Datron 4000A getting at test points??

• Datron 4000A getting at test points??
  Posted by essele on 11 Mar, 2019 09:01
• Hi,
  
  I've managed to get my hands on a Datron 4000A with a FAIL 7 error, which indicates a problem with the 50V supply. There are a couple of burnt looking resistors on the board, so I'm going to replace them anyway but I'd like to be able to test more thoroughly (btw the +/-15V supplies appear to be ok as I know that's the fault at least one other person had.)
  
  My question is really about how to make use of the test points ... this board in particular takes its 15V supplies from the motherboard connector, therefore you can't properly test it if it's not plugged into the motherboard, and then you just can't get at the test points???
  
  I managed to test the PSU board using test clips but it was a complete faff and there's just not enough space on the 50V board.
  
  I'm probably being completely dense, but how has everyone else seemingly managed to test this?
  
  Cheers,
  
  Lee.
• #1 Reply
  Posted by coromonadalix on 11 Mar, 2019 09:21
• you could solder some small wires onto the pcb to the tests points ? while being sure they touch nothing when the 4000 is powered ?  or connectors extenders ??
• #2 Reply
  Posted by Johnny10 on 11 Mar, 2019 11:42
• I have worked on my 4000A quite a bit and board extenders would be fantastic !!!
  
  But I did as coromonadalix suggested, attach some wires to the test points and carefully reinsert the board.
  I sometimes had 8 or more wires extending above the PCB boards, works in a pinch.
  
  I had the same problem as you with Fail 7.
  The 15V supply is critical.
  After replacing all tantalums on power supply board and -15V Voltage regulator(MC7915CT), problem solved.
  
  Also make sure that power supply board is seated properly. Had a problem with that also.
  
  xdevs has the power supply schematic on his site.
• #3 Reply
  Posted by essele on 11 Mar, 2019 21:57
• Thanks for the comments .. I've made a little more progress, but it now seems like I'm going to need to resort to soldering test wires on.
  
  Basically one of the thermistors on the positive heatsink assembly was faulty, and was actually leaking from the metal tab, so I was seeing -15v correctly on J1-2, but +28v on J1-1. I've unsoldered the thermistor and put a 190k resistor in it's place and all seems more normal (now matches the negative heatsink thermistor!)
  
  Does anyone know what a TS4-85 is? I know it's an NTC, but I can't seem to find any references anywhere??
  
  I'm still seeing a FAULT 7, and just looking at the connector voltages I can see that the +50V rail is perfect, but the -50V is only -42V, so lots to check, but at least I've narrowed it down to the negative rail.
  
  Cheers,
  
  Lee.
• #4 Reply
  Posted by Johnny10 on 11 Mar, 2019 22:27
• I find NTC Thermistor 85 degree Manufactured by Microtherm TS485
• #5 Reply
  Posted by essele on 11 Mar, 2019 22:56
• Thanks Johhny10 but where are you finding that? Even with all of that detail I still can’t find anything???
  
  Further progress though ... TP11 was also low, so I started to suspect either the op amp or the reference. Applying a small constant current voltage to the reference showed that the positive one was a good 2.45V, but the negative one was only 2.1V, but out of circuit it was fine! So I’m now fairly convinced that C17 (1uF tantalum) is the culprit... I’ve run out of time now, so hopefully I can get this sorted tomorrow.
  
  I’m slightly worried about the previous issue as there was 28V on a motherboard signal which was supposed to be 0V or -15V.
  
  Cheers,
  
  Lee.
• #6 Reply
  Posted by Johnny10 on 11 Mar, 2019 23:49
• I have various resource material collected from my own repairs.
  I have three Datron 4000 units: 2 working 1 for parts.
  
  
  
  
  

  ---

  
• #7 Reply
  Posted by essele on 12 Mar, 2019 09:02
• Thanks.
  
