Attempting repair of TDS540C - Option 1G - Fail Processor

[Tantratron]

Hum new unpected failure where after removing the RS232-Centronics PCB then did some soft antistatic ESD kind brushing to remove dust on some parts of the TDS540C oscilloscope... the TDS540C will never turn-OFF complete.

Upon pushing the ON/STDY button, the engine fan will constantly try to slightly run then stop then try to spin then stop. Of course no further function will work, no LED, no nothing so for some reason, the uC sees the failed fan hence prevents anymore starting the Oscope.

Any idea what connector or could explain the fan engine to not spin or I shall say spin 1° then stop then try spin 1°... ?

Thank you, Albert

[Tantratron]

P.S. I have now connected the fan-motor to a 25Vdc power supply, it runs nice so I've re-connected the fan to J20 connector on the A11 board. Just checked now electrical continuity between the J27-1 and J27-3 towards the J20 fan power supply, it seems totally OK. Finally used my DMM to measure the voltage between J27-1 and J27-3, the voltage keeps fluctuating or oscillating between 1V to 3V whereas the service manual states it should be 25Vdc.

Would you know if the A11 board with the fail processor board as discussed previously in the thread could affect the PSU or at least affect the Fan power hence blocking the complete start of the oscilloscope ?

[shakalnokturn]

First check all interconnect if you have removed any.
The most common mistake is the side interconnect PCB, it can be reversed but only works one way.
Your symptoms indicate PS is overloaded.

[madao]

If fan doesn't run , it would make no error-code .   25V is  trough wired of   PSU to fan and CRT-unit, nothing diagnostic here.
If it run more than 10 minutes, then turn unit off and added  one message into error-log. "overheating"

I have  buyed  a TDS754C from Israel, really cheap, with  blocked FAN.  But seller don't know about it and tell PSU is probably damage.
This is why, it was really cheap and easy to repair.

Error-logs is very interessing.

You can checked  on  flat cable connection on other side,  here a  picked schematic of  TDS520B ( on this Connector J27 and J26 is same with C & D Serie)

[Tantratron]

OK finally got the scope working again with its fan and of course the initial topic (processor fail).

I do not remember really to have disconnected anything excepting removing the RS232-Centronics board to test if it would affect the partial failure.

Anyway in order to troubleshoot the PSU over-loaded or whatever made the Fan to not work and no screen, no nothing. I did disconnect both J101 and J700 connector which links the PSU to the acquisition board A11. This time the fan would work, the scope would start but get stuck with all green LED and no screen. I did reconnect J101 only and no fan then I only reconnect J700, the fan starts then stop then starts along with relay clinking and LED blinking. For some reason, I tried again to reconnect only J101 no fan and then suddenly the fan worked, the scope started. At then end I reconnect the J700 and now the scope seems to work.

Not sure what happened really but now the scope works again except this damned "processor fail" as we have been discussing from the start in this thread.

At least now the scope has been cleaned from all its dust so I'll post some more pictures later.

One question I do have now, how can I tell visually from A11 and A10 boards if this scope has pre-installed 8M memories so in theory I could enable the 2M option ?

[shakalnokturn]

The SRAM IC references on the acquisition board and a little math should tell you that.

[madao]

Installed AS7C164 or similar =  1M Option
AS7C1024 or similar = 2M option.

I ask you again: Have you programmer for EPROM  (TL866 and other type) ?
Yes, heat  solder iron  and desolder  NVSRAM   (or: use tekfwtool)

[Tantratron]

I ask you again: Have you programmer for EPROM  (TL866 and other type) ?
Yes, heat  solder iron  and desolder  NVSRAM   (or: use tekfwtool)

Nope I do not have an EPROM programmer for the moment.

As for the tekfwtool, I need to check how to compile the C program in my Macintosh (using Xcode) which you mention on the other thread https://www.eevblog.com/forum/repair/gpib-usb-control-between-macbook-air-(macintosh)-and-tds540c/msg2836360/#msg2836360

[james_s]

Something I ran into that may possibly be of use here. My TDS784C has the video trigger option which is a board that sits above the processor board. I found that standard IC sockets raised the Dallas chips too high to fit, however low profile machined pin sockets did fit nicely.

