Tektronix TDS220 oscilloscope -Blank screen -suspect Flash memory corruption

[rh100605]

My TDS 220 oscilloscope suddenly stopped and would not power up, no LCD ccfl running and no voltages at the main board.
I checked out the power supply ( mains supplied via isolation transformer) and found the power pin on the 7552 chip at near zero. All diodes and the 3N70 power FET checked out ok as did the resistors.
The electrolytics showed no signs of aging so I replaced both the 7552A chip and the 22mfd capacitor (C6) on its 10 volt supply. Typical circuit attached below.

The power supply sprung back into life with no board connected. All voltages/ripple were in specification as listed in the service manual

When I connected the main board (679-3546-03 with 163-0792-03 ROM 1.16), the screen remained blank and the calibrator voltage had a dc offset  hf sine wave. The CCFL LCD illumination worked.

I pulled up the  the display flat cable by a few mm so that I could probe it. All signals were as per the service manual but the data lines were quiet.
This lack of data was strange as the Sharp LM32P0731 display datasheet (attached below) says that it has no internal memory for refresh.

I checked out the processor (68EC000) for activity on the address and data lines, There was activity and all the power regulator transistors had sensible voltages.

Previous experience with Flash memory said that it was quite possible for the flash to be corrupted with power glitches.  Many thanks to Jared here on eevblog for encouraging me to investigate the Flash.

I removed, using CHIP QUIK low temperature solder, the 1Mbit PLC33 AM29F010B Flash and tested it by reading-saving/ erasing then programming with the saved image. All verified fine.
I then erased the flash then, after putting a socket on the board, replaced it in the faulty board. I hoped to create a "Reset software" condition.

The scope came alive after several seconds of deep thought !!!!

Using the calibrator as a source, there was noise on the traces at the faster time base settings but it was usable .

There were some problems :-

1, It thought that it was a TDS210 model. As I do not have a TDS2MM/CM  module, I could not communicate with the scope to change the model using the password PITBULL and model number

2.The scope was in Service mode and the 53 step Factory calibration procedure required a stable 100MHz 1 volt pk to pk Sine wave source which I do not have.
   So I would not be able to make the model change "stick" by completing the Factory calibration.

In the meantime, as I now knew it was a board fault rather than the display, I had ordered a replacement TDS220 board from eBay (679 3546-04 v1.17)
I tried the new board in the scope and everything worked fine.

Now I copied the new board flash, saved the binary image and then put the flash image from the 04 board into the 03 board
The scope came alive and was a TDS220 and passed all calibration checks and tests. Service mode was off .

However surely this was consistent calibration from another board so not correct for the 03 board !

I ran self calibration, which took several minutes to run ( so long I nearly gave up), and it passed. The scope now has no noise on the traces at 5ns/division.
Surely calibration is now for the 03 board ?

To be safe I later paired a copy of the 04 ROM with the copied 04 flash so that both were at vsn1.17  before running self cal. This worked well on the board marked 03

The working Flash image is attached below.

I have yet to find a fault.

Any thoughts ?