybe Q1 is also damaged, that's no darlington, and what about CR14?

[kawal]

Wictor
Sounds like you got it working without the lights and I will have the lights working first.   
I am jealous that you where able to fix.  A good catch on the reed switches - I started measuring the resistance of the coils but maybe  it is better to try to energize them with an external power supply to confirm they work. I will add that to my to do list. 

If you have any more notes or any more details it would be great to post here.  Some people in the future will have it much easier when they try repairs on this 45 year old tech.

Maciej

[kawal]

Frank
First thing is - thank you so much that you are replying to this thread. Its good to know there is someone out there that has experience with the 3330B.    I am looking forward to your posts.

Q1 measured OK on a diod test  but gain is really high when doing HFE  for some reason. - maybe that is the indication that its not OK ?

Maciej

[wictor]

Wictor
Sounds like you got it working without the lights and I will have the lights working first.   
I am jealous that you where able to fix.  A good catch on the reed switches - I started measuring the resistance of the coils but maybe  it is better to try to energize them with an external power supply to confirm they work. I will add that to my to do list. 

If you have any more notes or any more details it would be great to post here.  Some people in the future will have it much easier when they try repairs on this 45 year old tech.

Maciej

Thread is here: https://www.eevblog.com/forum/repair/fluke-3330b-questions/msg1377140/?topicseen#msg1377140
But it does not have a lot info what was wrong with the unit... I'll try to collect some info about what parts I had to fix/change.

[wictor]

Relay K2 was broken on relay driver board also with my unit:


There was also one broken relay under preregulator board, replaced it with Meder relay:


Ladder board relay K3 broken solder joints:


I think that also orange wire(5th from top right) pin was broken:


Some mods done by someone:

[kawal]

Wictor
While looking at the K2 board I noticed a +35V line but when i was testing the power supply boards I did not see that voltage.  It seems that the +/- 35V supply is on the Current limiter board so i did not think to look at it.
I will be checking that next....

[kawal]

Ok Finally making progress.
The Circuit limiter board that contains the +35V and -35V  power supply was checked today. I first checked the transistor in the sockets -
Q3, Q5, Q6 all SS7504 measured HFE 53, 56, 34 respectively. So all good
Q7, Q4  are 203489 NPN 60V 0.8W measured Hfe of 105 and 140
power transistor Q1  did diod check = OK
Power transistor Q2 - 2N297A was a little puzzling - a 35W Germanium 5A Collector max. Was measuring .1V on junctions

Tried to power the board -the 35V positive was fine
The 35V negative was less then 1V . Turns out 36V zener is dead measures 0.05V in either direction on multimeter and about 0.8V with 5...20ma current using bench power supply.
I didn't have a zener to replace the broken one but installed a 3.3K resistor in its place.  The supply is outputting 23V or so.
I installed all the cards and the system now works. Will need to get the zener diod and finish the diods on the front.


First least significant digit decade has a problem with the 4th digit.  The relay for the 4 position is stuck or driver is broken - I measured the coils show the same resistance. of 2kOhm .  I am happy for today.

   
So I would call that progress

I updated the pictures

https://photos.app.goo.gl/1eqCgKT6KxZD8Hzh9

BTW - how to load more than 1 picture at a time to the post ?

Maciej
 

[wictor]

Hi,

I had similar digit problems caused by bad connection between the A4A10 ladder driver board and main board. It's the board nearest to lowest digit.

Wictor

[Dr. Frank]

Wow, that's great.   
Reed contacts may easily be available, if stuck.

[kawal]

Frank
For the reed contacts I removed each card that has reed switches and removed all the drive transistors. 
I then setup my power supply for 20V and cycled each reed switch a few times. I verified the Resistance in open and closed positions and moved to the next switch.
After that I replaced all the transistors  and reinserted all the cards. All reed switches worked via power supply  but each one was making a different "ping"  sound  when turning on. Some louder some less loud.

This took about 20 min and seems to have  fixed the issue with the first decade. I dont know what actually helped : either cycling the reed switches or by re seating  the drive  transistors. Either way the switches now function in each decade which i verified. 

I do get a big drift (300uV) and some inaccurate readings (scaling) but I will not worry about it till I replace the zener diod.  The -35V  supply is definitely not stable and definitely not at -35V. 
I ordered a lot of capacitors to do a total recap with the exclusion of the 1200V caps (these are expensive and most likely OK ).  After that  i will recheck the source and calibrate if its needed. So far i did not adjust anything hoping its within specs .
 I am so happy to get this working so fast and without getting shocked.

Maciej

[Dr. Frank]

These reed contacts really have a very long lifetime.

Anyhow, I've repaired a lot of analogue instruments (e.g.  P.A.R. Lock-In -amplifiers) from the 80/70ties, which had a lot of bad reed contacts. Cycling alone would help only temporarily.
Can't tell anymore, whether these made bad contact, or no contact at all, but anyhow, the clicking sound alone, which I also observed/listened to, was not the only criterion.  In the end, it's a quick fix to simply replace these reed contacts (apart from the coil).
Maybe you can make some electrical measurements, like on-resistance on these..should be in the mOhm range.

I think, over time these loose their spring force, and show higher than usual resistance.

[kawal]

Frank

I went back in and measured the reed contacts on all the ladders. The resistance of the contacts with 4 wire  Ohm measurement on the Keithley 192 andthe readings  are between 35mOhms and 50mOhms on all of the contacts. Based on the results I think that they are still in very good condition.
I  made some measurements and the calibrator is fairly spot on - The reference  was off by  off by a lot when put back in as the replacement transistor i put in the reference has Emitter and Collector reversed. Switching the terminals of the transistor made the  reference pop back in to 10.0000X volt.   
Now waiting on caps and zener diods which should arrived Tuesday.

P.S. The more I use the Keithley 192 the more I like it and the less I use the HP 3478. I think its time to sell the HP meter.
The Keithley needs a slight adjustment as it shows different results on the different ranges.
Maciej

[wictor]

Hi,
I mentioned that I had to replace two reed relays, one smaller NO relay and one bigger three pole relay. Now the smaller one seems to be fine, when measured in test bench. The on resistance is 20 mOhms, coil resistance is about 4.2k and I think it was about 2k with others same size relays.

The bigger one had problem with both NO and NC contacts. NC is showing resistance of 20-60 ohms when coil is not energized and is opening properly when coil is energized. The NO does not close when coil is energized and when coil is not energized, it shows resistance of few mega ohms. Coil resistance is 2.4k.

Wictor