EEVblog Electronics Community Forum

Electronics => Repair => Topic started by: motocoder on July 22, 2015, 03:25:38 am

Title: 34401A test failures (many)
Post by: motocoder on July 22, 2015, 03:25:38 am
I have a 34401A that I picked up off of eBay. When I run the self test, it is failing tests: 608, 609, 610, 611, 612, 613, 615, 617, 618, 619,621, and 623.

Test 608 is "Serial config readback failed". I am guessing that might be leading to other tests failing.

Any advice on how to troubleshoot this further?
Title: Re: 34401A test failures (many)
Post by: poot36 on July 22, 2015, 03:56:26 am
If you have the service manual it should say what to do depending on the error codes you are getting.  Try reseting the device if it is possible.  It may have been configured for remote control through the serial port.
Title: Re: 34401A test failures (many)
Post by: motocoder on July 22, 2015, 04:08:41 am
The service manual lists the error codes, and gives a very brief description of what they mean, but does not seem to give any troubleshooting steps. I don't think the device is in remote mode because it responds to front terminal commands just fine.

I've checked a few power supply voltages:

U552 (+18V)
U551 (-18V)
U751 (+5V)
U553 (VCC)

The first 3 were all within spec. VCC was +5V, but I haven't found in the manual if that's the correct voltage yet (but I suspect so).

Title: Re: 34401A test failures (many)
Post by: poot36 on July 22, 2015, 04:41:49 am
Is there some type of battery backed memory chip that could be holding the configuration of the serial port and now that the battery has probably gone flat it is corrupted or erased?  Do you have a link to the service manual?
Title: Re: 34401A test failures (many)
Post by: motocoder on July 22, 2015, 05:18:35 am
Is there some type of battery backed memory chip that could be holding the configuration of the serial port and now that the battery has probably gone flat it is corrupted or erased?  Do you have a link to the service manual?

Service Manual: http://www.keysight.com/main/redirector.jspx?action=ref&cname=EDITORIAL&ckey=765594&lc=eng&cc=US&nfr=-11143.0.00 (http://www.keysight.com/main/redirector.jspx?action=ref&cname=EDITORIAL&ckey=765594&lc=eng&cc=US&nfr=-11143.0.00)
User Manual: http://www.keysight.com/main/redirector.jspx?action=ref&cname=EDITORIAL&ckey=765584&lc=eng&cc=US&nfr=-11143.0.00 (http://www.keysight.com/main/redirector.jspx?action=ref&cname=EDITORIAL&ckey=765584&lc=eng&cc=US&nfr=-11143.0.00)

There are some folks here, like Dr. Frank, that are very familiar with these meters. Hopefully one of them will check in. In the meantime, I will go through the "Theory of Operation" section and see if I can figure out more.
Title: Re: 34401A test failures (many)
Post by: motocoder on July 22, 2015, 05:26:04 am
Ah man, there's nothing worse than seeing that some monkey has tried to repair something and failed. You then enter this no-mans land of not knowing what was the original failure versus what modifications has the monkey made before he gave up.

In this case, I see some monkey poo on this unit, specifically he has written in marker on the PCB, and there is a black bodge wire under the board going from one of the vias under U102 to what looks like some large trace (ground or power of some sort, probably)...
Title: Re: 34401A test failures (many)
Post by: motocoder on July 22, 2015, 05:30:08 am
Looks like U102 is the resistor network, which I suspect is the most expensive part ($191).  :-[
Title: Re: 34401A test failures (many)
Post by: Dr. Frank on July 22, 2015, 05:38:05 am
This failure was discussed before. I don't find it anymore, it was either here, or on Volt-nuts.

This failure is located somewhere in the daisy-chain shift registers, which run all over the analogue circuitry..see also block diagram on page 9-7 in the service manual.

The µP clocks in serially the steering signals for the different circuits, and afterwards reads back this information from the end of the chain, to test the serial bit stream.
This read-back fails obviously, so the daisy chain is broken somewhere.

You can easily track that in the schematics, and also on the PCB.. I assume you may find easily the failing shift register, or maybe the opto-couplers.

Frank
Title: Re: 34401A test failures (many)
Post by: motocoder on July 22, 2015, 06:33:40 am
This failure was discussed before. I don't find it anymore, it was either here, or on Volt-nuts.

This failure is located somewhere in the daisy-chain shift registers, which run all over the analogue circuitry..see also block diagram on page 9-7 in the service manual.

The µP clocks in serially the steering signals for the different circuits, and afterwards reads back this information from the end of the chain, to test the serial bit stream.
This read-back fails obviously, so the daisy chain is broken somewhere.

You can easily track that in the schematics, and also on the PCB.. I assume you may find easily the failing shift register, or maybe the opto-couplers.

Frank

Thanks, Dr. Frank. Can you tell mw if there is a patch wire on the bottom pf the PCB in the original, unmodified unit? The Pcb says Rev D if that is relevant. I suspect it was added bynthe previous owner, but thought I would check with you before removing it.
Title: Re: 34401A test failures (many)
Post by: Dr. Frank on July 22, 2015, 08:00:58 am
Sorry, don't know.
My instrument is 1st version, 1993, and has no cable, afair.
Frank

PS: I opened my instrument, PCB is Rev A, 1992 components, no cable on bottom side!
Title: Re: 34401A test failures (many)
Post by: KJDS on July 22, 2015, 08:40:54 am
I'll try and remember to look when I take some of these apart.
Title: Re: 34401A test failures (many)
Post by: motocoder on July 22, 2015, 04:00:45 pm
Here's a picture of the bodge wire. Doesn't look to me like typical Agilent manufacturing quality, so I'm thinking it was added by the previous owner.

Title: Re: 34401A test failures (many)
Post by: radioFlash on July 22, 2015, 05:13:01 pm
I had the 34401A which was fixed by replacing U400 (an AD706 opamp). In my case, I didn't have any error codes, but voltage readings were drifting. Here are photos of a Rev B and Rev D board.
Title: Re: 34401A test failures (many)
Post by: radioFlash on July 22, 2015, 05:13:48 pm
Here's a view of the bottom side of a Rev D board. No bodge wire.
Title: Re: 34401A test failures (many)
Post by: motocoder on July 22, 2015, 05:26:10 pm
Here's a view of the bottom side of a Rev D board. No bodge wire.

Excellent. Thanks, everyone for the pics. I will remove the wire and investigate the shift register issue that Dr Frank mentioned.
Title: Re: 34401A test failures (many)
Post by: Corporate666 on July 22, 2015, 05:57:10 pm
Are the different revision boards interchangeable?  I'd guess so since there aren't different P/N's for displays and such for each rev of the main board.

I have a Rev A in my non-working 34401, and I got a Rev D board pulled from a working unit for cheap $$.  Hoping I can swap out my Rev A board for the Rev D.
Title: Re: 34401A test failures (many)
Post by: radioFlash on July 22, 2015, 09:53:08 pm
They should be--I've swapped the display between my two units without any issues.


Sent from my iPhone using Tapatalk
Title: Re: 34401A test failures (many)
Post by: Vgkid on July 22, 2015, 10:12:14 pm
It is interesting looking at radioFlash's board, mine has several numbers scrawled in different places across the board.
Title: Re: 34401A test failures (many)
Post by: motocoder on July 23, 2015, 01:29:53 am
After removng the patch wire, the meter gives error 604 and stops the test. I don't hear any relays clicking, so I don't think it is running all the tests anymore. Also, the front panel shows "rear" as if the meter is switched to the rear inputs (it isn't).
Title: Re: 34401A test failures (many)
Post by: motocoder on July 23, 2015, 02:25:58 am
Ok, I see why there was a patch wire there. Whoever had a go at repairing this before me lifted one of the very fine traces on the back side of the board under the thin fim resistor network. It was very obvious where this trace was going, so he had just put that wire in there to fix the broken trace. I put a more proper patch wire - (minimum length, more appropriate gauge) - and now we're back to the errors I listed at the start of the thread.

Also, now that I have a chance to put some magnifying goggles on (my vision is failing badly, so I didn't catch this without them earlier), it appears that he's replaced a number of components on the board. It appears he has replaced U153, U103, U105, U106, U104, and possibly Q150, Q151, and C113 (may have just disturbed these in removing the neighboring ICs).
Title: Re: 34401A test failures (many)
Post by: motocoder on July 24, 2015, 10:53:40 pm
As Dr. Frank suggested, I traced the serial clock and data signals through the chain. When I get to U150, I can see serial clock and data going in, but do not see serial data (SERDATA2) exiting the chip on pin 4. I also checked pin 25 on U101-A, where SERDATA should be an input, and do not see a valid signal there either.

Now I think this leaves three options. Either previous owner has managed to short that line to something, the next chip in chain (U101) is so damaged that it's serial input line is a low impedance, or U150 is bad. I do hear the relays clicking as I change modes and ranges. So U150 is at least partially working. A replacement is $18 from Keysight. U101 does not seem to be orderable from them.

Suggestions?
Title: Re: 34401A test failures (many)
Post by: Kleinstein on July 25, 2015, 01:27:01 pm
One option may be opening the line from U150 to U101, just to see which chip is broken. Also double check that the serial clock is good at U150.
Title: Re: 34401A test failures (many)
Post by: motocoder on July 25, 2015, 03:26:51 pm
One option may be opening the line from U150 to U101, just to see which chip is broken. Also double check that the serial clock is good at U150.

Yes, I toyed with lifting the pin on U150, but was a bit intimidated by the fine pitch of the pins and thoughts of lifting a trace. Let me see if there is some other way to do it.
Title: Re: 34401A test failures (many)
Post by: motocoder on August 01, 2015, 09:54:49 pm
I finally got past the Agilent part ordering hurdles and got a replacement for U150. I can now see the serial data out from this chip, so that's good news. The bad news is that things are still not working correctly.

I checked the next chip in the chain, U101, and although I can see serial data and clock going in, there is no serial data coming out. I lifted the SEROUT pin (24) on this chip to verify that it wasn't something downstream shorting it out, and I still see no output on this pin. So it appears that U101 is also bad. Note that this chip is physically close to the other chips that were replaced on this unit by a previous owner.

The bad news is that U101, an analog switch with Agilent part number 1SK6-0001 is no longer available from Agilent.

Does anyone have a source for this part?
Title: Re: 34401A test failures (many)
Post by: radioFlash on August 02, 2015, 03:48:13 pm
There's one listing on ebay. I haven't used this vendor before, so standard caveats apply.

http://www.ebay.com/itm/1SK6-0001-DIP-8-/361222471615?pt=LH_DefaultDomain_0&hash=item541a897fbf (http://www.ebay.com/itm/1SK6-0001-DIP-8-/361222471615?pt=LH_DefaultDomain_0&hash=item541a897fbf)
Title: Re: 34401A test failures (many)
Post by: motocoder on August 02, 2015, 04:36:56 pm
There's one listing on ebay. I haven't used this vendor before, so standard caveats apply.

http://www.ebay.com/itm/1SK6-0001-DIP-8-/361222471615?pt=LH_DefaultDomain_0&hash=item541a897fbf (http://www.ebay.com/itm/1SK6-0001-DIP-8-/361222471615?pt=LH_DefaultDomain_0&hash=item541a897fbf)

Yes, I saw that one. I sent him a "Best Offer" yesterday, but no reply yet. Thanks.
Title: Re: 34401A test failures (many)
Post by: chihaxinh on August 05, 2015, 02:41:49 am
If you have part and replace U101 , what do you think error can relative with U102 ?

How to check U102 dead or alive when DMM power on ?
Title: Re: 34401A test failures (many)
Post by: motocoder on August 05, 2015, 02:46:12 am
If you have part and replace U101 , what do you think error can relative with U102 ?

How to check U102 dead or alive when DMM power on ?

U102 may have an issue, but I can see that the meter is basically reading OK - it's just stuck on the wrong range because U101 is not processing the serial data.
Title: Re: 34401A test failures (many)
Post by: alsetalokin4017 on August 05, 2015, 03:55:40 am
Another alternative:

http://www.utsource.net/ic-datasheet/1SK6-0001-9914.html (http://www.utsource.net/ic-datasheet/1SK6-0001-9914.html)

I've made several orders from UTSource and have always been happy. It's kind of hard to figure out their price structure though; I don't see a price listed for this part but apparently they do have it. You might email them with an inquiry; the ebay seller price shocked me.
Title: Re: 34401A test failures (many)
Post by: motocoder on August 05, 2015, 04:17:55 am
Another alternative:

http://www.utsource.net/ic-datasheet/1SK6-0001-9914.html (http://www.utsource.net/ic-datasheet/1SK6-0001-9914.html)

I've made several orders from UTSource and have always been happy. It's kind of hard to figure out their price structure though; I don't see a price listed for this part but apparently they do have it. You might email them with an inquiry; the ebay seller price shocked me.

I sent them a RFQ several days ago. They quoted me $80 for the part. No thanks!
Title: Re: 34401A test failures (many)
Post by: motocoder on August 16, 2015, 04:34:39 am
I got what I hope is not a counterfeit replacement part for U101 from an eBay seller. I replaced it, and now I am getting error #604.

Is there a way to verify that U101 is functioning correctly?

I saw another post here that this was caused by a bad U201, U102, and some zeners. It looks like the previous owner already replaced U201, and checking the voltages on its pins, it seems to be working.

I notice that the voltage across CR203 is 4.58V. Schematic says this is a 5.1V zener. I know the zener voltage does vary with current, but that seems pretty far off.