[Tantratron]

Something I ran into that may possibly be of use here. My TDS784C has the video trigger option which is a board that sits above the processor board. I found that standard IC sockets raised the Dallas chips too high to fit, however low profile machined pin sockets did fit nicely.

Thanks james but in this Oscope, I do not have the video trigger board so the A11 processor board sits on the top alone and the A10 acquisition  board bottom.

[Tantratron]

1G Option is BAD, normally TDS540C (also TDS754C) runs up to 2Gs/s.  (export restriction?)

It seems your analysis is correct and back-up by an older thread https://www.eevblog.com/forum/testgear/conversion-of-500mhz-tds744a-to-1ghz-tds784a/msg835530/#msg835530

However not sure exactly what jumper or resistor should be changed or removed in order to unlock the 1G restriction ?

[madao]

Pleas disassemble your TDS540C w 1G Option and pick  A10 Board out.
Pleas read Service Manual  for 500C/700C Series.  It is much step for disassemble to  resistor-ID

4x  ID-resisistor is on bottom of A10 Acquisition board.   
Location: near  4 golden koax connector to  rear BNC connector.  Then pleas flip it  to bottom , you'll found only one 0 Ohm  Resistor on R1061.  Added two  0 Ohm Resistor to R1062 & R1063


Pic show top-layer of  TDS520C/TDS724C A10 Board, location of ID-resistor (of coruse, it is on other layer)

[Tantratron]

OK Matt many thanks for this detailed instructions.

I guess best for the moment to focus on repairing if doable "FAIL ++ PROCESSOR" prior spending time to deconstruct totally the A10 acquisition board. Good to know now that this A10 has many components on its other face, I did not notice this first time I removed the cabinet. In fact the A11 processor board seems to have as well components on its other face... ouch !

With your information, now I understand why the 1M option offers 500K of memory because I did only count 32 memory chips AS7C164 which only made 256K. Now I see 32 mores AS7C164 using a flash light on on the A10 other face which makes the 500K count correct.

I assume then that all TDS540C with 8M of memory (option 2M) would have a slightly different A10 board because the AS7C1024 has 32 pins versus 28 pins for AS7C164. The soldering of these AS7C wether 8K or 128K cannot be universal with the same PCB foot print traces.

Maybe I need a break, tried different things to solve the PROCESSOR FAIL

One thing I wonder since this Oscope keeps correct the time, date... how can we be sure both NVRAM would have an issue explainingg the fault ?

This Oscope was stored few years it seems at the university unplugged so maybe the internal batteries of this NVRAM's almost died but then why it keeps track of the time stamp.

Albert

[Tantratron]

Pleas disassemble your TDS540C w 1G Option and pick  A10 Board out.
Pleas read Service Manual  for 500C/700C Series.  It is much step for disassemble to  resistor-ID

4x  ID-resisistor is on bottom of A10 Acquisition board.   
Location: near  4 golden koax connector to  rear BNC connector.  Then pleas flip it  to bottom , you'll found only one 0 Ohm  Resistor on R1061.  Added two  0 Ohm Resistor to R1062 & R1063


Pic show top-layer of  TDS520C/TDS724C A10 Board, location of ID-resistor (of coruse, it is on other layer)

Hello Madao,

This sunny day morning, I deconstruct the acquisition board of the TDS540C-1G unit per your instructions to find the 4 resistors in charge to allocate 1G 2G, please see attached picture.

Indeed there is 0R resistor for R1061 where the 3 other resistors R1062 R1063 and R1064 are open 

If I add two 0 Ohm Resistor to R1062 & R1063, do I need to leave 0R on R1061 to enable standard option 2G ?

By the way, what is the role of R1064 or more generally, what are the meaning of all combinations of R1061 R1062 R1063 and R1064 ?

If I make the 2G modification, do I need to recalibrate the unit ?

Do you know if upon starting - booting the TDS540C will automatically remove the 1G display and there is no need to remove the 1G option by software ?

Last question, the zoning of R1061 to R1064 shows two IC, these are X24C02 under U1055 and U1052. Are these ICs the EEPROM keeping the factory calibration constant and does firmware update could write-change their content ?

Thanks, Albert

[madao]

R1064-role ?

