EEVblog Electronics Community Forum

Electronics => Repair => Topic started by: Dave69 on December 27, 2019, 10:45:11 pm

Title: Marconi 2024 no display
Post by: Dave69 on December 27, 2019, 10:45:11 pm
Got a non working Marconi 2024 from an estate.
The unit powers up, but nothing on the display.
Display unit is ok, also all voltages are present.
Reset line is high, so i assume there are no problems with the voltage comperators.
How to proceed?
Any hints welcome.
Dave DL2MDV

Title: Re: Marconi 2024 no display
Post by: coromonadalix on December 28, 2019, 03:13:34 am
maybe something like this ??

https://www.eevblog.com/forum/repair/quick-repair-project-marconi-2024/ (https://www.eevblog.com/forum/repair/quick-repair-project-marconi-2024/)


Apparently the service manual is here

http://www.ko4bb.com/getsimple/index.php?id=manuals&dir=Marconi (http://www.ko4bb.com/getsimple/index.php?id=manuals&dir=Marconi)
Title: Re: Marconi 2024 no display
Post by: Andy Watson on December 28, 2019, 02:30:35 pm
As above, also, if you're not aware, it's worth joining:

https://groups.io/g/Marconi-Test-Instruments
Title: Re: Marconi 2024 no display
Post by: Dave69 on December 28, 2019, 05:07:22 pm
Today i had a look at the control board, looking for cold solder joints. I found IC3 completely unsoldered on one side. Wonder that it ever worked...
After soldering, the unit came back to life.
Display shows error 1: EEPROM checksum.
Will investigate what this means...
Title: Re: Marconi 2024 no display
Post by: TheMG on December 29, 2019, 07:32:38 pm
That means the EEPROM (which stores calibration data, presets, configuration) is corrupt, or it could also mean there is a logic fault causing the CPU to be unable to correctly read data from the EEPROM.

IC3 together with IC4 comprise the address latch, so there may still an issue with that part of the circuit and the CPU simply isn't addressing the EEPROM correctly, resulting in a checksum error.

The other possibility is that the EEPROM contents got corrupted due to intermittent operation of the address latch due to the cold solder joints.
Title: Re: Marconi 2024 no display
Post by: Dave69 on January 06, 2020, 12:46:40 pm
Many thanks for reply.
Will investigate further.
The service passwords seem to be stored in the Eeprom also.
As i don't know them it might be a good idea to replace the Eeprom?!
Can it be exchanged or is there more to know/do.
I assume a calibration has to be done.
As i get some more errors and no proper output signal, there might be an issue with the calibration data...
Title: Re: Marconi 2024 no display
Post by: George Edmonds on January 06, 2020, 07:05:30 pm
Hi

If you post on https://groups.io/g/Marconi-Test-Instruments, I am the person who is most likely to be replying to you.

May I make one point at this time, do not remove/destroy the EAROM (EEPROM).  Calibrating a 2024 is a none trivial task needing at a minimum either a Marconi 2305 or a HP8901/A/B along with a Marconi 6860B power meter and a 6912 sensor head.  A frequency standard is also required.  In particular adjusting the FM tracking is very time consuming.

Best wishes

George G6HIG
Title: Re: Marconi 2024 no display
Post by: TheMG on January 07, 2020, 03:26:58 am
You are correct that replacing the EEPROM with a blank chip will remove all of the calibration data. Calibrating this instrument is no trivial task. This would be an absolute last resort.

For the password, did you try 123456? That's the default password per the service manual. Most of the time people don't change the password from the default on these instruments.
Title: Re: Marconi 2024 no display
Post by: Dave69 on January 07, 2020, 03:14:46 pm
Ok, thanks for the advice. I thought calibration data is only stored from auto calibration procedure.
I tried the default passwords with no success.

If
"The other possibility is that the EEPROM contents got corrupted due to intermittent operation of the address latch due to the cold solder joints."
is true. --> How could i initiate a new checksum calculation? Do a simple write on the EEProm? How?

How could i determine if Eeprom or one of the latches IC3 or IC4 is faulty?
I'm not too good in SMD soldering...

If EEProm is really faulty, i assume i have no possibillity to change it and calibrate this unit.  --> Worst case :-BROKE

Also i don't know if the other error messages (501, 502) are just a result of the bad EEprom data?!
Title: Re: Marconi 2024 no display
Post by: George Edmonds on January 07, 2020, 04:14:38 pm
Hi Dave

You have asked a number of questions, I need a little time to think and make a appropriate reply to you
Please post the serial number from the tag on the rear and the electronic serial number from the menu, the master password can be derived from these serial numbers.

Regards

George.
 
Title: Re: Marconi 2024 no display
Post by: grumpydoc on January 07, 2020, 06:03:54 pm
Ok, thanks for the advice. I thought calibration data is only stored from auto calibration procedure.
I tried the default passwords with no success.

No, mine had nothing like the default in the manual.

Please post the serial number from the tag on the rear and the electronic serial number from the menu, the master password can be derived from these serial numbers.

