HP 3458A ; A/D board spare part availibilty and cost

[Extrasolar]

I found an interesting State Machine on top of the U180, with maybe more than 2 metal layers ... So far it has been very difficult to RE, but will get there nonetheless !!

[Extrasolar]

Hi All,

There are some nice updates from the U180 RE attempt.
After learning how to spot the various transistors, the overall process becomes very simple. Nonetheless, there are some areas that are way out of focus so I will have to scan em again.
So far, everything has been scanned and extracted, only the decoder remains (I wrongly assumed that the section was a sequential circuit thus a state machine, but it is just combinatorial). The decoder is easy so no hustle there .
There is another section which handles the FET Channel Modulation. I am going to use MOSFETS instead of JFETS so will think of an analog loop to "modulate" the MOSFET Switches.
There are 18 Drivers for the Switches, most probably those are Latches so not a problem

The Resistor network can be reproduced with 0.1% value tolerance and 0.1% matching tolerance (about 200 Euros/piece in a ceramic substrate inside a larger ceramic package with golden pins).

Will produce a couple of prototype PCBs with different features just to test out more possibilities... 

[Kleinstein]

Are the main reference switches JFETs or MOSFETs (lateral or vertical) ? From what I have read this is not so clear.

For a replacement, the choice of switch fets could be a bit tricky: the 3458 uses 50 K for the input path and 40 K for the main reference paths. To get good compensation of the switch TC one would need switching FETs with a resistance ratio of 5 to 4.

AFAIK there is some "substrate" modulation depening on the input voltage. This makes some sense to reduce the quare part in the switch resistance. When changing the FETs, this needs to be addressed too. With MOSFETs an alternative is changing the drive voltage for the input path.

The resistor value matching should not be so critical. A tricky part can however be TC matching and coupling for the input and main reference path. Self heating can result in V³ type nonlinearity as one principle part to the INL.

[Extrasolar]

Hi Kleinstein,

The main Switches are JFET !!
Since I'm going to use MOSFTES(charge inj. 3pC, leakage 10pA, fc at around 200MHz), I was thinking of placing a modulation scheme ahead of the switch to compensate(not sure if to proceed with this approach or use another). I am also considering extensive simulation for the JFET channel modulation, taken all the dimensions of the switches and mapped all on Cadence.
I believe it can be done easily but since I have not prototyped that, I cant tell for sure !!

[Kleinstein]

For most of the FETs one should not need anything special, just the normal gate control, maybe something to reduc the gate charge spike.
There should be only FETs with special needs: the ones for the input signal(s) 40K and 10K for the high speed mode.  These may want something like individual voltage for the gate drive, like 1/5000 of the input voltage added to the gate voltage.
The leakage specs of the FETs would be only a minor part. The interresting parameters are capacitance and on_resistance.  I would consider something like an ON resistance of about 1/5000 to 1/2000 the series resistor for the FET a suiteable order of magnitude. So some 10-25 Ohms for the 50 K reference resistors.

Anyway, with a recreated / coppied U180 chip one can not directly expect the same performance as the original 3458. There is a chance to get similar performance, but still would need the full testing much like a new design.

[iMo]

Do we know where the U180 in the new black edition come from?

[Extrasolar]

Hi Guys,

Some very nice updates on my U180 RE project are now on Master Ken Shirriffs Twitter account: https://twitter.com/kenshirriff !!

Kind Regards

[Extrasolar]

Hi All,

Here is a partial Decoder, with a few gates identified. It seems that it is all composed by NOR and NOT gates:

[Extrasolar]

Hi All,

I was able to complete the Decoder section of U180, although errors might be present so, take it with a pinch of salt :

[wutieru]

Nice done.

[TheSteve]

https://www.keysight.com/my/partDetail/03458-66533

42 in stock!

[niner_007]

https://www.keysight.com/my/partDetail/03458-66533

42 in stock!

Royal cool, I have a stash of 5 from when they were selling them, let me see if I can extend my stash

It's very cool seeing Keysight committed to reparability like this.  <3

[wolfy007]

https://www.keysight.com/my/partDetail/03458-66533

42 in stock!

Royal cool, I have a stash of 5 from when they were selling them, let me see if I can extend my stash

You need more than 5 spares? How many 3458A's do you have? 

[TheSteve]

https://www.keysight.com/my/partDetail/03458-66533

42 in stock!

Royal cool, I have a stash of 5 from when they were selling them, let me see if I can extend my stash

Do you test each board before tucking them away? One of my original "refurbed" A3's had excessive noise and I had to return it.

I'm flying fast and loose now, no spares on hand.

and down to 40 in stock now.