Four  Resistor  play a role for  ID-numbers.  16 possible combination exists.

only R1061 =  TDS540C/TDS754C with 1G Limit.
R1061,1062, 1063 = TDS540C/TDS754C standard.

I don't know about other combination. Possible: only  R1064 = TDS580C/TDS784C  (not sure..)

I don't know about  1G option on software site,  you can  solder brigde at R1062 and R1063 and let surprised, if 1G is removed.

Yes, EEPROM  24C02 hold calibration constant ,  but not factory calibration constant, it hold current calibration (only partial  constant).

[Tantratron]

OK Matt many thanks for this detailed instructions.

With your information, now I understand why the 1M option offers 500K of memory because I did only count 32 memory chips AS7C164 which only made 256K. Now I see 32 mores AS7C164 using a flash light on on the A10 other face which makes the 500K count correct.

I assume then that all TDS540C with 8M of memory (option 2M) would have a slightly different A10 board because the AS7C1024 has 32 pins versus 28 pins for AS7C164. The soldering of these AS7C wether 8K or 128K cannot be universal with the same PCB foot print traces.

Stupid mistake of me where now I realize both TDS540C with each 1M option do share the same acquisition board layout for 2M. Upon closer look on the attached picture I took, we see 2x2 additional PCB pad if tektronix would have soldered the 8meg (2M memory option) chipsets. In my case, I do have 1M option so 500K memory using 2x14 chips but the board accepts 2x16 chips hence 8 Meg memory.

Do you confirm that all D series systematically solder the 8Meg memories, namely 23 set of 1S7C1014 where we would just play by software enable the no option, the 1M option and the 2M option ?

[madao]

not 100% confirm.

Reason: D Serie, from this serie exists  2 version.  One old and new. Old look same as C Serie (but not same function), it has  AS7C1024.
Newer Series uses BGA-housing and much faster synchronus SRAM (same type on Pentium II /III CPU), wich one has  256kb (2Mbitx16) memory. One channel has 8 pieces synchronus SRAM.

[Tantratron]

R1064-role ?

Four  Resistor  play a role for  ID-numbers.  16 possible combination exists.

only R1061 =  TDS540C/TDS754C with 1G Limit.
R1061,1062, 1063 = TDS540C/TDS754C standard.

I don't know about other combination. Possible: only  R1064 = TDS580C/TDS784C  (not sure..)

I don't know about  1G option on software site,  you can  solder brigde at R1062 and R1063 and let surprised, if 1G is removed.

Yes, EEPROM  24C02 hold calibration constant ,  but not factory calibration constant, it hold current calibration (only partial  constant).

Just soldered wire-strip on resistors then re-assembled the unit. Unfortunately the boot will never complete, all the front panel LEDs stay ON and nothing is displayed on the CRT screen. No choice to go back, de-solder the shorts where it seems hacking 1G to 2G might required software enable or something else on the boards.

[Jwalling]

Do you confirm that all D series systematically solder the 8Meg memories, namely 23 set of 1S7C1014 where we would just play by software enable the no option, the 1M option and the 2M option ?

On the TDS700D series, serial prefix B03XXXX and higher always have 8MB of memory installed. I would think the same is true for 500D...

[madao]

R1064-role ?

Four  Resistor  play a role for  ID-numbers.  16 possible combination exists.

only R1061 =  TDS540C/TDS754C with 1G Limit.
R1061,1062, 1063 = TDS540C/TDS754C standard.

I don't know about other combination. Possible: only  R1064 = TDS580C/TDS784C  (not sure..)

I don't know about  1G option on software site,  you can  solder brigde at R1062 and R1063 and let surprised, if 1G is removed.

Yes, EEPROM  24C02 hold calibration constant ,  but not factory calibration constant, it hold current calibration (only partial  constant).

Just soldered wire-strip on resistors then re-assembled the unit. Unfortunately the boot will never complete, all the front panel LEDs stay ON and nothing is displayed on the CRT screen. No choice to go back, de-solder the shorts where it seems hacking 1G to 2G might required software enable or something else on the boards.

Important:  D1 Bus connector, a side with 3 connector (two is empty) must  show to CPU board.  If side with 3 connector show  to acquisition board, it is worng, flip it again.
This mistaking with D1 Bus connector  happened  also by me. 

