ANRITSU instrument boot up failed discuss(MS462X VNA)?

[miketubby]

I have an Antrisu MS4622A (serial 023202) that is bricked - won't boot.

I have read this thread and have opened it up.  Both the CR3032 coin cell and the SnapHat (M4T28-BR12SH1) are flat - both under 200mV.

How do I get the contents back into the BBRAM/RTC chip - mine has an M48T18-100MH1 SMD soldered to the board. I have ordered a new SnapHat.

I have all the discs that came with the unit but I don't think it will boot the monitor/debugger without the BBRAM being valid?

Does anyone have:

1. A copy of an alternative boot rom that will bypass the BBRAM corrupt problem?

2. A copy of MVME Bug162 that I can program in a 27C040?

3. A copy of a correct BBRAM file for an MS4622A?


Regards


Mike

[ve2ulu]

http://www.bitsavers.org/pdf/motorola/VME/MVME162/firmware/MVME162-633A/

https://rescue.sunhelp.narkive.com/0md0Dts1/looking-for-mvme162-rom-images

check and read...
I dont know if is work...

With good rom (very hard to have), boot and push 1, start Monitor
With terminal on serial port (att: check if reverse tx and rx), 9600n81, type "CNFG;M" and type this data:

"32 " Clock speed
00E0A0000xxx Mac-Address  (sticker on back side of CPU board.)
16 DRAM Size in MBytes, 40 with Option 24 installed
2048 SRAM Size
0128 DSP shared memory
02 Flash Size
08 extended Flash Size, 02 and 08 works both for me

now you have minimal setup to boot.

read carefully BBRAM-RESTORATION.PDF
The serial number dont work for me in this doc, i need to add manualy all hex number and subtract hex 56 to make good checksum!

[miketubby]

http://www.bitsavers.org/pdf/motorola/VME/MVME162/firmware/MVME162-633A/

https://rescue.sunhelp.narkive.com/0md0Dts1/looking-for-mvme162-rom-images

check and read...
I dont know if is work...

With good rom (very hard to have), boot and push 1, start Monitor
With terminal on serial port (att: check if reverse tx and rx), 9600n81, type "CNFG;M" and type this data:

"32 " Clock speed
00E0A0000xxx Mac-Address  (sticker on back side of CPU board.)
16 DRAM Size in MBytes, 40 with Option 24 installed
2048 SRAM Size
0128 DSP shared memory
02 Flash Size
08 extended Flash Size, 02 and 08 works both for me

now you have minimal setup to boot.

read carefully BBRAM-RESTORATION.PDF
The serial number dont work for me in this doc, i need to add manually all hex number and subtract hex 56 to make good checksum!

Okay, I found three copies of 162Bug... 162LX-bug v2.3, 162-bug 2.6 and 162-bug 4.0 - not sure which to use?

My serial number is 023202 so what does that make the checksum using your method?


Mike

[ve2ulu]

I receive today one 27c040 and two 29f040 to test few bin file in my bricked ms4623b.
I test and post the result.


In the restoration document, at the last page, to found right checksum, the document say to use CS command but for me in my MS4622B, the result of CS command make wrong number.
My machine always reject the serial number and write *INVALIDE* with new checksum, after test many case, i found right number with this method: add in hex all  carater (6 from your ascii serial number and 7 from ascii of ANRITSU) and subtract hex 56.

In the pdf, the checksum 0481 are the sum of all caratecter, my machine have another checksum number with *INVALIDE*
 *INVALIDE*=25Bhex,   ANRITSU=226hex,   25B+226=checksum 481hex   like 04 81 in the doc.

ascii 023202= hex 30 32 33 32 30 32 = sum hex 129
ANRITSU 226hex + 129hex = checksum 34F hex



The serial number is last step to do, invalide serial number dont hurt your machine, after you stabilise your machine, you take time to put serial number in BBram.

[ve2ulu]

I try all 162Bug... 162LX-bug v2.3, 162-bug 2.6 and 162-bug 4.0  and also MS4623B - Kopie.bin  on 29F040 flash eprom.
Nothing work for me.