I never had any luck with this for mine - ended up sniffing the data bus for EEPROM reads "reasonably" close to keying in the code - which worked without too much difficulty but was a bit time consuming (especially as I was using a 'scope, not a logic analyser).

I posted the details somewhere - either here or possibly the mailing list.
Title: Re: Marconi 2024 no display
Post by: George Edmonds on January 07, 2020, 07:04:32 pm
Hi

There is a factory master password that is derived from the instruments serial number, this is in place to unlock instruments that the user has set their own passwords and then forgotten them.

The procedure was you telephoned the local Marconi /IFR office and they gave you a master password derived from the serial number.  The only problem with this is that most, if not all, of the local offices have now closed.  The one in Germany did last autumn.

I have a contact in the UK that can still obtain the master passwords on a very limited and intermittent basis.
 
Currently I am trying to collect as many of these serial number derived master passwords so that someone clever than I can  deduce the algorithm in used to generate them.

Regards

George G6HIG

 

Title: Re: Marconi 2024 no display
Post by: George Edmonds on January 08, 2020, 09:50:34 am
Hi Dave

First passwords, there is a super user password which you MUST have to replace the EEPROM, my previous posting details how this was/may be obtained from the instrument serial number.  My suspicion is that this super user password is not held in the EEPROM, but is calculated in real time by a routine in instrument’s firmware from the instrument serial number that is held in the EEPROM.  Only a super user can enter the necessary instrument serial number in a new EEPROM.

Prior to the 2023/4 instruments there was a second level provision to recalculate and store the EAROM/EEPROM checksum if required in a number of Marconi Instrument types.  This appears to be no longer the case with the 2023/2024 instruments.  What I would suggest that you do is to do a dummy calibration run without changing any data, it looks like as you complete the possible 20 individual calibration steps and quit each one the EEPROM checksum is recalculated and reset.

This calibration procedure starts at page 58 in the service manual that I am using.  It is manual part number 46882-245M dated 8th September 1998.

The numerous tables at the end of the calibration procedure show the multiple adjustments that have to be completed if you change the EEPROM in a 2023/2024.  Having done a number of Marconi instruments calibrations I can assure you that they are a real time consuming nauseous PITA.  They were done in the factory by ATE kit using the GPIB interface

Currently I do not have a 2023/2024 in my workshop, if one is required to help you resolve your problems I can have one early in February.

George.

G6HIG.
Title: Re: Marconi 2024 no display
Post by: grumpydoc on January 09, 2020, 08:22:27 am
First passwords, there is a super user password which you MUST have to replace the EEPROM, my previous posting details how this was/may be obtained from the instrument serial number.  My suspicion is that this super user password is not held in the EEPROM, but is calculated in real time by a routine in instrument’s firmware from the instrument serial number that is held in the EEPROM.  Only a super user can enter the necessary instrument serial number in a new EEPROM.

The EEPROM is definitely used to store a version of the Level 2 password as I managed to retrieve mine by sniffing the data reads from the chip after entering an attempt on the keypad - it is just stored as a 24-bit binary number with no special encoding.

Presumably there is also a serial-number derived default. I have no idea whether the default is used to populate the EEPROM password when the unit is initially set up and calibrated - in which case a user-set password would possibly over-write it. It might also be interesting to know whether the 2024 knows its own serial number because it is coded in the hardware or just lives in the EEPROM and is entered during configuration - it must have been possible for Marconi to set-up a machine starting with a completely blank EEPROM.

Perhaps George could confirm whether, if you know both the default and user-set passwords for a unit, whether either will unlock it or just the user-set one?
Title: Re: Marconi 2024 no display
Post by: George Edmonds on January 09, 2020, 09:36:55 am
Hi

The super user password (the one from Marconi using the instrument serial number ) must be used to enter the instrument serial number into the EEPROM, or so the service manual states.  I have never had to use it this way as I have never had to replace an EEPROM.  I have used it to replace the second level password when it has been forgotten/lost.

I doubt if the instrument serial number is hard coded into it, but it could be,  my guess and it is only a guess is that the routine for writing the serial number into the EEPROM checks that the password in use matches the one for instrument serial number that is being entered in the EEPROM.  Such a check would be required to ensure the validity of the serial number.

The problem now is that there is little to no support available for Marconi Instruments and soon there will be none at all  VIAVI Solutions will focus on 5G cellular only.

My intial concern was that the OP was going to change the EEPROM without realising just what doing so involves.

George G6HIG
Title: Re: Marconi 2024 no display
Post by: grumpydoc on January 09, 2020, 11:37:46 am
My intial concern was that the OP was going to change the EEPROM without realising just what doing so involves.
Yes, that would potentially be a bit of an issue.

Of course we don't know that the EEPROM is not sufficiently corrupted to need a complete re-cal anyway.

