Tektronix TDS 820

[dgimb]

Hello folks.
I bought a Tektronix TDS 820 for repair.
The main screen is dark, but the segment indicator on the motherboard runs.
After connecting to the debug console port, I have the following log:
-------------------------------------------------------------------------------------------------
Int V
   Bootrom Header Checksum passed.
   Bootrom Total Checksum passed.
   BootRom Check Sum passed.
   Bus Error Timeout test passed.
Kernel Diagnostics Complete.

Calling SDM (monitor) Routine.
   Enabling Bus Control register. Value = 0x67
   IMR 1 Register test passed.
   Misc. Register test passed.
   Timer Interrupt test (Auto-Vector) passed.
NVRam DSACK FAILED, wrote 0x12345678, read 0x78ffffff

NVRam DSACKS test failed.
-------------------------------------------------------------------------------------------------
Also I replaces DS1245Y to new one but the result still the same.

I am also interested in whether it is necessary to flash the DS1245Y memory after replacement to new one.
What exactly firmware need to upload in the DS1245Y memory?

Also on eBay i found diskettes for my oscilloscope model...
I'm curious about the left one on the photo.... what does the "Оptions key" mean?

Thanks!

[TERRA Operative]

All options can be unlocked via GPIB, the procedure is around the various forums.

You will need to program the new NVRAM, try pulling the data out of the old one, otherwise you might need to get a copy from a good unit if your old one has lost its data.

[dgimb]

Solved!

The problem was that there was no signal on the data bus from the buffer chip (CPU data bus).
This was the reason that the communication between the CPU and the NVRAM was not working properly.
The reason is that the signal line in the inner layer of the PCB was broken.

Thanks!

P.S: Also, earlier (before the real cause of the problem was found) all data was extracted from the old NVRAM and loaded into the new one, but the result was the same.
For now, the oscilloscope starts up and works correctly.

[dgimb]

The memory of the DS1245Y may be empty at startup (erased before installation).
The DS1245Y only stores the calibration data in my oscilloscope model (TDS 820).

[TERRA Operative]

I think you will probably find that the calibration data is stored on some EEPROM's on the acquisition board, it's where the cal data is stored for the 500, 600 and 700 series scopes (I could be wrong about the 800 series though, I haven't played with an 800 series scope directly).

The DS1386 (or equivalent) 'Ramified Timekeeper' or 'Ramified Watchdog' holds firmware and the DS1245Y (or equivalent) is the NVRAM that holds user settings, captured waveforms and option unlock data etc. There should be no actual calibration data in there.

[dgimb]

Thanks for the advice.
I'm going to backup the oscilloscope firmware and use the "NI GPIB-USB-HS" adapter to do it.
Can you give me some information about this process?
What are the important points here?

Thanks!

[David Hess]

Solved!

The problem was that there was no signal on the data bus from the buffer chip (CPU data bus).
This was the reason that the communication between the CPU and the NVRAM was not working properly.
The reason is that the signal line in the inner layer of the PCB was broken.

Wow, how did you figure that out?  I might have checked every address and data line at the NVRAM for activity.

[dgimb]

Solved!

The problem was that there was no signal on the data bus from the buffer chip (CPU data bus).
This was the reason that the communication between the CPU and the NVRAM was not working properly.
The reason is that the signal line in the inner layer of the PCB was broken.

Wow, how did you figure that out?  I might have checked every address and data line at the NVRAM for activity.

The falling edge signal of DIR signal was wrong.

The reason was not signal on the RD*\1  pin on U1178.

Please see attached image.

[David Hess]

The falling edge signal of DIR signal was wrong.

The reason was not signal on the RD*\1  pin on U1178.

So with none of the buffers producing the correct output, something earlier in the logic was the problem.

I would not expect that kind of output from any TTL gate because open inputs usually pull high, but FAST TTL has a lower high level input current.

[TERRA Operative]

Thanks for the advice.
I'm going to backup the oscilloscope firmware and use the "NI GPIB-USB-HS" adapter to do it.
Can you give me some information about this process?
What are the important points here?

Thanks!

Here's the instructions and tektool DOS app attached (Can be run from command prompt in Win10).
Also info on unlocking options too.

I've also added some floppy disk tools that will allow you to copy in and out the NVRAM and Calibration data (To backup your cal data. I have seen the EEPROMs go bad once or twice).
Be very careful about what you copy in! Make absolutely sure you copy OUT first, maybe twice then generate checksums and compare to ensure you have a good read.

[dgimb]

The falling edge signal of DIR signal was wrong.

The reason was not signal on the RD*\1  pin on U1178.

So with none of the buffers producing the correct output, something earlier in the logic was the problem.

I would not expect that kind of output from any TTL gate because open inputs usually pull high, but FAST TTL has a lower high level input current.

Pin RD*\1 did not receive a signal from another part of the circuit(see screenshot).
RD*\1 had about 2.3V and this was an undefined state for trigger.

It was resolved with a long wire through the main board

[dgimb]

Thanks for the advice.
I'm going to backup the oscilloscope firmware and use the "NI GPIB-USB-HS" adapter to do it.
Can you give me some information about this process?
What are the important points here?

Thanks!

Here's the instructions and tektool DOS app attached (Can be run from command prompt in Win10).
Also info on unlocking options too.

I've also added some floppy disk tools that will allow you to copy in and out the NVRAM and Calibration data (To backup your cal data. I have seen the EEPROMs go bad once or twice).
Be very careful about what you copy in! Make absolutely sure you copy OUT first, maybe twice then generate checksums and compare to ensure you have a good read.

Thank you very much for your help!