  Interestingly I've not managed to find any data or references at all to the TS4 range. I have found the TS3 range, but unfortunately they are no longer manufactured (post 2016) due to source materials and production line equipment challenges.
  
  The interesting this about these is they have a highly specific transition region at the rated temperature, so I'm really not sure what to do about a replacement.
  
  To quote the datasheet...
  
  Within this transition region the devices typically exhibit a negative temperature coefficient ranging from 40% per °C to 200% per °C. Below the transition region they exhibit a linear “thermistor” characteristic of -2 to -4%/°C.
  
  The only thing I can find at the moment is the 75 degree version, so I've ordered one anyway, hopefully it won't be an issue.
  
  EDIT: Actually it may work with a simple normally open thermal switch ... I'll have a look around.
  
  Cheers,
  
  Lee.
  
  
  
• #8 Reply
  Posted by coromonadalix on 12 Mar, 2019 11:00
• Is there any complete schematics out there ?  this is a collection   of found items on the web.
  
  found this :    ko4bb  has them too
  
  http://www.mediafire.com/file/d4lrsde7zv0cyry/Datron4000A.rar        27mb
  
  
• #9 Reply
  Posted by essele on 12 Mar, 2019 23:19
• I’ve been using the schematics here: https://xdevs.com/doc/Datron/4000A/Datron_4000_Fail_7.pdf
  
  Some more progress ... I was completely wrong about the cap, I’ve actually tracked the issue to the negative 50v current limit. I’ve disconnected one leg of D27 so the current limit isn’t active and it all works fine (well almost!) ... so something in that current limit section isn’t right. Everything tests ok, so I think I’m going to have to go back to test wires again. (I am assuming there isn’t actually a large current flow, I should probably check, but nothing seems overly warm.)
  
  The voltage and resistance ranges all seem to work ok and are reasonably well calibrated. The 1A current range doesn’t work, but the others are fine. I need to do some work on the front panel keys as they are a bit hit and miss.
  
  Good progress for this evening!
  
  
• #10 Reply
  Posted by Johnny10 on 12 Mar, 2019 23:34
• A couple of points to remember:
  
  The calibrator has built in routines that shut the output down making it difficult to discover certain problems.
  
  I replaced the same IC on the front keyboard PC on all my units. M4 Dual Binary 1 of 4 Decoder MC14555BCP . It is handy that they are all socketed.
  
  On all units, connecting to GPIB allowed me to access remotely all functions of the Calibrator without using keypads.
  
  After replacing all tantalum capacitors I had no further errors.
  
  
• #11 Reply
  Posted by essele on 13 Mar, 2019 11:18
• Thanks -- would be good to know about the keyboard IC. Some of the keys work, but some don't, some definitely need a clean as you can get them to work through different pressures.
  
  Do you remember what voltage your 15V rails were? ... mine are +/-14.6V which I was a little concerned about, but they are within 7915 specs. I've ordered all the tantalums for the 15V supply anyway so will replace them when they arrive. I don't think this is the issue though as the 15V rails don't play a part in the over current control, and everything is fine when that bit is disconnected.
  
  I'll also change the ones in the overcurrent limit, but they all test ok, so I don't really think it's them. I've also ordered all the transistors and I suspect they are most likely at fault -- if I understand the schematic correctly then some leakage could cause this to happen. I'm also a bit worried that this is a power-on transient issue that causes the over-current circuit to latch on -- more sophisticated tested needed!
  
  You don't happen to have a schematic for the current section do you? I get a few mA on the 1A range, it does change with setting, so hopefully something simple. The other current ranges are fine.
  
  Cheers,
  
  Lee.
  
   
  
  
  
  
  
  
• #12 Reply
  Posted by Johnny10 on 13 Mar, 2019 12:54
• I chased problems around different boards for hours.
  It was not until I changed all tants that the errors stopped.
  
  I also had a problem with leaving unit sit for a few months in between uses, turn it on and blow a few more tants. AArgh!
  
