Author Topic: Advantest R3463 trouble (solved) (Read 4592 times)

Maurizio1957 · « **on:** October 14, 2022, 02:34:19 pm »

Hi all, what do you think about a R3463 spectrum analyzer that does not start until 24 - 48 hours have passed since the last start? (excluding 'when it starts do not turn it off'

) It loads the files but the screen remains stuck on the Advantest logo...
Cheers Maurizio

Maurizio1957 · « **Reply #1 on:** October 15, 2022, 08:13:43 am »

Preface: i'm not an electronic professional, i'm just a hobbyist .
A year ago an electronic engineer friend of mine gave me a lot of non-working equipment,
telling me "repairs too expensive, use them to make parts ..." including: HP 8753C (zero signal output
and keyboard not working: repaired with new YIG oscillator and keyboard rebuild) and Advantest R3463 (" don't start").
I have checked the voltages of the power supply and they are ok; I have opened all the boards and there is no visible damage;
before replacing all SMD electrolytic capacitors i took one apart and tested with ESR meter, and it looks good ...
it starts every now and then only when it has been turned off for 1 - 2 days, and then it seems to work ...
is there anyone who has had similar experiences?
Tnx, Maurizio

rainbow · « **Reply #2 on:** October 19, 2022, 06:23:36 am »

A very lucky guy to get an HP 8753C for free ;-)

First step, get the documentation:
- for the R3463 there is only the Operation Manual https://xdevs.com/doc/Advantest/R3265A/ADV%20R3463%2C%20R3465%20Series%20Operation.pdf
- but many circuits/boards are the same of the R3265 which have the fully service manual. Get it at https://xdevs.com/doc/Advantest/R3265A/

Second step, fix the power-on intermittent issue:
- disconnect the power supply from the board and check it with DC load or power resistors. You should load the power supply at half the power at least.
- if the power supply is working fine you need to move to the dc/dc circuits on the boards.
- check if there is some back signal to the power supply that disable it to start
- check all DC/DC circuits with an external power supply (regulating voltage/current)

When you have resolved the power issue you can move on. Many of the Advantest's spectrum analyzer of those years have issues with compact flash and battery.

Maurizio1957 · « **Reply #3 on:** October 19, 2022, 08:18:18 am »

Hi rainbow, thanks for the reply! I downloaded everything and now I have something to read ...
I believe the power supply is working, the voltages are in order, I have measured the output currents with a DC clamp and they are 'reasonable', I have started to test the electrolytics on the outputs of the power supply and (for now) they are ok ...
Thanks again, Maurizio

rainbow · « **Reply #4 on:** October 19, 2022, 09:21:21 am »

Try to understand which power line (5V, 12V, ecc.) doesn't power up or if there is some low resistance on the line voltage when doesn't power up.
Good luck

73

Maurizio1957 · « **Reply #5 on:** October 20, 2022, 05:12:31 pm »

Hi Rainbow,
unfortunately it is not that easy ... the analyzer loads the operating system files, but remains stuck on the initial screen with the Advantest logo. I think the system expects a lock from some pll which doesn't happen.
I checked the output voltages from the power supply and on the 25V (the voltage that feeds the main coil of the IYG) there are> 200 mV of ripple, I will try to replace the electrolytics on that branch of the power supply (the manual states 5 years of life for electrolytics, these have at least 20 ...).
Ciao Maurizio

edpalmer42 · « **Reply #6 on:** October 20, 2022, 07:14:52 pm »

On my R3465, as it's booting up it shows info on the boot process in the bottom left corner of the screen. Do you see anything like that? It would give you a clue what's failing. If you don't see anything, you might have corrupt firmware or bad digital rather than RF hardware.

Ed

rainbow · « **Reply #7 on:** October 21, 2022, 05:45:01 am »

Quote from: Maurizio1957 on October 20, 2022, 05:12:31 pm

Hi Rainbow,
unfortunately it is not that easy ... the analyzer loads the operating system files, but remains stuck on the initial screen with the Advantest logo. I think the system expects a lock from some pll which doesn't happen.

This is why you need to check/test power lines before everything else (also DC/DC converter in the boards).
As @edpalmer42 said, if the problem is the RF part you should see some error message in the left corner or at the end of the self-test procedure.
If there is something wrong in the digital part it can be: corrupted firmware (CF?) or bad SRAM, eeprom components.

You should find the troubleshooting procedure in the service manual.

Maurizio1957 · « **Reply #8 on:** October 22, 2022, 07:22:11 am »

"You should find the troubleshooting procedure in the service manual."
Hi Rainbow, I don't know how to define the troubleshooting section of the manual: ridiculous or outrageous ...
the whole section is 3 pages long, it practically says: if it doesn't go, call assistance ...
ok, the hunt continues ...

rainbow · « **Reply #9 on:** October 22, 2022, 07:34:55 am »

R3265A service manual I has quite good troubleshooting and which voltages you need to check. As always (and correctly), the operating manual said to call the assistance.

