Rohde & Schwarz ZVRE hanging

[dxl]

Hi,

i got my hands on a Rohde & Schwarz ZVRE Network analyzer. Of course it is not working

It boots Windows NT fine, but either hangs at the Splash info screen, or directly after the VNA application is started. It doesn't react to keypresses or anything else, but Windows NT keeps running, so the Main CPU is ok.

Sometimes it says ZVR instead of 'ZVRE' at the splash screen. The log files show:

Code: [Select]

Boot Log File created
Date: 09/24/22
Time: 08:03:48
*****************************************************
Mailbox created 'TRIGGER_MBX' [8] ID : 385
Mailbox created 'SPOOLER_MBX' [20] ID : 386
Mailbox created 'PARSER_MBX' [8] ID : 387
Mailbox created 'OPC_MBX' [8] ID : 388
Mailbox created 'MOUSE_MBX' [8] ID : 389
Mailbox created 'MAINAPP_MBX' [8] ID : 390
Mailbox created 'LPTIO_MBX' [8] ID : 391
Mailbox created 'LINK_MBX' [8] ID : 392
Mailbox created 'IECSTATE_MBX' [8] ID : 393
Mailbox created 'GUI_MBX' [20] ID : 394
Mailbox created 'GUI_LNK_QMBX' [8] ID : 395
Mailbox created 'FPNL_RGI_QBX' [8] ID : 396
Mailbox created 'FPNL_PGI_QBX' [8] ID : 397
Mailbox created 'FPNL_PER_QBX' [8] ID : 398
Mailbox created 'FPNL_MRC_MBX' [8] ID : 399
Mailbox created 'FPNL_MPL_MBX' [8] ID : 400
Mailbox created 'FPNL_MBX' [8] ID : 401
Mailbox created 'FPNL_GUI_QBX' [8] ID : 402
Mailbox created 'FILE_TRC_MBX' [8] ID : 403
Mailbox created 'FILE_MBX' [8] ID : 404
Mailbox created 'EXTDEV_MBX' [8] ID : 405
Mailbox created 'ERROR_MBX' [50] ID : 406
Mailbox created 'COM2_MBX' [8] ID : 407
Mailbox created 'COM1_MBX' [8] ID : 408
Mailbox created 'BEEPER_MBX' [8] ID : 409
Mailbox created 'API_TP_QMBX' [8] ID : 410
Mailbox created 'API_MBX' [20] ID : 411
Mailbox created 'API_LNK_QMBX' [8] ID : 412
Mailbox created 'API_GUI_AMBX' [8] ID : 413
Mailbox created 'API_GEN_QMBX' [8] ID : 414
Mailbox created 'API_COM2_QMB' [8] ID : 415
Mailbox created 'API_COM1_QMB' [8] ID : 416
Mailbox created 'ALARM_MBX' [8] ID : 417
Started thread 'InitialisationMain', Priority NORMAL
Loaded DLL 'LINK_JOB'
Loaded DLL 'ERR_JOB'
Loaded DLL 'API_JOB'
Loaded DLL 'GUI_JOB'
Loaded DLL 'TIMER_JOB'
Started thread 'TimerControlLoop', Priority NORMAL
Started thread 'TimerControlLoop', Priority NORMAL
Loaded DLL 'PARSER_JOB'
Spooler DLL Initialising
SpoolController Object Created
Loaded DLL 'SPOOLER_JOB'
Started thread 'Spool', Priority NORMAL
Loaded DLL 'EXTDEV_JOB'
Calling init function: FPNL.DLL init function
Calling init function: LINK.DLL init function
Started thread 'FPNL key', Priority NORMAL
Started thread 'FPNL config', Priority NORMAL
Started thread 'FPNL macro player', Priority NORMAL
Started thread 'FPNL macro recorder', Priority NORMAL
Started thread, Priority NORMAL
Link job: Error downloading C:\R_S\INSTR\TP\R4_RUN_Z.BTL
Link job: Error downloading C:\R_S\INSTR\TP\R4_RUN_Z.BTL
Calling init function: ERR.DLL init function
Calling init function: API.DLL init function
Started thread 'Link send', Priority NORMAL
Started thread 'Link receive', Priority NORMAL
Started thread 'Link file I/O', Priority NORMAL
Started thread 'ERROR_TASK', Priority NORMAL
Calling init function: GUI.DLL init function
Started thread 'API_CDT', Priority NORMAL
Started thread 'API_TRGCT', Priority NORMAL
Started thread 'API_HWDUMP', Priority NORMAL
Calling init function: PARSER.DLL init function
Started thread 'GUI_CDT', Priority NORMAL
Started thread 'IECSTATE thread', Priority NORMAL
Started thread 'Parser monitor', Priority NORMAL
Started thread 'Parser', Priority NORMAL
Started thread 'Parser monitor', Priority NORMAL
Started thread 'Parser', Priority NORMAL
Calling init function: EXTDEV.DLL init function
Started thread 'Parser pipe server', Priority NORMAL
Started thread 'PARSER OCAS/OPC', Priority NORMAL
Started thread 'RSIB State Channel thread', Priority NORMAL
Started thread 'RSIB wait for connection thread', Priority NORMAL
Started thread 'Parser', Priority NORMAL
Started thread 'Parser', Priority NORMAL
Started thread 'TimerControlLoop', Priority NORMAL
Started thread 'Parser', Priority NORMAL
Started thread 'TimerControlLoop', Priority NORMAL
Enable message sent to 'GUI_MBX'
JOB_ENABLE received
Enable message sent to 'API_MBX'
Watchdog Task: Watchdog disabled
Started thread 'MainApp', Priority NORMAL
Started thread 'MainApp', Priority NORMAL
Started thread 'API_CMOS_ST', Priority NORMAL
Deleting SpoolController Object

