53210A RF Counter repair [mostly FIXED]

[eb4eqa]

Hi all,

I have a 53210A counter in nice shape but with (at least) one problem. It fails AutoCal with error "Computer calibration correction factor out of range" and Selftest with "Error 1". No trace of those erros on the manuals I have found. Interestingly, the counter is partially working. Let me explain 

Depending on gate time, it can actually count up to a certain frequency. For example, with a 100mS gate time, it can go up to (almost) 5.110 000 000 00 KHz. It actually shows 0.000 000 000 00 if I reach 5.11Khz, so I can go up to 5.109 999 543 22 (that is a real frequency I have seen registered in the max register, I haven't been able to get a higer Reading due to phase noise). At that point, if I increase gate time by 1mS to 101mS, I only get dashes. If I decrease the frequency to about 5.059 404 I start to get readings again.

If I increase the gate time by x10, to 1000mS, then it will only measure up to .510 999 900 103 (actual reading, the limit seems to be now 0.511KHz). Notice that max frequency is now /10 than before.

If I decrease the gate time by /10 to 10mS, then it will count up to 51.099 835 447 6 KHz (actual reading, the limit seems to be then 51KHz. Notice that max frequency is now x10 than before.

As you can guess, the highest frequency it will count happens at 1mS gate time and it is 511 KHz.

Supply voltages are fine, as described in the almost-useless service guide. I can see no internal damage. I've checked with both internal and external 10MHz references (that should make no difference anyway).

It can measure the amplitude of the signal up to 350MHz without problem, it is just the frequency reading that seems to "overflow" somehow.

As expected, signal conditioning circuits are fine and show a nice square wave going into a CMP567 ultrafast comparator. Without schematic or block diagram (and complete lack of knowledge about how this counters works), I am stuck. There are two big FPGAS on the board, I hope they are fine (lower frequencies read OK, does that mean they are fine?)

I'd like to think that the counter has resources to measure low frequencies, which work fine, and other means to measure higher frequencies. Whether the difference is all within the FPGAs (I hope not) or in support circuitry I don't know...

If someone has a block diagram, pointers to information, or any kind of help, it will be very appreciated.

(Sorry if you saw this help request on YahooGroups too, I did try there first)

Thank you!

Roberto EB4EQA

www.rbarrios.com

[singapol]

Hope this helps:

literature.cdn.keysight.com/litweb/pdf/53210-90001.pdf

https://www.upc.edu/sct/ca/documents_equipament/d_337_53220a.pdf

Agilent 53200 series 350mhz. freq. counter/timer assembly level service guide:
http://cp.literature.agilent.com/litweb/pdf/53220-90010.pdf

operator's manual?
https://www.upc.edu/sct/ca/documents_equipament/d_337_53220a.pdf

[eb4eqa]

Hello,

Thank you for the pointers to those documents. Sadly, I had already seen them and they are not a lot of help for repairing the instrument.

I am still probing around with the scope from time to time to see if I find something helpful.

Schematics or block diagram would be sooo handy.

Thank you,
Roberto

[DimitriP]

We might just  have to wait for the service manual  to show up on wikileaks ....

[singapol]

Hello,

Thank you for the pointers to those documents. Sadly, I had already seen them and they are not a lot of help for repairing the instrument.

I am still probing around with the scope from time to time to see if I find something helpful.

Schematics or block diagram would be sooo handy.

Thank you,
Roberto

Try measuring power voltages on the motherboard.Refer to page 158 and 159 of the assembly-level service guide for a start. check if there is a reference 10Mhz. clock signal from the oscillator.

PS- Searched hard and long, as this is a current instrument not likely that a service manual is available.
Even if you know how a basic counter works as this is processor controlled, we don't really know how keysight  system is organised.Unless someone dissassemble the firmware to see how it runs.

"Computer calibration correction factor out of range" and Selftest with "Error 1". Of course this can mean a
circuit failure but what? Error numbers start with 100 onwards so error 1 is cryptic    Best you can do is do the usual component troubleshooting and hope you get lucky.

[eb4eqa]

Let me add something to this old thread. I mostly fixed it. One of the two big FPGAs had one pin simply RESTING on the the pads. A tiny drop of shiny flux made it look like a nice solder joint but under it, there was no solder at all. It may seem a silly fault and simple to spot but at that size, let me tell you it was not obvious at all. Now the counter measures the right frequency always 

It still does not calibrate though. I can see the DAC sweeping the TCXO control in coarse steps and then finer ones until it finds the best value. But it somehow fails updating a register somewhere and reverts to the old value after giving those errors.

For now, it works beautifully with an external reference. I'll keep trying to fix the other issue, just because it would be nice to know what happens.

A simple solder joint after all...

Regards,
Roberto

[KE5FX]

Did you reflow the whole FPGA, or just the open pin that you found?

[eb4eqa]

Hi John! Long time user and fan of your GPIB Toolkit here... 

I did not reflow the board but I cleaned and re-soldered every joint with fresh flux and new solder, of both FPGAs. I am certain that is not an issue anymore. There is obvious damage here and there caused by some poor soul doing his best to repair the counter. The calibration issue could be anywhere.

I also checked the DS2432 EEPROM that seems to me the best candidate to hold the calibration constants and the signals seem fine. I even removed it just to find that the counter fails on power-up so it seems to be working.

Thank you,
Roberto EB4EQA

[TheSteve]

This is a heavily delayed reply but I believe error 1 simply refers to an incorrect time/date. You will get this error if you remove the backup battery or remove the processor card.

[eb4eqa]

Hi Steven,

Thank you for your reply. I was actually able to fully repair the counter a few weeks ago.

The remaining problem was the same that the first one I fixed. One pin of the second FPGA seemed soldered but it was actually lying freely on the PAD, the dry flux residue made it look as it was soldered. I went and ultra-carefully cleaned and soldered all the pins one by one and I found that one. Once soldered, the unit now calibrates OK too. It is 100% functional. Those pins are so small that it is difficult to see what is going on.

Regards,
Roberto EB4EQA

[HighVoltage]

Hi Steven,

Thank you for your reply. I was actually able to fully repair the counter a few weeks ago.

The remaining problem was the same that the first one I fixed. One pin of the second FPGA seemed soldered but it was actually lying freely on the PAD, the dry flux residue made it look as it was soldered. I went and ultra-carefully cleaned and soldered all the pins one by one and I found that one. Once soldered, the unit now calibrates OK too. It is 100% functional. Those pins are so small that it is difficult to see what is going on.

Regards,
Roberto EB4EQA

Nice repair !
I also had a 53210A a while back that had a similar problem and Keysight fixed it under warranty.