It must starting also with worng option and resistor-ID. It doesn't block starting. In few case ,  TDSxxx is showed.
 But flipped  D1 Bus blocks it.    Did you check, calibration switch, it must be protect. ?

[madao]

Do you confirm that all D series systematically solder the 8Meg memories, namely 23 set of 1S7C1014 where we would just play by software enable the no option, the 1M option and the 2M option ?

On the TDS700D series, serial prefix B03XXXX and higher always have 8MB of memory installed. I would think the same is true for 500D...

propably is this notice correct.  I have checking my  NVSRAM-Dump collection, all above B03xxxx, but only one is B04xxxx.
Only TDS714L-case (B01xxxxx), it is excluded, because, it is a "frankenstein" from pre-owner.

[Tantratron]

R1064-role ?

Four  Resistor  play a role for  ID-numbers.  16 possible combination exists.

only R1061 =  TDS540C/TDS754C with 1G Limit.
R1061,1062, 1063 = TDS540C/TDS754C standard.

I don't know about other combination. Possible: only  R1064 = TDS580C/TDS784C  (not sure..)

I don't know about  1G option on software site,  you can  solder brigde at R1062 and R1063 and let surprised, if 1G is removed.

Yes, EEPROM  24C02 hold calibration constant ,  but not factory calibration constant, it hold current calibration (only partial  constant).

Just soldered wire-strip on resistors then re-assembled the unit. Unfortunately the boot will never complete, all the front panel LEDs stay ON and nothing is displayed on the CRT screen. No choice to go back, de-solder the shorts where it seems hacking 1G to 2G might required software enable or something else on the boards.

Important:  D1 Bus connector, a side with 3 connector (two is empty) must  show to CPU board.  If side with 3 connector show  to acquisition board, it is worng, flip it again.
This mistaking with D1 Bus connector  happened  also by me. 

It must starting also with worng option and resistor-ID. It doesn't block starting. In few case ,  TDSxxx is showed.
 But flipped  D1 Bus blocks it.    Did you check, calibration switch, it must be protect. ?

Well I did pay special attention to re-install the D1 bus connector correctly.

As for the NVRAM write enable switch S1002, it was always OFF.

Hacking or removing 1G option to make it standard 2G might require more modification (HW, SW) but the R1061,1062, 1063 short do not work unfortunately 

[Tantratron]

Hello again Matt,

Maybe if you have some spare time and access to a TDS500-700 C serie, try the other way around and leave only R1601 shorted to verify if it will down-grade from 2GS/s to 1GS/s

Another option would be to solder wires with remote-jumpers on R1061,1062, 1063 and 1064 then try boot on all 16 combination possibilities.

Albert

[Tantratron]

Hello again Matt,

This morning I've decided to re-open the standard TDS540C Oscope in my lab, namely the one with 2GS/s to check its Acquisition board layout against the TDS540C-1G board which is restricted to 1GS/s. Please see attached pictures of both PCB's as installed, I did not remove this time to see other face.

There are some difference between 2G and 1G regarding the 4 resistors: R1061 R1062 R1063 and R1064. With the 2G standard, no need to remove the PCB from the Oscope assembly and we can see your statement to be correct, namely R1061 R1062 and R1063 are shorted with 0R resistor. Take notice of R1066 resistors to be also 0R in both 1G and 2G versions.

Anyway both PCB are not exactly the same so I suspect this is why not so easy to hack or convert the 1G to 2G version. What is certain, you're right that the 1G resistors R1061 to R1064 are installed on the other face of the 1G board hence requires more hand work.

Albert

[madao]

What you have, 2 difference version of  A10 Board (also from CPU Board existis  two difference version). Resistor on Bottom is newer.
679-4039 is old,  679-04204-00 is newer. But this numbers is only 2Gs/s ACQ Board. I guest: 671-3825-01 by your instrument with 1G Option EDIT:  671-3825-01  is seen in few photo from albert.

Servicemanual for 500C/600B/700C tells me,  only this A10 board is difference.   

I can checking of Resistor-ID, but  i have version with Resistor on bottom.
I believe yet,  you have put not correctly D1 Bus or other stuff at first try. It is happens also by me.  Wrong ID doesn't block starting.