Author Topic: Fluke 5440B repair  (Read 5125 times)

0 Members and 1 Guest are viewing this topic.

Offline leighcorrigall

  • Frequent Contributor
  • **
  • Posts: 470
  • Country: ca
  • Nuclear Materials Scientist
Re: Fluke 5440B repair
« Reply #25 on: July 17, 2024, 01:38:35 pm »
If you have another 5440B that you do not mind opening, I would suggest swapping those cards, including the transformer, to see if you can isolate the problem.

This may be my shortest path to success, as I do have my 5440A, and they're very very similar, if not outright identical (other than the display board). In fact, I seem to recall I identified a couple other less know differences going by just the SMs, which may be documented in my 5440A thread.

In any case, I need to collect some guts to begin such an endeavor, as maneuvering two behemoths like these guys are are not for the faint of heart (or people liking theirs backs working OK). I need to wrestle them in a way where I could have both open side by side, and this in my cramped workshop will be tough.


I have reviewed the manual and seen enough teardowns to agree with you that there are very few differences between the A and B versions. The rear interface panels are different, too.

Be careful. The plastic handles are prone to breaking and should not be used for moving the unit. I move the instrument by sitting it on an office chair and moving it around my office when I need to work on it. One of my long-term goals is to replace these handles with steel equivalents so that I don't have to worry about breaking something off. My 5440B will be with me for a long time because it is tremendously reliable and has excellent specifications.
MASc, EIT, PhD
 
The following users thanked this post: Dr. Frank

Offline RaxTopic starter

  • Super Contributor
  • ***
  • Posts: 1110
  • Country: us
Re: Fluke 5440B repair
« Reply #26 on: July 18, 2024, 12:12:33 am »
One of my long-term goals is to replace these handles with steel equivalents so that I don't have to worry about breaking something off. My 5440B will be with me for a long time because it is tremendously reliable and has excellent specifications.

If you ever get to, I dunno, generate some cnc files (or other machining) you can share, or maybe have a few more made, I'd gladly get that if you're willing to share. Cover costs, etc, of course.
 

Offline leighcorrigall

  • Frequent Contributor
  • **
  • Posts: 470
  • Country: ca
  • Nuclear Materials Scientist
Re: Fluke 5440B repair
« Reply #27 on: July 18, 2024, 02:54:31 pm »
One of my long-term goals is to replace these handles with steel equivalents so that I don't have to worry about breaking something off. My 5440B will be with me for a long time because it is tremendously reliable and has excellent specifications.

If you ever get to, I dunno, generate some cnc files (or other machining) you can share, or maybe have a few more made, I'd gladly get that if you're willing to share. Cover costs, etc, of course.

If you are suggesting that I make a CAD model, I have already done so. Merry Christmas.

I have not formally shared them on EEVBlog because I still have to confirm the dimensions. It should be a near-identical copy.

Please let me know if I should pursue this. The material selection should also be correct. We will have to think about that.

EDIT:

According to the service manual, the handles are as follows:

Table 5-1 (5440B Final Assembly) and Table 5-2 (5440A Final Assembly)

Reference Designator, Description, Stock No., Supply Code, Part No., Quantity

MP 2, Corner Handle 8.75", 656199, 89536, 656199, 2
MP13, Corner Handle 8.75", 656199, 89536, 656199, 2

A private conversation with Rax confirms that these handles are interchangeable between the Fluke 5440A and the 5440B.

A single part (656199) can be used on all four corners because of the symmetry.
« Last Edit: July 23, 2024, 07:53:56 pm by leighcorrigall »
MASc, EIT, PhD
 
The following users thanked this post: Rax

Offline RaxTopic starter

  • Super Contributor
  • ***
  • Posts: 1110
  • Country: us
Re: Fluke 5440B repair
« Reply #28 on: July 25, 2024, 06:35:35 pm »
The other things I'd like to see as a 3D print file is the spacers used inside these 5440s (see pic). I'll dig up the part number and real name, but anyone having opened one of these has stumbled upon these little knoblets. They become very brittle with time, and I have not run into the situation where I'd pull them with "no victims" in a new unit having landed at my desk.

I'll give a shot to making some 3D files for this and see what comes out.
 