Maurizio1957 · « **Reply #10 on:** October 22, 2022, 07:53:22 am »

Hi Edpalmer, yes, I see the boot process, but, when the boot is finished nothing happens and the analyzer remains blocked, Advantest logo on the screen and no error message at the bottom left.
Ciao maurizio

edpalmer42 · « **Reply #11 on:** October 22, 2022, 09:01:45 am »

Remember that this is for an R3465, but here's my info from the bottom left of my screen. Some of these entries go by so fast that you'll need to make a video at 60 fps to see them. My first attempt was at about 20 fps and I missed some of them!

----------------------------------------- start of display ---------------------------
Read system files ...
LOADING ATOS.SYS
LOADING INIT.SYS
SYSPATH = SD:/
LOADING MONS.X
LOADING GPIB.X
LOADING STATUS.X
LOADING REMOTE65.X
LOADING SIOCTL.X
LOADING ROMTASK.X
LOADING DH.X
LOADING VOO.X
LOADING MENU.X
LOADING PCCLOCK.X
LOADING SYSCONF.X
LOADING SRM.X
LOADING ISUPER.X
LOADING MMM.X
LOADING CMPNSTR.X
LOADING DFM.X
-------------------------------------- end of display -------------------------------

Does this match up with your R3463 or does the process stop somewhere?
The time from power on until 'LOADING DFM.X' disappears is about 16 sec. The main display appears about 12 sec. later.

Maurizio1957 · « **Reply #12 on:** October 26, 2022, 02:56:37 pm »

Hi Ed, these are the files that I was able to see at power up:
Read sistem files ...
LOADING ATOS.SYS
'' INIT.SYS
SYSPATH = SD: /
LOADING GPIB.X
'' REMOTE & £ .X
'' ROMTASK.X
'' VOO.X
'' MENU.X
. ? (too fast)
.
'' SRM.X
'' CMPNSTR.X
'' DFM.X
pause...
LOADING RXSRV.X
'' BASIC.X

at this point sometimes it starts, sometimes it doesn't ...

lately part, but I don't know why ...

Ciao Maurizio

edpalmer42 · « **Reply #13 on:** October 26, 2022, 06:11:18 pm »

So it looks like your unit completes my list (more or less) and then continues on to load other things. My R3465 has no options. Maybe yours has other options that use the extra files.

RXSRV.X - maybe some kind of server process?
BASIC.X - some form of the programming language called BASIC?

My unit doesn't include these files at all.

There is a way to copy the internal files to a compact flash card. I did it many years ago, but I don't remember how. I do remember that the info came from the users manual. If you think it would be helpful, you need a PCMCIA to Compact Flash adapter and a *small* Compact Flash card. 32 MB cards work, 64 MB cards don't work. Yes, that's MB as in MegaBytes.

The adapter and Compact Flash card are also useful for saving and copying screen captures. Pick up a USB Compact Flash reader and then you can move the captures to any Windows or probably, Apple computer.

FYI, I've attached a PDF that details my investigations regarding the use of Compact Flash cards.

But at this point, it seems like the digital side of your unit is probably working correctly. But it seems really odd to me that there's no error message and the system just hangs.

When it does boot up properly, have you checked all the outputs (30 MHz, 10 MHz, IF outputs) to make sure they're good? Have you run through the internal calibration procedure? If there is a PLL that's marginal, maybe that will retune it.

Ed

Maurizio1957 · « **Reply #14 on:** October 27, 2022, 08:26:21 am »

Hi And, thanks for the info.
I will do other tests, but first I have to reassemble the power supply, it gets too hot out of its seat.

edpalmer42 · « **Reply #15 on:** October 27, 2022, 07:17:52 pm »

By the way, if you're feeling daring (or desperate) you could try resetting the NVRAM in case something got corrupted. But this will also wipe out the calibration data so don't try it unless you're sure that the autocal function is working!!!

To reset the NVRAM, hold the "." (dot, period, whatever) key as you power the unit up. I've never done it so I don't know how long you have to hold it for or if there's a message that tells you that it's done. I'm guessing that one of the things that gets wiped is the clock.

Have you checked the battery on the CPU board? If it hasn't been changed, it's probably stone dead.

Ed

Maurizio1957 · « **Reply #16 on:** October 28, 2022, 07:13:12 am »

Hi Ed , I apologize for the typo in my last post.
The battery is new, I changed it; now I am waiting for a new set of capacitors for the power supply .
My 3463 has a label with " installed option n. 09, 61, 15" but i don't know the meaning...
until now I aven't read the user manual carefully ... $:=\$
As I said lately the analyzer starts and works apparently well, in my search for a fault I removed all 4 internal boards and the various coaxial cables between them, opened the shields and checked the inside.
maybe there was a bad contact on the boards and now it works?
more news will follow ...
Ciao Maurizio

edpalmer42 · « **Reply #17 on:** October 28, 2022, 07:59:46 am »

When you replaced the battery you probably lost all the calibration info. The intention is that the autocal will replace it all.