  Voltages on 480455 DC PCB Assy
  + 14.90
  - 14.90
  
  What board do you need more info on?
  
  This may be helpful

  ---

  

  
• #13 Reply
  Posted by Johnny10 on 13 Mar, 2019 15:58
• 91

  ---

  

  400448.pdf

  430448.pdf
• #14 Reply
  Posted by essele on 14 Mar, 2019 08:46
• Thanks Johnny10, they are all really useful -- I'll get to that once I've figured out this -50V issue.
  
  The final batch of tants should arrive today, so hopefully I can get them all replaced this evening.
  
  I didn't have a lot of time yesterday but I did check the voltage drop across the current sense resistor, and there does seem to be a power on spike well into the limit area on the -50V side, the +50V side spikes but is half the magnitude. But I only measured this with my multimeter (fast and graphing) as I didn't have time to satisfy myself about attaching my scope without isolation (and my isolated probe is x200, so not really the most use in this case.)
  
  If I'm understanding the schematic there is a 47u tantalum in series with 150R that looks like it should slow the reaction of the current limit by a few ms presumably to stop it triggering on short transients and the turn-on spike. So I think it leaves me with four scenarios...
  
  1. This is all tantalum voodoo and they just need replacing (which I will do anyway.)
  2. The current limit is working fine, and the turn-on spike is indeed much bigger than it should be - so something else is wrong.
  3. The current limit is triggering too quickly (most likely the cap or associated transistor.)
  4. The current limit has a fault and is on all the time anyway.
  
  If recapping doesn't work, then I'm going to try shorting out the current sense resistor to see if the current limit is being triggered this spike or not ... so that will prove or disprove number 4. Hopefully it proves it as I have all the parts to replace everything in that section now!
  
• #15 Reply
  Posted by essele on 14 Mar, 2019 21:23
• Aaarrrggghhhh!
  
  I'm pleased to say the +/-50V rails are now working and my FAIL 7 error is gone, but I'm not particularly pleased with how long it took and how stupid I am!
  
  So I shorted the current sense resistor and yet the problem still persisted, so that was good, it meant the issue was squarely with the current limit circuitry and not anything to do with turn-on spikes etc.
  
  So I replaced the two tantalums, no change. Replaced the three transistors (one at a time), no change. I replaced the 1N4148's, no change, I removed and re-tested the two zeners, they seemed fine. I retested all the resistors, they were fine. I swapped the ceramic cap.
  
  So I'd changed or tested every single component and it still wasn't bloody working!
  
  Because I'd removed two of the heatsink assemblies I had a bit more room, I could start putting test clips on various components, and I eventually found that the 5.2v zener anode, which should have been at -75V (ish) was at -35V ... how can that be? It's directly connected to the darlington and I can probe that on the connector and it's -75V??
  
  Anyway, it turns out that it's not directly connected to the darlington, because a small bit of the copper track had come away, so I've just spent hours (and a fair few quid on components) to solve a bloody track defect.
  
  Sigh. Lesson learnt.
  
  Before and after pictures attached for embarrassment. I've also included a picture of my multimeter with 1v output set, this was only a few minutes after power-up (ignore the historical stats) ;-)
  
  Now to move on to the 1A current range problem ... although I'm going to save that for tomorrow and crack open a beer instead.
  
  Thanks for your help Johnny10.

  ---

  

  

  
• #16 Reply
  Posted by coromonadalix on 14 Mar, 2019 21:26
• nice finding for the pcb trace,  but i would have gone from thru hole to thru hole,  i think the pcb trace may fail again ...
• #17 Reply
  Posted by Johnny10 on 14 Mar, 2019 21:39
• I did the same thing with a HP3457A, checked and changed a number of parts and still didn't work... then flexed the board and it worked???
  Split in a trace .  AArgh !
  