Title: Re: Marconi 2024 no display
Post by: Dave69 on January 10, 2020, 12:45:53 pm
Many thanks for all the information.
Attached pic with Serial Nr. 112255049
Title: Re: Marconi 2024 no display
Post by: grumpydoc on January 10, 2020, 01:35:00 pm
Currently I am trying to collect as many of these serial number derived master passwords so that someone clever than I can  deduce the algorithm in used to generate them.
George - if you want to PM me a couple of combinations I'll take a quick look.
Title: Re: Marconi 2024 no display
Post by: George Edmonds on January 12, 2020, 10:59:02 am
Dave

I have asked for the superuser password for your serial number, can now but wait.


George
Title: Re: Marconi 2024 no display
Post by: Dave69 on January 19, 2020, 10:17:52 am
Many thanks George,
I‘m in no hurry, so will gladly wait and give it a try.
Title: Re: Marconi 2024 no display
Post by: George Edmonds on January 20, 2020, 06:12:54 pm
Hi Dave

For your serial number 112255049 the master password is 356273

George G6HIG
Title: Re: Marconi 2024 no display
Post by: Dave69 on January 24, 2020, 02:40:39 pm
Thanks George,

i made a fault entering the passwords.
One has to confirm with "ENTER" and not "SELECT".  |O
Was not aware of this difference, noticed it when reading the original Manual.
I managed to unlock the device with the default passwords.

Changed level 1 password, so it is written to eeprom, bit error "EEPROM checksum" is still there.
Output fed to oscilloscope shows more than one frequency. Unfortunately i don't have a spectrum analyzer.
Error 501 and Error 502 "Fractional-N loop low, Fractional-N loop high" also still there.
I'm unsure if i shoul investigate in this errors first or if they are a follow up from the eeprom fault...

Dave DL2MDV
 



Title: Re: Marconi 2024 no display
Post by: George Edmonds on January 24, 2020, 08:10:10 pm
Dear Dave.

Good to hear that the serial number derived master password unlocked access to the EEPROM.  I did not expect that simply gaining access to the calibration routines would have cleared the checksum errors.  At least now if you do have to change the EEPROM you can at least get past first base for which you must have the serial number derived master password.

To quote my previous posting.

“What I would suggest that you do is to do a dummy calibration run without changing any data, it looks like as you complete the possible 20 individual calibration steps and quit each one the EEPROM checksum is recalculated and reset.  NB you MUST follow the full exit step/s correctly for an EEPROM write to occur.

This calibration procedure starts at page 58 in the service manual that I am using.  It is manual part number 46882-245M dated 8th September 1998.

The numerous tables at the end of the calibration procedure show the multiple adjustments that have to be completed if you change the EEPROM in a 2023/2024.  Having done a number of Marconi instruments calibrations I can assure you that they are a real time consuming nauseous PITA.  They were done in the factory by ATE kit using the GPIB interface”

I doubt if the 501 and 502 errors are directly related to the EEPROM checksum errors, but they may be.  It would well be worth carrying out the “Adjustment procedures” steps 1 and 2 on page 63 as they require no additional test equipment and they do relate to the synthesiser.  What I do know is that finding synthesizer errors in a 2024, or any other instrument, is not a simple task.

At one time Marconi had problems with surface mount electrolytic capacitors manufactured by Philips leaking electrolyte and going open circuit.  If you get a fishy smell when working on the signal generator it is the electrolyte from the surface mount electronic capacitors.

73 George G6HIG.
Title: Re: Marconi 2024 no display
Post by: Dave69 on January 27, 2020, 09:22:48 am
Many thanks George,

i will follow your suggestions as soon as possible.
Also i wonder how to read that operating time. 126hrs? Not that much isn't it? Wonder if this device ever worked properly  ;)
Until now i didn't remove the RF case, but will lookout for bad tantals.
The control board was very poorly soldered, so i wouldn't be surprised if i find more unsoldered parts...

Dave
Title: Re: Marconi 2024 no display
Post by: TheMG on January 28, 2020, 01:59:38 am
Also i wonder how to read that operating time. 126hrs? Not that much isn't it? Wonder if this device ever worked properly  ;)

The number of hours on a piece of test gear really depends on how it was used. For instance, where I work we've got some test gear that is used in the field, which means they are only powered up for anywhere from a few minutes to a few hours at a time, to perform a few tests, then turned off and packed away until needed again which could be days or weeks.

At the other end of the spectrum, you have test gear that is used in automated test racks controlled via GPIB and that pretty much never gets turned off.

And everything in between.
Title: Re: Marconi 2024 no display
Post by: Dave69 on January 28, 2020, 09:08:07 am
So i ran through the calibration routines yesterday and saved the values, most unchanged.
EEPROM error still there. Got no errors during save so i could assume that the read write to the Eeprom is basically working?!
Also i ran into Err. 23 when doing Cal. 108 FM Tracking.
Will try to investigate in that 501 and 502 Errors this week and leave the Eeprom as it is now.
Title: Re: Marconi 2024 no display
Post by: George Edmonds on January 30, 2020, 01:44:59 am
Hi Dave

First may I say that I was referring to the surface mount aluminium electrolytic capacitors NOT the solid Tantalum capacitors which almost always fail short circuit. Finding 501/502 errors start on page 90 to 92 of the service manual.  I think that you are correct in chasing the 501/502 errors first  George