TDS540 repair series - FINALE

[vaualbus]

Finnaly I was able to repair my TDs540 scope. After I've changed all the 54 acq capicitor now it pass SPC.
It still have a error regading one of the chip of channel 4 memory. (address 0x73c0046, the errors is: exp data to be 0x5555 but found 0x5551).
A part of them everythink is ok and it also seem calebrate decently.
The good think is also that the display is very very good still after 20 years.

So guys what do you think about the this scope?
PS: If anybody know whic memory ic corrispond to that adress please let me know and also if anybody know whic ram ic I can use to replace the old one, please answer me.

PROBLEMS FOUND:
      - A transistor shorted on the boost suply
      - Two TVS diode cocked
      - The main switch ic break
      - capicitor on acq board
      - I adedd one ic that was missed

Best regards, Alberto Vaudagna   

[PA4TIM]

nice job, can not help you with that error. But I know how you must feel. I'm just fighting with a Tek 2710 SA. It had no readout and no trace. The whole digital stuff is over complicated but back then probably the only way. In my case all 1uF tantalums were leaking, a lot of 220uF caps were bad,dead diode, transistor, PLS and now I am working on the YIG driver because there were 2 shorted caps on the interface board. I have readout, only one error left and that is hopefully caused by the YIG interface.

Is the TDS also stuffed with custom PLS ?

[vaualbus]

ya, all the the trigger logics is done with hybrid chips, the adcs are of course hybrid and the dacs also.
Also there is four pre amp hybrid and the attenuator.
Basicly all the front end of the scope is done by hybrid ic.
There is also two vlsi asic hribid on the processor board.
on the A version of the scope I know that there is also a hybrid transistor on the psu (the sm called it intergral transistor).

[vaualbus]

Guys i want to update a little bit. After I was able to get the real tds540 acq board schematics from tek and also the super secret calibration software, i was able to probably find the bad sram chip of channel 4. But as now i haven't replaced it yet.
I have also change the non volatile sram to a new one. ( the original have datecode of the 18th week of 92 and still work more ir less, after 3 day of not power the scope didn't mantain the configuration)!. I've replaced with a new ram in which i've programmed a modify version of the original data that enable option af and 2f. Before do that i was a little bit scary because that image  is for firmware fw_2.16 meanwhile mine was fw_2.13.
Eventually it work perfectly!
 And now i also have a ftt into the scope for free!
Next phases: change the original fan that make too much noise, change the ram and found a knob for the frontpanel encoder that now is missing.

[Kjelt]

I have the TDS510 still runs as a charm but I am going to replace all smd elco's just to be sure.
Does your scope also have the two Dallas battery backupped chips, one is a SRAM which I think you mention in your last post, the other one is a RTC and also should be readout and copied in a new one or if it runs out you probably will have some issues.
I would rather have a different solution with an external rechargeabel battery because these chips are not made anymore and to find them in 10 years will be problematic.

[vaualbus]

On the non revision ther eonly one non volatile chip that contain the cal data and the options information.

[vaualbus]

By the way this chip are still avaible and are free in freesample from maxim (that acquire dallas semicondactor).

[nctnico]

I have the TDS510 still runs as a charm but I am going to replace all smd elco's just to be sure.
Does your scope also have the two Dallas battery backupped chips, one is a SRAM which I think you mention in your last post, the other one is a RTC and also should be readout and copied in a new one or if it runs out you probably will have some issues.
I would rather have a different solution with an external rechargeabel battery because these chips are not made anymore and to find them in 10 years will be problematic.

If you scope is from around 1995 or later then the capacitors won't leak. Its better to leave it as is than mess with it. Leaking capacitors can be spotted by looking at the PCB under different light angles. In some angles you'll see a greasy area around the capacitor. The smell of cat pee is also an indicator. About the NVRAM: some people grinded the top off and replaced the battery. Another option is to use an FRAM.

[Kjelt]

If you scope is from around 1995 or later then the capacitors won't leak. Its better to leave it as is than mess with it. Leaking capacitors can be spotted by looking at the PCB under different light angles. In some angles you'll see a greasy area around the capacitor. The smell of cat pee is also an indicator. About the NVRAM: some people grinded the top off and replaced the battery. Another option is to use an FRAM.

Ok thanks, much appreciated, I will investigate in the christmass holiday, then is a nice time to clear the desk and mess two days with it. A colleague of mine has a lot of experience with these Tek scopes and he says they almost all have this elco issue.
Do you have a sort of list which Tek scopes with manufacturing dates or serial numbers are effected, that would help much.

[nctnico]

Its just my experience with two scopes I own. Either way if the caps where bad they would have started leaking by now (even if the scope is from 1995 it's age is close to 20 years). If they don't leak then I wouldn't replace them. I have restored a TDS644A and a TDS544A from the early 90's. Replacing the capacitors is a lot of work and it is easy to damage the pads. It also takes skill and a good iron to solder the new capacitors.

[vaualbus]

So after too much time I finally repair the scope.
Turn out to be two memory ic broke. U118 an U117.

[GreyWoolfe]

Congratulations on your repair and persistence to keep at it this long.  Enjoy the scope.

[twobeseen]

what is the best way to replace the Dallas battery powered SRAM chip? Any desolder advice?

I thougt about cutting it out and putting it on a programming device to read it out. This way the old chip would not be overheated by desoldering.

[nctnico]

Use a desoldering iron and/or desoldering braid. Except for the ground and power pins desoldering the NVRAM is dead easy. The holes are large so it is easy to suck the solder out.