Keithley 237 repair

[rolkinas]

Hello,
I bought Keithley 237 from ebay "used and 100% working" condition.
It appeared to be not outputing any voltage/current due to some failures such as all caps in resonant converter are dry and converter not working, 4 caps 100uF for +-30V are dry and there are no 30V present, triac in 50V supply is leaky and one 1N4577 in output stage was open.
bit too much for used and perfectly working condition (as stated in listing) from Italy (:
fit temporary parts to see functionality
Now I have a question. if I input by keypad 1 volt - unit outputs 10V. Moving decimal point to right I manage to get 1 volt but in other range. the same is with all other voltages from 1 to 11 volts. 12V and up - are fine (12V means 12V 50V means 50V etc. everything below 11V are multiplied by ten except when decimal point is before 2 digits right to left. The same is for negative and positive values.
current programming is OK
Just for clarification I added pictures.
Will be looking for issues

[rolkinas]

Moving towards last steps in repair.
Replaced all electrolytic caps in resonant converter. In power supply section replaced C52, C52, C53, C54, C56, C57, Q7
Replaced shorted Q33 and defective U31
Last issue almost solved. It is burnt 2.77K PTF56 series resistor in feedback divider network.
Have anyone idea where to get stable 2.77K ?
Also, what is original fan specifications in regards of required air volume for K237 ?

[coromonadalix]

For the 2.77 k  resistor   you have a few

Vishay Dale RNC50J2771BSB1431

PHILIPS B0207C2K770B1A metal film resistor. 2.77 kOhm1/2W 0.100% tolerance


It depends on the wattage  and availbility in your country

Digikey, Avnet, Arrows, Future Electronics   

[rolkinas]

I looked in Farnell, Mouser, Avnet, Digikey, RS, TME, Distrilec, but unfortunately they have no stock of 2.77K . Only lead time of approx 20 weeks+
I am open to offers for 2.77K 0.125W - 0.25W

[MiDi]

Hope it is ok to hijack this thread for this interesting unit:





If it would be only that, I would be happy...



Clear indication to turn it apart before turning it on - as I would not do that anyway 



The sticker that states "does not turn on" probably has falling off somewhere...

After pulling out the transformer - which was an easy part as preowner did preliminary work already - the first check was resistance of all windings followed by checking voltages @242V~:

RD - BK	6	121V~
RD/WH - BK/WH	5.5	121V~
GY - GY	0.35	11V~
YE - YE/WH - YE	25 each	157V~ each
OG - OG/WH - OG	2.7 each	21.8V~ each
BN - BN	0.7	10.7V~
BU - BU/WH	2.2	42V~
VT	High-Z	-

Seems to be reasonable, but transformer drawing 580mA and 9.5W (real power) without any load might be a bit high.

After initial checks on all other boards and cleanup of the inside of the whole unit, the tranformer went back into it.
First only digital board was connected to see if unit powers up.
It did without any problem, everything nice and cool despite the LM323 Heatsink which gets rather hot - this seems to be ok.
The other boards were connected and unit came up without any issues - sorry, no magic smoke escaped for you

Power consumption is 58W and ~600mA @~230V when outputting 10V.
The fan consumes a whopping 18W @ 121V~, so rest of unit should be ~40W.

It came with a 1A fuse installed, but for 230V a 0.5A fuse should be used, which would not last long in this particular unit.
So current drawing from line might be a bit to high - would be nice if someone could measure current @230V~/50Hz to compare.



First test of source part succeeded and gave reasonable values - just some digits of.
Have to test the measurement part and will give an update...

PS: Have taken pics from all boards top/bottom and details - so if there is interest, I can post them, but there are already a couple e.g. Keitley 236 Source Meter - Teardown and Review.

[rolkinas]

Hi, feel free to hijack this thread
My 237 consumes about 330mA either in idle and voltage source mode (without any load)
I can't remember exact power of fan. I replaced it while it was not original and worn out.
I suggest to check temperature of MAC223 triac. It may be semi open and cause high intermittent current draw as in my unit. Therefore I had to replace it.

[rolkinas]

Small follow up to my repair.
2.8k instead of 2.77k resistor in input divider worked well. Calibration-adjustment of lower ranges were performed without any problems.
In my case upper voltage range initially was quite off. Looks like already failing unit was calibrated by previous owner. It might be that Keithley has calibration boundaries in terms of call- to call value difference and in terms of absolute value. Had to perform several adjustment sessions to be able to adjust it to real value. Adjusting firstly negative and then positive 1000V  (as opposed to manual) helped a lot.