29F040 work in place of 27C040?

Finaly a found, no, 29F040 dont work in place of 27C040.

[miketubby]

I have now received a new SnapHat battery for the RTC and fitted it.  Fitted 2 x AA cell to replace the CR3032 coin cell. and the unit still does not boot - as expected.

I programmed a 27C040 with MVME 162 BUG 4.0 debugger and placed that at U75 instead of the Anritsu boot PROM and I get:

Exception: Access Fault
RAW-BUG>

on the back serial port at 19200bps 8N1 using TeraTerm and a cross over cable.

I think it should be possible to get 162-BUG running and use it to reset the system.  The 162-BUG guide talks of setting GPIOs 1 through 5 on the normal MVME 162 board.  Do we know where the equivalent GPIO bits are on the MS462X motherboard?  For example are they the switches near the FLASH?

Has anyone got MVME 162-BUG working on an Antrisu MS462X motherboard?

Does anyone have the schematics?


Mike

[ve2ulu]

Nice move, try this.

"md fffc0000:2000;b"   Is all BBram dump

"md fffc1ef0:170;b"  The board configuration aerea from 0xfffc1ef8 to 0xfffc1ff6 the last byte at 0xfffcff7 is a one byte checksum.

Try "CNFG;M"

and follow bbram-restoration.pdf

To write use MS command.
MS adress data
example: ms fffc047f 3031 3034 3037 0000 0000
This is write the serial number, ascii format, in right place in BBram.

[Staffan321]

Hello,

I've been following this thread and notice that there are quite a few people trying to get the boot process working on old units. Many seem to be connecting to the machine debug port using a serial interface.
On my 37247 there is no serial connector on the outside, but there seems to be pins available on the MVME-162 CPU board. Is the MS462 using that board too? I'm hoping someone could show me the inside of one in the hope that there is a cable from the CPU board to the backside serial connector. Perhaps I could setup the same kind of connector on mine.
If you have a picture of an opened machine, that would be greatly appreciated.
Regards,
  Staffan

[ve2ulu]

https://www.abex.co.uk/esales/test/anritsu/vna/37247/001703_s02297/hires/p7021744.jpg

The 37247 have serial connector.


https://www.abex.co.uk/esales/test/anritsu/vna/37347/992103_s02854/hires/p1280738.jpg

The 37347 dont have serial connector!

[Staffan321]

Mine hasn’t got the serial connector on the backside. Would like to be able to connect that way if there is debug info available. Ideally would be able to read out eprom.
Regards,
 Staffan

[miketubby]

Nice move, try this.

"md fffc0000:2000;b"   Is all BBram dump

"md fffc1ef0:170;b"  The board configuration aerea from 0xfffc1ef8 to 0xfffc1ff6 the last byte at 0xfffcff7 is a one byte checksum.

Try "CNFG;M"

and follow bbram-restoration.pdf

To write use MS command.
MS adress data
example: ms fffc047f 3031 3034 3037 0000 0000
This is write the serial number, ascii format, in right place in BBram.

It won't work.

While I have 162-Bug firmware it is crashing on start-up and dropping to the very basic RAW-Bug>

To do what you say I need 162-Bug to run normally and not generate:

Exception: Access Fault
RAW-Bug>

If you read the documentation for 162-Bug software it tells you that you need to set GPI1 through GPI5 to specific patterns to get 162-Bug to run.

When 162-Bug is running properly you get the:

162-Bug>

prompt, not the emergency:

RAW-Bug>

prompt.  I am getting the emergency prompt probably because Jumper GPI1 through GPI5 are wrong?

GPI1-5 are jumpers on a real Motorola MVME-162 VME board but where are the equivalent inputs on the MS462X motherboard?


Mike

[ve2ulu]

Oh, i dont know is ms462x cpu board have this feature.

I work to make some adaptor/socket for the BBram to make extenal programmable capacity of BBram without lost memory when i remove from motherboard.

With the right board configuration aerea in BBram, the machine boot and is possible to complet the configuration with the monitor.

[miketubby]

To see if your MS462X motherboard is working.  Burn 162BUG v 4.0 into a 27C040 and place in U75.

Connect a cross-over cable from the rear panel serial port to a PC/Laptop using TeraTerm or similar at 19200bps.

When you turn on the power you should get a message:

Exception: Access Fault
RAW-Bug>


Mike

[miketubby]

All,

Having become quite frustrated with this MS4622A VNA of mine that is a brick and with everything that has not worked so far, I have, today taught myself enough M68K assembler to write an 'unbrick' ROM.

I have created an all new ROM for the MS462xx that has only one purpose - to copy a good set of BBRAM data to the RTC and then halt.

The idea is that you remove the official boot ROM (27C040) and burn this ROM, power up the instrument, wait a few seconds, while it copies new data in to your BBRAM/RTC, then power down the instrument and re-fit the original Boot ROM.

The code was developed using Easy68K and has been dry-run in the simulator.  I have NOT YET tested the code because I had to pack my analyser away due to visitors at the weekend.

I am sharing this early so that other people with the same problem can look at the code, try it, critique it, etc.

Again - I have not yet tried this code. If your Anritsu Analyser explodes, your house burns down or your wife leaves you - its not my fault!


Mike G8TIC

[ve2ulu]

Nice and very simple move!
I dont have 27c040 now but when i receive i try this.

On my side, i work on SMT adaptor to make BBram remove and  program without lost battery power.
And also add bigger battery to keep BBram for 20 to 40year.

[miketubby]

So, I have now had a change to try my 'Unbrick ROM'....

Before: With normal Boot PROM - switch on an all keypad LEDs lit, no activity at all.

After installing running 'Unbrick' and then re-installing normal Boot PROM, on power up:

    1. Keyboard LEDs light briefly and go out - this is different..
    2. Screen flashes briefly but remains blank
    3. Floppy drive did a seek

Cycle the power a second time - now keypad LEDs are jammed on again - back to where we were before!

So, it seems my Unbrick ROM must have done something but now the instrument has reverted to fully bricked again.


Does anyone else have a fully working instrument from which they can load the debugger and make a full backup of the 8Kb BBRAM/RTC area?


Mike

[ve2ulu]

This is my BBram TXT capture of my MS4622B, no cal but fully fonctionnal.
I use Putty for terminal.

[Aiy]

I happened on this other thread
https://www.eevblog.com/forum/repair/looking-for-application-disk-for-anritsu-ms4623b-vector-network-analyzer/

Perhaps the documentation at the mentioned bitsavers site can be of help ?
http://bitsavers.org/test_equipment/anritsu/MS462xx/firmware/

Best regards

[ve2ulu]

Today i remplace two module in my 4622B, is RF source module.
Some time before, my vna dont have nice curve in low level, reflect and transmit, on port 3 in first but also port 1 and 2, some time i have some noise start at -65db with source at +7DBm
After i do Non-ratio Parametric Test, i see the problem i have.
I change the IF mixer with no change, i change the first source, port 1 and 2 and i have nice result, i change second source and now the port 3 work great.
I put the picture before and after of Non-ratio Parametric Test on second source port 3.

[mrf245]

http://www.bitsavers.org/pdf/motorola/VME/MVME162/firmware/MVME162-633A/

https://rescue.sunhelp.narkive.com/0md0Dts1/looking-for-mvme162-rom-images

check and read...
I dont know if is work...

With good rom (very hard to have), boot and push 1, start Monitor
With terminal on serial port (att: check if reverse tx and rx), 9600n81, type "CNFG;M" and type this data:

"32 " Clock speed
00E0A0000xxx Mac-Address  (sticker on back side of CPU board.)
16 DRAM Size in MBytes, 40 with Option 24 installed
2048 SRAM Size
0128 DSP shared memory
02 Flash Size
08 extended Flash Size, 02 and 08 works both for me

now you have minimal setup to boot.

read carefully BBRAM-RESTORATION.PDF
The serial number dont work for me in this doc, i need to add manualy all hex number and subtract hex 56 to make good checksum!

Hello VE2ULU,
Could you tell me what is you mentioned "good ROM"? Is it VNA original BOOT ROM?
And connect RS232 to PC and press 1 in monitor, is it press key 1 in VNA's panel?

[whitebear]

Just in case someone is still struggling with this ...

"RAW-Bug>" is a prompt for ROM-only monitor. Luckily, this monitor has basic memory inspection and modification functions which are enough to change BBRAM contents.
Pressing "h" at the prompt gives help message:

Code: [Select]

RAW-Bug>h
h
D <ADDR>...........(DISPLAY)
M <ADDR>...........(MODIFY)
R <ADDR>...........(READ LOOP)
W <ADDR> <DATA>....(WRITE LOOP)
X <ADDR>...........(EXECUTE TARGET)
RAW-Bug>

Addresses and data words should be entered as 8-digit hex values (e.g. 000012FF).

A simple script can be used to type in necessary commands to write BBRAM. I've managed to unbrick my MS4623B using this method.

[frankly]

Hello whitebear

Thanks for identifying this, that's extremely helpful for the rest of us and completes the effort without the use of a soldering iron.

I counted at least 3 of the 8K images in this thread. They are all different, with "ms4622B_OK" being the most divergent. What image were you successful with? And I assume that this was edited somewhere to load your particular MAC address, eh?

thanks & regards, Frank

[DaJMasta]

Does anyone have any info on the LEDs on the motherboard or the boot process for these units?  Should I just be referencing the materials from the board the CPU board is based on?

I've got one with similar symptoms that just lights most of the front panel lights and doesn't do anything (nothing on screen, nothing on external screen, nothing on RS232), and I've replaced the snaphat, the RTC battery, and copied the contents of the U75 ROM, but I haven't gotten any farther in the process yet.  The four LEDs on the board light and then go out very shortly after boot, and when a keyboard is attached, the caps lock and other lights flash like a normal boot, but nothing else (and no activity light flash on the USB port that's in the place of the 3.5" floppy on most units).  When the U75 ROM is pulled, the four CPU board LEDs light and don't extinguish, so I think the process is at least getting to reading or preliminary checking of the ROM, but I have yet to find signs of light beyond that.

I've tried holding the 1 key on the front panel to try an boot from an alternate source, but again no activity on the FDD replacement LED and no difference on the board (since there's a beeper on it, I would have expected that or LEDs lit if there was a CPU/memory error).  I will see if the linked disk utilities will work on a USB stick in this unit, but since the floppy described doesn't sound like it's bootable, I don't know if my unit will even get far enough through the sequence to try to boot from something.

Is the BBRAM with the fresh snaphat to blame for preventing the boot altogether, or is my unit hanging before it gets to that stage?

[DaJMasta]

I've done a little more digging but made no progress.  I've actually got two units based around the same CPU board - an MS4622 and an MS4623 - and both show the same symptoms at every step.  With the original ROMs, there is no output, all the keypad lights light, the four lights on the board go out, no floppy activity, and the halt LED behind does not light.  I tried the 4.0 version 162BUG in a ROM and haven't gotten any sign of life out of either - no activity on the serial port, nothing on the front panel or screen, all four status LEDs on the CPU board stay lit, no activity from the floppy.

I'm wondering what I could be doing wrong.  Do I need to be using the 2.65 version or the Probe+ version?  Is there some other step that needs to be taken after power on to access the new ROM?  Is there another interface that can provide some insight into whatever problem is going on?

Since it's happening the same with both, I doubt it's a one-off CPU board fault, but I even went as far as to scope all the serial port lines for activity and got nothing, and it really doesn't seem like the 162BUG ROM is actually doing anything on either unit given the onboard LEDs.  I will check the GPIB connection on a whim to see if there's any activity there, at least, and I don't think I've tried clearing the CMOS (jumper near the RTC battery) on the new one, but short of just trying to reseat the CPU or something, is there something else I should be considering?

[ve2ulu]

The BBram are empty or corrupted, you need special boot rom U75 to load BBram with minimal data to boot with original U75 rom.