Yokogawa 2858 AC Voltage Standard Service Manual

[asp]

Hi
   Does anyone have any service manual for the Yokogawa 2858 ? Or anything that might help debugging the unit. My unit is faulty as it's waaay off by some 1000 orders of magnitude or more sometime (and I don't think that's tweackable). It does not seem to be constant even so it has to be some comparator somewhere on the board I guess.

Thanks

[asp]

So what I have thus far is this:

There are 11 boards on the back of the unit ( plug boards ) numered p01-p11

P01 deals with switching on the range selection leds (err... light bulbs)
P02 high voltage resistor divider part 1
P03 high voltage resistor divider part 2
P04 *think it's the 3v voltage standard for the oscilator as it has that z1 there and it's near the oscilator*
P05 oscilator ( works as it changes freq to the one displayed selected on the front panel )
P06 ?
P07 ?
P08 ? (has one relay on it)
P09 ?
P10 ? (another relay on it)
P11 ?

Below are some pics of the boards

What should I check on next?

[asp]

More pics

[asp]

And some more

So any thoughts what should I do next? I did check all the caps on the boards and all have low esr ( except the ones in the pF range which are ~1K ). I did not check any transistor on the boards but I don't think they are the problem

PS: the voltage standard was made in 1974 and last calibrated in 2008

[acbern]

If you say the pF capacitors have 1k of resistance, thats definitely not ok.

If your problem is intermittend, then potential candidates for this normally are relays (bad contact), cold solder joints, cracked components, bad PCB traces, bad active part (causing noise oickup); in that sequence

[asp]

acbern: 1k at lease in circuit. haven't taken them out

I did some measurements of the acutal selected voltage and output voltage and here are the results:

On the left it's the selected voltage on the ac voltage standard and on the right the actual output voltage.

10v range:
1.000v -> 0.9469v
1.100v -> 1.1357v
1.200v -> 1.3246v
1.300v -> 1.5136v
1.900v -> 2.652v
1.999v -> 2.837v
2.000v -> 4.024v
3.000v -> 5.886v
4.000v -> 7.736v
5.000v -> 9.601v

20v range:
9.00v -> 13.913v
10.00v -> 7.732v
14.00v -> 13.912v
16.00v -> 17.000v

50v range:
10.00v -> 9.015v
19.00v -> 25.09v
19.99v -> 26.86v
20.00v -> 39.99v

[lowimpedance]

Okay without any circuit diagrams its down to checking as many parts on each of the plug in boards then the main unit itself (pictures!).
First check the precision resistors (the ones with the more odd values on them), allow that in cct testing may give some differing results.
Clean the plug in connectors and edge contacts.
Check the diodes.
Check the reed relays are actually working  ie N/O and N/C contacts are good , coil is not open etc.
What are you using to measure the output. what frequency is it?.

[asp]

The output frequency is the selected frequency on the front panel. ( for measuring the freq I am using a Agilent U1272A ). I was thinking about some opamp maybe?
Still looking for the precision resistor ( might be 2 of 10K since they look different but don't know ) (not sure even on what board to look ). The unit itself is just selector switches and 2 big transformers. The plug boards and 2 x 2SC1116 mounted on heatsink is the whole electronics.

I was thinking if I can find any drop-in replacement for the MC1439 single opamp (as I would change them all to some dip ones) they seem to be almost compatible with LM318 (with some differeces). On the mc1439 pin 1 is Compensation, on LM318 is BAL/COMP. Pin 5 on MC1439 is Output Lag on LM318 is BAL/COMP and pin.
Actually pin 5 is not used on the board

Later Edit: I did check the reed relays and they seem to work and most of the resistors seem to be ok (haven't tested them all).

[lowimpedance]

I would not be immediately be thinking opamps, perhaps the gain of a stage which is external circuit paths. Did you test all diodes and transistors?.

The precision resistors like the brown ones!. A change in resistor values can shift bias points/ affect gain ratios etc !.

Do you own an oscilloscope?, if yes what does the output look like (ie not distorted or dc offsets etc)
Are you sure the switch contacts and edge connectors are clean (particularly the front panel switches with gear of this vintage they will have tarnished somewhat.
Can you post some photo's of the main unit insides around the switches if easy to get to ?.

If you must change the opamps go ahead as there should be plenty suitable types that can be made to fit.
Just remove the compensation components.

The output frequency is the selected frequency on the front panel. .

And that was??.

[asp]

I will try to get some pics of the unit tomorrow (also will post some screenshot from the oscilloscope of the actual output (if I can get the ethernet on it to work) ). I also cleaned all the contacts before I actually asked on the forum.

There are plenty of transistors around the unit and just unsoldering them takes some time. I was also thinking that maybe it's the sense/feedback circuit as it might increase the voltage over the selected one.