The 'Error downloading...' Message above is likely the problem (although sometimes it says succeeded but the analyzer also doesn't work). Anyone here with experience working on these Analyzer? I'm not sure where to start digging...

[dxl]

Looks like there are four transputers in this beast. memory seems to be ok, i would expect mtest to have printed an error message otherwise.

[dxl]

I just noted that i can change settings and switch menus in the VNA frontend. It just hangs as soon as i change some RF parameter (like Center frequency, span etc). So i guess there's a communication problem between the PC parts and the transputers.

The good thing is that the ZVRE worked once for a few seconds - and showed a pattern in the smith chart that matched the open port. So it seems like it's not totally broken. I think i probe around the MCU signals later, at least the Service manual has the pin assignments of the connectors.

[dxl]

I've read the mtest output a bit more carefully, and now it looks like the memory on the MCU board might be the problem. There are two T825 transputers and two 16MB memory modules populated with AS4C4F1 4Mx4 DRAMs. Running the transputer memory utility with the -x option shows more output. As you can see in the attached pictures, there's is a stuck bit error reported.

The error moves with swapping the modules, so it looks like one of the memory modules is broken. A bit odd is that both the address and the data changes when swapping the modules, but i don't know how the modules are connected to the transputers.

So i have to figure out which DRAM is broken, but this gives me at least an idea where to start debugging.

[dxl]

It's not the DRAM module. As you can see in the attached picture, there are two DRAM modules. I removed one of them. If i populate ony the left socket the test passes for both modules. If i put them in the right socket, both are failing. So it must be related to the right Transputer, and not the DRAM module itself.

I'm also attaching a boot floppy here which runs the transputer test - if anyone else wants to try that, i would be interested in the results.

[dxl]

Looks like it's the transputer itself (D7 on the A100 MCU module) - if i keep it cool with freezer spray the unit is working. So i need to figure out where to get a replacement - seems like the non-PGA T805 are pretty rare. Anyone got one to spare or a complete A100 board?

BTW, the error in ispy/mtest is still present, but it changed from 'stuck bit test' to 'address' error.

[Jarek]

Hi !
Good job 

I am trying to buy such a zvr and maybe someday I will have to use your experiences.

Regards Jarek

[Georg - PY5ZSE]

dxl,
while you are at it: make a backup (image file) of the instrument's hard disk. It WILL fail one day, and if you have the image you can replace it by a Compact Flash card with a simple passive adapter. Done so in a ZVCE and a FSEB. I actually consider this replacement good practise when bringing any of the big R&S network- and spectrum analyzers (FSEx, ZVx, FSIQ, FSIG. ESI, maybe others) back to life. When you get these 20+ year old boxes today, the HD's are well beyond their specified life span.

[dxl]

dxl,
while you are at it: make a backup (image file) of the instrument's hard disk. It WILL fail one day, and if you have the image you can replace it by a Compact Flash card with a simple passive adapter. Done so in a ZVCE and a FSEB. I actually consider this replacement good practise when bringing any of the big R&S network- and spectrum analyzers (FSEx, ZVx, FSIQ, FSIG. ESI, maybe others) back to life. When you get these 20+ year old boxes today, the HD's are well beyond their specified life span.

Definitely a good idea. The Harddisk is surprisingly "new" for the instruments age - it's a 20GB disk. However, i already ordered a compact flash + usb ide adapter to make a backup of the disk and replace it with Flash. Do you have any experience whether the RMX based versions can be exchanged with the Windows NT version? Right now it has the Windows NT version, but i wonder whether the RMX version would be faster to boot. But i guess initialization of the measurement Hardware takes most of the time anyways...

[Georg - PY5ZSE]

Having that "rather new" 20 or 30 GB disk most likely means that the original disk already had failed and was replaced by one bought from R&S as a spare part. You could order these indicating your instrument's S/N, and got one that contained the operating system plus your instrument's factory calibration.
AFAIK you cannot change the operating system. The instruments came in 2 varieties., The older ones had a Processor board with an 486 CPU, the enwer ones a Pentium, and the 486 came with the DOS/RMK system, the Pentium with Windows NT. I don't know if anybody ever tried..
And, yes, a significant part of the "boot time" is loading firmware into the transputers and self tests.

[dxl]

I did a small test: The A100 MCU module contains two transputers, one for Measurement correction and one as a general control interfaces. I swapped one with the other and expected that i now have to cool the other to make it work, but in this configuration the ZVRE works - maybe some I/O pin is on the edge of its specification, and in the other circuit it's still ok.