Option 09 = CDMA Test Source Control

Option 15 = A version of the BASIC programming language. Also called Program Loader Option. So that explains the 'BASIC.X' program that loads on boot.
Option 61 = CDMA Measurement

I have manuals for Options 15 and 61. Let me know if you need them. They total about 6 MB. I don't know what the size limit is for attachments. I probably got them from the Yahoo Advantest group. It's now moved to groups.io .

A bad connection is a definite possibility. Reseating cards and connectors is always a good idea for troubleshooting, even though that can cause intermittent connections. That just means that the connector needs to be cleaned.

Good luck, Maurizio! Keep us informed!

Ed

Maurizio1957 · « **Reply #18 on:** October 28, 2022, 02:58:11 pm »

a strange fact: if I disconnect the 24V line coming from the power supply,

the YIG oscillator does not work, but there are no anomaly warnings ...

Maurizio

Fraser · « **Reply #19 on:** October 28, 2022, 03:55:43 pm »

Thank you to those contributing to this thread. I have an SA from this series that died after I cleaned the processor PCB that had electrolyte leakage contaminating it. I used a professional swept frequency ultrasonic tank so may have damaged crystals as well

I had removed the battery before the clean so clearly destroyed the calibration data set. Thank you for detailing the auto cal process

Fraser

Gyorgy Albert · « **Reply #20 on:** October 28, 2022, 04:15:06 pm »

Hi fellows,

Don't worry about the "calibration data" backupped in the RAM. It is the "user calibration", which is the result of connecting the CAL output to the RF input, and running the self calibration routines. This has nothing to do with the calibration data of the instrument, which is in the eeproms of each module. So this calibration data is safe, even if your battery runs out. Indeed, pressing the dot key while booting (powering up) the analyzer, is clearing the RAM, together with the user calibration data. But this does not have any effect on the performance of the analyzer. And the user calibration is a procedure which can be made as often is needed.

BR,
G

edpalmer42 · « **Reply #21 on:** October 28, 2022, 05:14:13 pm »

After I replaced the dead battery in my R3465, I had a lot of trouble getting the autocal process to complete successfully. My guess is that due to age and component drift the eeprom data was so out of date that the autocal process gave up. Sort of like "I've had to move things too far. This can't possibly be right!". The problems were related to levels rather than frequency or filters. The process might fail and state that the reference signal level was too low. No, it wasn't. Or later, it might say that the attenuator was bad. No, it wasn't. I had to use an amplifier and attenuator combination to *slowly* move the results. This allowed the autocal process to keep up with the changes. Eventually, after adjusting the pads and running the individual calibration routines over and over, (the unit includes both 'calibrate all' and multiple seperate calibration processes) I was able to get the unit to complete the 'calibrate all' function successfully. It was very strange!

By the way, the entire autocal procedure only works on the low frequency band (9 KHz to 1.7 GHz). How do you calibrate the higher frequencies? Some Advantest spectrum analyzers have a hidden maintenance menu, but AFAIK, the R3463 and R3465 don't have that.

Ed

Maurizio1957 · « **Reply #22 on:** November 23, 2022, 03:49:57 pm »

Hi all,
I replaced all the electrolitics in the power supply, checked all voltages are fine, reassembled all and now there is the same old problem:
analyzer stuck on the Advantest logo... it load files until DFM.X, and then stop loading other files ( RXSRV.X , BASIC.X ).
Now I don't know what to do anymore...

Greetings, Maurizio

edpalmer42 · « **Reply #23 on:** November 23, 2022, 04:25:02 pm »

Two other thoughts ......

If it always hangs when loading BASIC.X, that won't have anything to do with the RF section. But since it always gets that far, it suggests that the digital section is somewhat working. During the boot process, it's copying those files into RAM so maybe you've got a flaky RAM chip. When the system tries to load BASIC.X, it happens to write to a bad address and hangs the system. It works after being off for 24 - 48 hours because it's nice and cool. Try freeze spray and a heat gun to see if you can find one bad memory chip.

I had a problem on my R3465 where the trace was about 2 divisions wide. It turned out to be a bad surface-mount electrolytic capacitor on the RF deck. I started checking other similar capacitors and found that a bunch of them had high ESR. Maybe some of the electrolytics on your CPU board have gone bad.

Ed

Maurizio1957 · « **Reply #24 on:** November 23, 2022, 06:05:54 pm »

< Try freeze spray and a heat gun to see if you can find one bad memory chip.
Thanks Ed, I'll do!!
Maurizio


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Advantest R3463 trouble (solved) (Read 4592 times)

Share me