[BU508A]

I looked in Farnell, Mouser, Avnet, Digikey, RS, TME, Distrilec, but unfortunately they have no stock of 2.77K . Only lead time of approx 20 weeks+
I am open to offers for 2.77K 0.125W - 0.25W

How about combining a 70.6R and a 2700R resistor?
Okay. the 70.6R has only 100mW but I think in series with 2700R resistor this shouldn't be an issue.
All come with 0.1%.

70.6R:
https://www.mouser.de/ProductDetail/Vishay-Dale/RN55C70R6BB14?qs=sGAEpiMZZMsPqMdJzcrNwsAoK60ltMOoV2kySbHX6ZI%3D

2700R:
https://www.mouser.de/ProductDetail/Vishay-Dale/RN60C2701BB14?qs=sGAEpiMZZMsPqMdJzcrNwsDBSEuVlBlrkPFjbJ8wHr8%3D

[Neomys Sapiens]

I looked in Farnell, Mouser, Avnet, Digikey, RS, TME, Distrilec, but unfortunately they have no stock of 2.77K . Only lead time of approx 20 weeks+
I am open to offers for 2.77K 0.125W - 0.25W

How about combining a 70.6R and a 2700R resistor?
Okay. the 70.6R has only 100mW but I think in series with 2700R resistor this shouldn't be an issue.
All come with 0.1%.

70.6R:
https://www.mouser.de/ProductDetail/Vishay-Dale/RN55C70R6BB14?qs=sGAEpiMZZMsPqMdJzcrNwsAoK60ltMOoV2kySbHX6ZI%3D

2700R:
https://www.mouser.de/ProductDetail/Vishay-Dale/RN60C2701BB14?qs=sGAEpiMZZMsPqMdJzcrNwsDBSEuVlBlrkPFjbJ8wHr8%3D

2700R + 69.8R, but only in 1% 50ppm/°K 0.125W

[rolkinas]

50ppm/C might be too much for such use.
I put this one:
https://www.mouser.de/ProductDetail/Vishay-Dale/PTF562K8000BZEK?qs=41svs21nsHB8T7Nf7ZeGtg%3D%3D
PTF56 series as original but 2,8k instead of 2,77
Currently I am going to leave that as is. At least since 2,77 with better than 5ppm/C becomes available somewhere

Almost forgot to mention that before performing calibration unit being in voltage source mode started to be noisy after several hours being ON. That was caused by faulty 150pF NPO !!! capacitor.

[HighVoltage]

I have found it to be easier sometimes to buy a second unit in working condition and then compare signals.
My problem usually is, that I keep the second instrument as well.

[rolkinas]

I have found it to be easier sometimes to buy a second unit in working condition and then compare signals.
My problem usually is, that I keep the second instrument as well.

I have same sickness   My inventory tends to be growing in pairs
unfortunately (or maybe fortunately) there is only one K237 in my lab this time
K237 is quite straight forward unit. But as always we need some time to familiarize with design

[MiDi]

RD - BK	6	121V~
RD/WH - BK/WH	5.5	121V~
GY - GY	0.35	11V~
YE - YE/WH - YE	25 each	157V~ each
OG - OG/WH - OG	2.7 each	21.8V~ each
BN - BN	0.7	10.7V~
BU - BU/WH	2.2	42V~
VT	High-Z	-

Would appreciate if someone could do measurements on unit, especially transformer to compare:
-Current/Real power consumption from line and line voltage
-Voltages on all tranformer taps (see above table)
-Resistance of primary windings (unplugged!)
-Bonus: Temperature difference on top of LM323 to environment (warmed up unit)

[rolkinas]

RD - BK	6	121V~
RD/WH - BK/WH	5.5	121V~
GY - GY	0.35	11V~
YE - YE/WH - YE	25 each	157V~ each
OG - OG/WH - OG	2.7 each	21.8V~ each
BN - BN	0.7	10.7V~
BU - BU/WH	2.2	42V~
VT	High-Z	-

Would appreciate if someone could do measurements on unit, especially transformer to compare:
-Current/Real power consumption from line and line voltage
-Voltages on all tranformer taps (see above table)
-Resistance of primary windings (unplugged!)
-Bonus: Temperature difference on top of LM323 to environment (warmed up unit)

Measurements taken on my unit:
current draw at 240Vac
0.33A with fan,
0.25A without fan
My fan is 230Vac, It's connected directly to mains not through transformer as it should be when using 120V fan

RD - BK                   8.7 ohm             120V~
RD/WH - BK/WH   8.2 ohm             120V~
GY - GY                   0.3 ohm             9.1V~
YE - YE/WH - YE     25  ohm each   134V~ each
OG - OG/WH - OG   2.7  ohm each   18V~ each
BN - BN                   0.8 ohm             8.9V~
BU - BU/WH          2.0 ohm            40V~

LM393 temperature approx 29 degree C above ambient (top cover removed, fan running).
Ambient 20 degree C, actual temperature on LM393 - 49 degree C

Transformer is barely 5 degree C above ambient

Temperatures taken on unit which was powered for about 40 min.