Output frequency is 50, 60 or 400 Hz depending on what I select on the front panel ( if I select 50Hz than the output it's 50Hz (actually 49.9xx), if I select 60 it's 60 ), but the output voltage does not seem to be affected by selected frequency (the measurements above have all been done @50Hz)

All the brown resistors seem to be in specs as well ( will post tomorrow the actual values )

I did manage to break one of the output fuses 2 times as well but I have no clue how (maybe at the > 500V range??! )

[lowimpedance]

I would not bother to remove the transistors as yet, just a quick 'diode' test between B-C and B-E and no shorts from C-E etc to eliminate anything obvious. same with the diodes.

[ddrl46]

I should have access to the service manual for this thing, I'll try to get you a scan if I don't forget.

The unit I had to work on a while ago had 2A fuses on the output, even said so on the rear panel, but the service manual seems to insist on installing 4A fuses. You might want to check that on your unit as well.

Also don't forget to switch the voltage setting all the way back to 0 before going to the higher voltage ranges or it goes into alarm, same goes for changing output frequency.

[asp]

The unit I had to work on a while ago had 2A fuses on the output, even said so on the rear panel, but the service manual seems to insist on installing 4A fuses. You might want to check that on your unit as well.

Mine has 2A delayed fuses. I could install 4A fuses but that won't solve the problem (maybe it won't blow that many fuses).
Would be great if you can find and scan the service manual for this.

Below is the screen capture from my osc. ( at the same time I tested the output voltage with 2 DMMs and they ware 5.650V AC (Uni-T UT71E) and 5.680 (Agilent U1272A)

It seems that the precision resistors are out by a bit (most notably the 50K and 30K ones) but tweaking the trim pot on the board did no do anything (so now it's back to it's original place)

so here are the tested values:

50KA ( A seems to be 0.05% ) = 50.05K, 50.06k, 50.01k, 50.04k, 50.04k, 50.02k, 50.06K (all seem to be out as the range should be  between 49.975k and 50.025 K)
30KA ( same 0.05% ) = 30.02K, 30.05K (all seem to be out as the range should be between 29.985 and 30.015 K)
13.333KA (same 0.05% ) = 13.355K, 13.354K ( all seem to be OK as the range should be between 12.66635k and 13.99965k)
3.333KA ( same 0.05% ) = 3.335K ( seems to be in range ( 3.16635k - 3.49965k ) )
580ohm C ( C seems to be 2% ) = 0.5791K ( seems to be in range ( 0.5684k - 0.5916k ) )

Edit: @60hz you can see some wobble effect (with persistance on) that is visible with the naked eye as well
Later Edit: Added photos with the unit.

[asp]

So I tested almost all the transistors on the board and all seem to be ok... I still have the LM301 and MC opamps ( and some 4 pin mosftes ( 3SK14 ) ( I know I can still threat them as normal fets and test them but ... yeh... ) ).

I think I will start replacing the actual transistors ( C538 with some BC548C also it's counterpart BC556 and the C1116 with 2SC1116A ( unless I can find a better option ) ). I will also start replacing the 2 IC's on the last board.

I still hope that ddrl46 can find and scan the service manual.

[ddrl46]

Here you go, scanned the manual: hhttps://www.dropbox.com/s/d36ojrbzapvdnya/Yokogawa%20Electric%20Works%20Type%202858.pdf?dl=0

I hope it helps.

[asp]

thanks ddrl46 for the service manual.
   I am still trying to debug the thing. the one thing I noticed is that I don't get the 5.000V on pin 18 on P06 ( I actually get 7.5V ), but I did simulate the schematic in proteus and it does not come to 5V anyway ( so I am kind of stuck at that part ).
   Below is the simulated screenshot and the corresponding schematic from the manual. Should I use a 5V refferece instead of the opamp and resistors? (I could easily swap them out for a AD586JNZ which I can trim to 5.000V (hope it's + not - 5) )

Edit: My unit uses a BZ820 which is equivalent to 1N4884 which in turn is equivalent to 1N4747A ( the zener I use in the simulation ), and in the simulation I am not able to change the voltage either by changing the pot ( or resistor values )

[asp]

So, in the end, after 6 years of having the unit in a corner of the lab I decided to give it another go and have found lots of issues with it. It had a bad resistor on rotary switch nr1 (this one was really hard to get to in order to measure it because of other components getting in the way), another bad resistor on the medium amp stage board, 3 opamps on frequency generator board, 2 opams on FET controll board. I had to adjust the circuit on P04 and add a trimpot in series with R9 so i can adjust the -15V without changing +15 as they had to be as close as possible in module.
I am working now on replacing the VU meters (as they have gone bad) with a LCD that will display the load, I will change the trimpots on the P* boards some multiturn trimpots and I'll also change the front panel light bulbs with leds so the unit gets a more modern design. Once I have everything done I will post some photos of the unit