Offline Tony_G

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: us
  • Checkout my old test gear channel (link in sig)
    • TGSoapbox
Re: Fluke 5440B repair
« Reply #29 on: July 25, 2024, 08:13:15 pm »
Random question Rax, did you replace/check the relays? One of my 5440s Fluke rejected for cal due to relays and I've never got around to addressing.

Just wondering,

TonyG

Offline RaxTopic starter

  • Super Contributor
  • ***
  • Posts: 1110
  • Country: us
Re: Fluke 5440B repair
« Reply #30 on: July 25, 2024, 08:18:41 pm »
Random question Rax, did you replace/check the relays? One of my 5440s Fluke rejected for cal due to relays and I've never got around to addressing.

Just wondering,

TonyG

Tony,
Not really, no. I've just ran some of the corresponding diagnostics and I think my next step will be to swap modules to localize the issue (as I have this other 5440A that works correctly).
On checking the relays as its own thing, I am not entirely sure how I'd ago about that, need to think about it and figure it out. It's a good recommendation though.
 

Offline Tony_G

  • Frequent Contributor
  • **
  • Posts: 967
  • Country: us
  • Checkout my old test gear channel (link in sig)
    • TGSoapbox
Re: Fluke 5440B repair
« Reply #31 on: July 25, 2024, 11:49:41 pm »
I wouldn't have bothered getting the unit calibrated but I live down the road a bit from Fluke so I could drive the unit to their location rather than shipping it - I'm not sure how they determined that the relays were an issue (metrology, volt-nut noob...) but I haven't tried some of the suggestions of cleaning the relays.

Interested to see what you discover.

TonyG

Offline leighcorrigall

  • Frequent Contributor
  • **
  • Posts: 470
  • Country: ca
  • Nuclear Materials Scientist
Re: Fluke 5440B repair
« Reply #32 on: August 02, 2024, 06:03:12 pm »
The other things I'd like to see as a 3D print file is the spacers used inside these 5440s (see pic). I'll dig up the part number and real name, but anyone having opened one of these has stumbled upon these little knoblets. They become very brittle with time, and I have not run into the situation where I'd pull them with "no victims" in a new unit having landed at my desk.

I'll give a shot to making some 3D files for this and see what comes out.

If you keep the parts, they can be glued back together with epoxy and care. Once dry, you can use a razor blade to clean off the excess. I think I have a CAD file for them somewhere, as I was also concerned about them breaking. Unfortunately, I have no time to print them right now, but if you remind me after August 12th, I can see if I can fabricate them with a high-quality printer. You have my email.

I think I also saw a talk about calibrating this unit. I highly recommend that you do not calibrate a unit until you have had plenty of experience with it (> 6 months of operation and use). If there is a subtle fault, you may be wasting your money. Believe me, I made the mistake of being hasty a while ago with an Advantest R6581T.

MASc, EIT, PhD
 

Offline RaxTopic starter

  • Super Contributor
  • ***
  • Posts: 1110
  • Country: us
Re: Fluke 5440B repair
« Reply #33 on: August 03, 2024, 10:21:24 pm »
The other things I'd like to see as a 3D print file is the spacers used inside these 5440s (see pic).

I think these are Fluke 421776 ("DORCAS?..."). So I made this in Fusion.
 

Offline RaxTopic starter

  • Super Contributor
  • ***
  • Posts: 1110
  • Country: us
Re: Fluke 5440B repair
« Reply #34 on: August 04, 2024, 12:38:46 am »
This afternoon I finally got the earthmover (...) to carve out some room at the bench, enough to have two of these behemoths open side by side. Not sure I'll be able to get the car we usually garage inside the garage tonight...

So, I was able to swap modules from the known good unit to the issue unit such that I'd isolate the issue. And this far tests are a bit confusing. Per SM (p.4-14), given the error message, A8, A9, and A4 are prime candidates for issues. A4 and A8 A9 seem to change nothing, essentially, so I'm deeming them good. Swapping A9 A8 doesn't fix the issue, but instead returns a different error. Actually, errorS! Initially, there was something about, I think, "inside guard" (it was one time and I didn't write it down immediately), and now I'm consistently getting "OUTPUT LIMIT FAULT. OUTPUT OVER VOLTAGE." But, this was upon me asking it to output 10V, and I didn't see it outputting more, so that's a bit weird. It seems there's some false flag errors triggering or something.