[eplpwr]

https://www.keysight.com/my/partDetail/03458-66533

42 in stock!

I think KS have different stock locations; yesterday the number of stocked was 23, now it is 17. So, probably different warehouses and different # in stock. I'm in Europe, as my country flag signifies.

[niner_007]

https://www.keysight.com/my/partDetail/03458-66533

42 in stock!

Royal cool, I have a stash of 5 from when they were selling them, let me see if I can extend my stash

You need more than 5 spares? How many 3458A's do you have? 

Too many, you don't want to know

[tonyz663]

I am a newbee on this forum and I just purchased one of the new A3 boards (66533) from Keysight and installed it on my older HP 3458A.  My old A3 board was not working properly.  I am now getting a "202 Hardware Error - Slave Test: Convergence" when I turn on the meter and each time I try to reset it.  When I Autocal it, it gives me a "209 Unable to Read A/D" error and a "101 Calibration Error" after going through the routine.  I was curious if anyone else was experiencing this issue as well with the new A/D boards and if there is something I am missing?  FYI - I have replaced all of the A5 NVRAMs a short time ago.  Is there a software update I need to do?  Just not sure at the moment.  Let me know your thoughts!

[wutieru]

I also bought 66533 but haven’t received it yet. After I test it and tell you the result.

[TiN]

Tonyz663 , sounds like you might have problem with A6 or A5 boards (or less likely but possible A1 and\or power supply), if you see those faults with replacement A3. Do you get any readings from meter? How about self-test?

[tonyz663]

Thanks for the reply TIN.  I've enjoyed your in-depth repair discussions for the 3458A!  I'm beginning to think there is an issue with the new A3 board.  I just re-inserted my drifty original A3 board and it passed the self test and no errors.  And it is actually reading voltage now.  So, I don't think I have issues with the other boards at the moment.  Do you have any experiences with this latest A3 board?  There could be something I am missing?

[TiN]

Interesting. If no errors with your old A3, I'd reach out to KS and asked them for replacement of your replacement A3 board. It's not first time that those refub boards were actually not working and needed additional refub once more  .

[branadic]

It's not first time that those refub boards were actually not working and needed additional refub once more

I have the impression that 03458-66523 designed by Wekomm GmbH is backwards compatible, while 03458-66533 redesigned from 03458-66523 by KS for some odd reason isn't. The boards offered are not refurbed but new.

https://www.keysight.com/my/partDetail/03458-66523
https://www.keysight.com/my/partDetail/03458-66533

Edit: Which firmware version are you running, the latest 9.1?

-branadic-

[TiN]

Perhaps tonyz663 can take some hi-res photos of his board and shine some light for us here. That would be nice.
"Wekomm" board works fine in old 3458A, even with older Rev4 firmware.

[tonyz663]

Well, after a phone call with Keysight and a follow-up e-mail, they tell me that I need a firmware update to 9,2 on the A5 board.  The new 66533 boards will not work with 7,2 or 8,2 version (I currently have 8,2).  My A5 board has one ROM on it.  Is there a way to update my current A5 board to 9,2?  I don't have a UV eraser on hand.  I can program chips with my BP programmer however.  Is there a different chip I could put in and program instead of using the original UV chip?  Is there a 9,2 file available to program the ROM with?  Let me know!

In checking with the Parts ID team, it sounds like you need to update the firmware. Please let me know if you have any questions or if I should close the case.

"Customer can refer to Service Note 12C. 3458A, serial number 2823A15426 is an older instrument.  This one left the factory pre-2001. Firmware upgrade (ROM replacement) is the only solution for an older 3458A. 3458A OLDER Models SN Prefix 2823 are not compatible with the new 03458-66523/333458A  A3  A/D Converter incompatibility issueP/N 03458-66533 REV 4 A/D converter board for the 3458As but I get a error stating "ERRSTR 202, "HARDWARE FAILURE – SLAVE TEST: CONVERGENCE".Regarding the problem description, the assemblies are not DFS/DOA. The issue from what we know at this time is older 3458A firmware revisions (7,2 – 8,2) are not compatible with the new 03458-66523/33 assembly. The latest revision 9,2 – shipping for several years, is compatible. The serial number specified, particularly the ‘2823’ prefix, indicates this an older unit likely manufactured in Loveland when we were HP. My guess is the customer’s unit has revision 7,2 or 8,2 installed.
One solution now is to upgrade the firmware which requires a new A5 assembly. Kit 03458-80049 is available and in addition to containing the latest firmware, it also enables extended memory within the 3458A."

[branadic]

Please check: https://xdevs.com/fix/hp3458a/#firmware

-branadic-