  Good Find.
• #18 Reply
  Posted by essele on 14 Mar, 2019 22:51
• Hi coramonadalix - yes I did think that myself, and then got overcome with the joy of fixing it quickly. I was also considering putting some epoxy on it, but you’re probably right and I should do it properly.
  
  Johnny10 - do you have the other half of the schematic and board layout for the I/Ohms board? I couldn’t help but have a quick look, the 0R1 is fine, so I think it’s either the relay or the relay control signals.
  
  Also ... is there a way to change from 2 wire to 4 wire resistance? I notice that calibration can be set for each, and I’m assuming it should display accuracy for each, but I can’t find a way to switch?? (1k 4-wire is a bit off, but everything else is great [or my meter is off!!])
• #19 Reply
  Posted by Johnny10 on 14 Mar, 2019 23:20
• What happened to going for a beer?
  
  Are you getting OL error on Current?
  
  I believe pressing OHMS key and SENSE key puts it in 4 wire mode.

  ---

  

  
• #20 Reply
  Posted by essele on 15 Mar, 2019 08:07
• Don't worry ... the beer was consumed!
  
  I'm not getting OL, I'm getting a much smaller output than expected, it's actually just over 50 times smaller than it should be (1A gives me 19.926mA, 100mA give me 1.994mA) so I have a strong suspicion that it's using the 5R shunt rather than the 0.1R shunt.
  
  I'm slightly confused because that would imply that RL1 isn't switching properly (which I can believe), but also RL3 would need to be closed when it's supposed to be open, and I'm fairly sure it's ok when using the other ranges, so may well be relay driving problems.
  
  Is there another schematic covering the relay drivers and logic?
  
  I also just managed to get the 2-wire/4-wire switching as you suggested ... it seems to switch every relay in the box in sequence rapidly several times when you change the setting. Is that normal, probably lasts best part of a second, sounds horrible? I'll record it this evening.
• #21 Reply
  Posted by Johnny10 on 15 Mar, 2019 12:54
• Switching to Sense makes the repeated clicking sound on my unit also.
  

  ---

  
• #22 Reply
  Posted by essele on 15 Mar, 2019 18:57
• Hmmm .. I took the board out again and soldered some test wires onto the relay coils so I could see if they were being driven (for RL1) and when I put it back it was working fine!
  
  Desoldered the wires, and it stopped working!
  
  Soldered them back on, and the left hand relay coil wasn't being driven, traced the tracks and they went to the sub-pcb, so I've just reseated it and it all seems to be working fine (if not slightly low in current reading, 100mA on the 1A range is reading 99.966mA ... which is below the 99.970mA min, but it hasn't warmed up yet, so hopefully this will recover.)
  
  So I'm hoping it was a badly seated board, but it could well be another dodgy track or a dry joint ... I will test for a while now.
  
  2w and 4w both read spot on so I think everything is functioning now ... just need to look at these dodgy buttons.
  
  Thanks for all your help again.
• #23 Reply
  Posted by Johnny10 on 15 Mar, 2019 19:21
• As I said before I had problems with boards fully seating.
  My own error.
  
  Great all is up and working!
  Are you going for calibration?
  
  Another member suggested I buy an 8 1/2 digit calibrated DMM instead.
  
  Got to save the pennies for that one.
  
  Behind the keys pic.
  
  Carefully pry off the key top and underneath there is simply a spring.
  Key snaps on and off to this.
  

  ---

  
• #24 Reply
  Posted by essele on 15 Mar, 2019 20:31
• I was thinking about seeing if I can find somewhere to get it calibrated without having to remortgage, I do have a 34470A that's about 9 months old and still in factory calibration, it's "only" 7.5 digits, and actually 7 months on, I probably would have bought a Keithley rather than the Keysight, but it is showing me that this 4000A is doing ok, but is maybe very slightly off .. although what do you trust more?
  
  But them I'm also running it with the top off at the moment, so I probably have some more test to do before I decide what to do.
  
  Thanks for the keys pic ... that's tomorrows job.