I'll dig down deeper, but if anyone sees any sense in the above, I'm all ears. All I can conclude this far is that the issue may have to do with A9, though a known good  module didn't fix it. For anyone asking, the module in both units is a "5440A" module, so I don't suspect any updates to the "B" unit modules make the swapped "A" unit unfit.
« Last Edit: August 04, 2024, 01:09:55 am by Rax »
 

Offline RaxTopic starter

  • Super Contributor
  • ***
  • Posts: 1110
  • Country: us
Re: Fluke 5440B repair
« Reply #35 on: August 04, 2024, 12:42:41 am »
With the replacement A9 A8 board, if I try to output a "previously error condition triggering output" (say, 100V), I'm still getting the "UNDER VOLTAGE" error. If I try to output, say, 10V, I'm getting an "OVER VOLTAGE" error. Outputting 10V was successful with the original A9 board. So there seem to be new errors of sorts with the replacement board.

I think I'm just going to have to take a very close look at the trouble unit's A8.
« Last Edit: August 04, 2024, 01:44:28 pm by Rax »
 

Offline RaxTopic starter

  • Super Contributor
  • ***
  • Posts: 1110
  • Country: us
Re: Fluke 5440B repair
« Reply #36 on: December 03, 2024, 10:10:27 am »
This project's still active. I've not had a ton of time to spend on it, but during the winter holidays there may be some additional time so I am getting this back to the front of the docket. I am hoping for some input on what's going on here and where to head next.

A quick and probably partial recap and status report.

          1. The unit initially was completely dead, which I tracked to A19 not being supplied with 120VAC. Which meant all outguard modules were essentially off.

          2. Once the above was corrected, the unit started largely working. Except that, when I attempt to output more than about 60V, the unit initially successfully outputs the requested voltage (say, 70V), but just for a few moments, then it drops to about 57V, which triggers the fault mode "OUTPUT LIMIT FAULT. OUTPUT UNDER VOLTAGE."

          3. My next step at this point was to swap modules (A4, A5, A13/12, and I may have tried others) from my fully working (successfully repaired) 5440A with inconclusive results.
 
Then I went ahead and tracked the symptoms above (the lower than set output voltage) backwards through the output board and back to A13 FILTER B board. Technically, I think the issue resides on what Fluke calls the "HV Loop," and they have some pretty contorted and not illuminating (at all, to me) steps in section 4-23.

I've done this by going by section 4-19, where I've applied a DCV proportional to the set output vs. full scale between A4 (OUPUT board, NOT A5, as per Service Manual, so please note the error in the manual) A4TP5 and A4TP13. For instance, a set 70V out would want this applied DC to be about 112mV. I did this with my HP8904A, which has the great convenience of being able to key-input voltage settings directly. This step is in order to disable the monitoring system, so an out-of-bounds condition (more than 5% difference between the output and the set voltage) would not reset the unit and further troubleshooting would be possible to make at length.

Under these conditions, the events at #2 above occur and stay active past the few seconds it'd take the unit to actual its reset. To me, this confirms the monitoring is operating correctly and the issue originates with the generation of the HV.

       4. I looked at +TSV and -TSV by probing A13TP1 to A13TP2, and I am seeing the voltage go up to about 5V (as Vout stabilizes to 70V), then dropping to about 2.15V (as the voltage out is falling from the 70V set to the 57V is settles for), and then going back up a bit to something like 2.5V or so. I do see some ripple riding on top of that (per my Fluke 189), though it varies pretty wildly (it starts low, maybe mV, but as the fault occurs, it jumps to maybe 250-300mV, then goes down in the tens of mV and stays there). I am not 100% sure what I should be seeing or looking for at A13TP1 and A13TP3, so I don't have a satisfying analysis of my own of what this means.
       5. I am not seeing anything significant happening over R20 (the +TSV .33 ohm) when the fault occurs. Barely in the mV range if even that. So there seems to be no protection actuation on the HV that could be causing this.
 

Offline MegaVolt

  • Frequent Contributor
  • **
  • Posts: 931
  • Country: by
Re: Fluke 5440B repair
« Reply #37 on: December 09, 2024, 02:09:14 pm »
As I understand it, the problem is with the voltage drop at the output in the range of 275V.

First, take a multi-channel oscilloscope (4 channels) and study the picture "Figure 2-8. Sample String PCA, Simplified Circuit Diagram"

Put the oscilloscope probes in many points of the device at the same time and see in which node the first voltage drop occurs.