[MiDi]

Thanks rolkinas!
The transformer seems fine, despite it is a Japan Version (TR-262 100/200VAC) instead of usual (TR-261 115/230VAC) 
This explains the ~20% higher voltages on secondary, the elevated current & real power draw (saturation of core) and the higher temperature on LM323 - a new TR-261 is on its way.

[rolkinas]

Wow, what a finding... 
What about fan? isn't it 100Vac ?

[MiDi]

Guess the fan is same on all units: Papst 8500 C 115VAC 16W
Line filter: Delta 115/230V 6A 2x0.55mH 15nF-X2 2x3.3nF-Y

[rolkinas]

Looks great, especially that gorgious PAPST fan 

[balage]

RD - BK	6	121V~
RD/WH - BK/WH	5.5	121V~
GY - GY	0.35	11V~
YE - YE/WH - YE	25 each	157V~ each
OG - OG/WH - OG	2.7 each	21.8V~ each
BN - BN	0.7	10.7V~
BU - BU/WH	2.2	42V~
VT	High-Z	-

Would appreciate if someone could do measurements on unit, especially transformer to compare:
-Current/Real power consumption from line and line voltage
-Voltages on all tranformer taps (see above table)
-Resistance of primary windings (unplugged!)
-Bonus: Temperature difference on top of LM323 to environment (warmed up unit)

Hi Guys,

My 237 was repaired before I've bought it. The fan is connected to the PCB as it is a 125V one.

-Current consumption on a 230V line is 320mA in idle. When I turn the output on (without load) it draws around 330mA.
-The LM323 is around 40°C hot, ambient is 25°C.
-Resistances and voltages are the followings:
RD - BK                   ?                 118V~
RD/WH - BK/WH     ?                 119V~
GY - GY                   ?                 9,4 V~
YE - YE/WH - YE      ?    each     132V~ each
OG - OG/WH - OG   ?    each     19V~ each
BN - BN                   ?                 8,7V~
BU - BU/WH            ?                 41V~

I haven't got time to disassemble the instrument that is why the resistances are missing.

[MiDi]

Thanks balage,
error was a japanese Transformer:

Thanks rolkinas!
The transformer seems fine, despite it is a Japan Version (TR-262 100/200VAC) instead of usual (TR-261 115/230VAC) 
This explains the ~20% higher voltages on secondary, the elevated current & real power draw (saturation of core) and the higher temperature on LM323 - a new TR-261 is on its way.

Got a new TR-261 and now power supply is fine (and RoHS compliant ):

Instrument consumes ~350mA and ~50W, without fan ~260mA and 34W @235V.

LM323: 40°C, dT=20.5K
MAC223 triac: 31.5°C, dT=12K

Transformer is drawing 230mA and 2.4W (real power) @241V without any load.

RD - BK	121V~
RD/WH - BK/WH	121V~
GY - GY	9.6V~
YE - YE/WH - YE	136V~ each
OG - OG/WH - OG	19V~ each
BN - BN	9.3V~
BU - BU/WH	41V~

Next is performance verification 

PS: If someone needs TR-262 - PM me

[aleruggeri87]

a new TR-261 is on its way.

Hi, sorry to reopen this but, where did you find it?
I fear, I burned mine yesterday... 

Thanks,
Alessandro.

[merox]

Hi, sorry to reopen this but, where did you find it?
I fear, I burned mine yesterday... 

Thanks,
Alessandro.

I think I should have a spare one somewhere in my lab, I'll have a look for it...

[DavidKo]

...
Hi, sorry to reopen this but, where did you find it?
I fear, I burned mine yesterday... 

Thanks,
Alessandro.

I think I have found the source (on ebay) - they do not have another one.

[aleruggeri87]

To follow up on my previous post:

I fear, I burned mine yesterday... 

At the end of the day the transformer overheated a lot, but it is still fine.
The 10 mF capacitor on the unregulated digital rail (C4) released some electrolitic, which went under the diode bridge CR1 and shorted it out.
The bridge exploded and the PCB was also damaged (see picture below).
I replaced these components and the two 100 nF capacitors near the bridge. It took about 1 month for the burning smell to pass, but everything work like before 

I also replaced the two 560 uF on the ±150V rails (C58-C59), as they are sitting really close to the 15k drop resistors for the ±15V rails, which get very hot during normal operation (~120 °C).

My unit (actually of the workplace) is a kei236, and power consumption from the 230VAC is 280 mA when ouput is on without any load current.