Flashy Clock - Yet another DEV board

[exxos]

Header strips removed.. Was a total PITA again. Starting to question if the pump has blocked up on my desoldering station at this point :roll:

These strips are the easiest to fit but are not great if you plan to insert and remove stuff repeatedly...

IMG_4754.JPG

The sockets I sell in the store have a centre bar which has to be removed as it is in the way...

IMG_4755.JPG

Also found the plastic is a little wider on the socket so I had to file it down a bit...

IMG_4756.JPG

IMG_4759.JPG

The result is a win as the plastic which was acting up just is now working properly ! I have six other ones I need to retest now..

[exxos]

One board comes up a random number of errors, typically around 10..

I wrote a STOS loop to verify the TOS206 file to whats in FLASH...

IMG_4761.JPG

Got AI to figure it out..

Capture.PNG

Thing is, the format is SOURCE,DESTINATION, ADDRESS. so bit 3 isn't always stuck high as other values pass.. unless all other bits just happen to be high on bit 3 :stars:

But at least it gives me a starting point to check the tracks for bit 3.

[exxos]

I don't believe it.. :shock: :shock: :shock:

1.PNG

IMG_4763.JPG

:dizzy:

[exxos]

Currently I have three boards where the flash fails verify. Though one board the RTC also fails. So will look at that one later. One had a ALT-RAM fault, luckily that turned out to be a solder whisker on the SRAM chip. There are two other boards with weird faults that need to look into yet as well..

One of the boards which failed was interesting, because the amount of failures was basically random (in the order of 40,000) every time. While the previous fix helped on some other boards, there seemed to be still a underlying cause somewhere.

I suspected the problem was data was not latched correctly before the bus cycle ended. So again I just got AI to check my code with that thought in mind.. And it actually found 2 more bugs :shock: !

Capture.PNG

So I had one bug where the latching of data wasn't working as expected. It was trying to latch data to late. Hence random faults. And another bug where I had a "inversion" on the WE line latch reset :roll: AI gave me the code snippets to change, which saved me a bunch of typing as well !

TL;DR
Another odd write flash failure sorted !

Now on to diagnosing the other board failures...

[exxos]

Two boards the RTC would not count.. Long story short the oscillator was bad on one board and I seem it is bad on the other board as well :roll:

IMG_4771.JPG

IMG_4772.JPG

So now its counting correctly !

EDIT:

I thought these boards have been in the US cleaner so many times now.. The datasheets suggests not using a US cleaner at all :roll:

RTC modules can be successfully processed through conventional wave-soldering techniques as long as temperature
exposure to the lithium energy source contained within does not exceed +85°C. However, post-solder cleaning with waterwashing techniques is acceptable, provided that ultrasonic vibrations are not used to prevent crystal damage.

So will have to manually wash future boards, oh the fun ! :roll:

EDIT2:
Annoyingly the oscillators I am actually using don't say anything about cleaning :roll:

[exxos]

9 now in store. That's all I could rescue from the current batch.


https://www.exxosforum.co.uk/atari/store2/#0314


It's unlikely I will do any more of these because the cost would be manufacturing them from scratch and recycling the SRAM chips. So unless there is a significant up rise in demand, then it is unlikely any more will get produced in the future.

Thanks for everyone's patience with all this! At least I can finally cross something else off my to-do list !!