1. DAC output.
2. PREAMP SUMMING JUNCTION point
3. PREAMP PCA OUT
4. OUTPUT

Looking at these graphs, you can understand which of the nodes breaks through first. This will greatly narrow the search area.
 
The following users thanked this post: Rax

Offline RaxTopic starter

  • Super Contributor
  • ***
  • Posts: 1110
  • Country: us
Re: Fluke 5440B repair
« Reply #38 on: December 17, 2024, 05:05:50 am »
Thank you for thinking on this with me, MegaVolt.

Yes, the 275V, but also the 1100V. So everything over about 60V seems to trigger this fault state. Apparently the unit is unable to output more than the about 57V + 5% for fault to be triggered, but that seems to be the case.

Right now, I'm trying to get a four channel digital scope at my bench (it just so weirdly happens I don't have a fully working one right now, just one that's a project...). Before I do that, I'll be trying a few other things, but if they go anywhere, I'll post here with results.
 

Offline RaxTopic starter

  • Super Contributor
  • ***
  • Posts: 1110
  • Country: us
Re: Fluke 5440B repair
« Reply #39 on: January 06, 2025, 01:01:52 am »
Well, project's done... Hugely gratifying and a gianormous project in a few different ways for sure.

I hope I'll not miss anything, but this unit was inflicted with more separate and possibly disconnected issues than anything I've seen before. Attempt at a summary:
  • The HV issue tracked back to A13, where HV gets cooked. That's because the TS +/- lines coming from the digital drives in seemed correct. Almost accidentally, I discovered that voltages over 876V were output successfully, so that put the TS1, TS2, and TS3 triac drives under suspicion (TS4 deemed good). Replacing the corresponding triacs (Q2-4, 5-7) restored operation of the corresponding range for TS2. No luck on TS1 and TS3.
  • Given A13 seemed to be the origin of all evils, I went ahead and replaced the optocouplers on the faulty paths. Also, all bridges (including TS4, while I had my hands dirty). This way, those paths had essentially all new parts in them. No luck.
  • Further investigation revealed the TS AC inputs into those triac drives were present at A11 and missing at A13. Stripping the unit bare revealed that A11/J3 and A13/J6 were not talking to each other. Taking a magnifier to A3 (MB) revealed the MB card edge connectors had broken pins. This included TS5 (magnifier was a literal must to figure this one out) which wasn't a suspect before, but a perfect tracking of both + and - sides may have been explained by this third AC path being broken. IMHO, the mechanics of this is a design flaw - the massive transformer attached to A11 should be fastened to the A11 board with plastic washers, or something else that would provide a tiny bit of leeway. The massive transformer "manhandles" A11 during insertion into the MB connectors during servicing; or any shipping event, for that matter (even if the transformer is fastened to the chassis, ultimately). Not very smart of you, Fluke.
  • Once the edge connector was fixed, the unit had one issue left, namely an "INSIDE GUARD FAULT. CHECK GUARD CTRL BUS" error, which became more and more persistent.
  • This was a tricky one, and close examination of the GPOP line revealed some odd handling of the +5 TS AC signal by A10/U13. Replacing C31 fixed this in a climax I have not frankly anticipated.
Exceedingly happy camper. Thanks again, MegaVolt, for your kind pitching of thoughts on here. Effusive kudos for your availability.
« Last Edit: January 06, 2025, 09:39:58 am by Rax »
 
The following users thanked this post: Dr. Frank, MegaVolt

Offline RaxTopic starter

  • Super Contributor
  • ***
  • Posts: 1110
  • Country: us
Re: Fluke 5440B repair
« Reply #40 on: January 06, 2025, 01:22:02 am »
Lessons learned on the 5440 series?
  • A11 has a slight design issue, where, imho, the transformer needed a different type of fastening to the board. Take a magnifier at the edge connector if symptoms point to issues there, especially if some HV ranges work and others don't.
  • Recap. Not indiscriminately (though, really, up to you), but suspect caps being guilty of symptoms you observe.
  • Triacs on A13 are suspect. To my surprise, more suspect than the optocouplers (which I've seen go bad before), so examine (both of) them closely.
  • Dallas SRAMs tend to go bad. Check the data on them and replace with known good chips with correct bits on them. Data is available from different sources, here and elsewhere
« Last Edit: January 06, 2025, 09:41:31 am by Rax »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf