exxos's Atari, Amiga & retro forum

It's bad enough when I had the TF534 on top of the IDE board and all the cables everywhere!

PhilC wrote: 04 Jun 2019 20:36 It's bad enough when I had the TF534 on top of the IDE board and all the cables everywhere!

I can imagine

Back to "normal" :)

Still having reset issues :roll:

I added another 10K on the reset cap and its behaving again now.. odd..

I've not had it on the scope yet to see whats going on yet.. but as the 555 reset circuit is going, theres probably not much point in diagnosing it.

I'm tempted to do a small kit to replace the reset circuits on the STE,STFM etc as they always seem to be causing issues with something or another.

I thought I would post this snippet of info here my blog as it will be easier to find later..

I think I have worked out what the problem is.. It is really temperamental on powering up. I either get bombs, or just get stuck in the reset loop, or it just goes to desktop without accessing the floppy drive.. All manner of unrelated crazy..

I have 1K on DTACK but oddly each time I put another 2.2K on it, the machine boots right up and works fine. This is around 700R which seems a bit low..

But I also think this could be the same oddity as to why the 16Mhz line needs a 47pF on it.. Basically it would delay the CPU clock by just enough time for it to work.

I think the problem is that with other chips running on 8MHz, DTACK arrives based on that clock, but as the CPU is now running double speed, it reads DTACK a lot faster than it would do normally... As I found with my booster work, DTACK actually arrives slightly earlier than data is actually valid. So slowing down the CPU or speeding up DTACK I think is actually the same fault.

What this would basically mean is a inverter on the 16Mhz line to basically delay the CPU in reading DTACK.. however, the bus transceivers are still LS types.. I am going to quickly look through Farnell and see if there is anything faster which I can get in for tomorrow and give that a try first.. Obviously something is borderline timing -wise, so saving 10ns or so on them might be enough to solve it.. otherwise we will have to look into the 16MHz line, or delaying DTACK...

ahh look what I found :D

Sadly no better... in fact replacing the LS373 with a F373 caused it to white border and black screen :roll: Changing the LS244 with a F244 didn't make any odds.. so back to first idea then....

I have tried various delays on each the clocks, various resistor values, even tried the MMU generated clocks.. Tried capacitors on various lines.. I give up, I can't figure it out. Definitely the resistor on DTACK helps.. but I am still randomly getting malfunctioning floppy drive..

I really wonder if Icky isn't seeing this problem as he has the 1772 buffer boards in place ( I do not) . Is it possible to remove it Icky and see if your floppy starts acting up ?

exxos wrote: 06 Jun 2019 19:44 I really wonder if Icky isn't seeing this problem as he has the 1772 buffer boards in place ( I do not) . Is it possible to remove it Icky and see if your floppy starts acting up ?

Removed buffer board and have read / written to floppy but not seeing any issues.

Icky wrote: 06 Jun 2019 21:57 Removed buffer board and have read / written to floppy but not seeing any issues.

It is more on power up... I have a autoloading floppy which loads GB6... But half the time it just goes to desktop, and if I try and access the floppy, I get the busy bee, and it just goes back to the normal arrow without opening the drive.. If I power down for about 15 seconds, and turned back on again, it generally then works okay again.

What I have done is run GB6 with a stock 8MHz setup with the 161 Providing the clocks.. All scores are basically looking correct at about 100% ( a couple were 99%)... So far the floppy seems to be behaving, but now I have another really odd problem that the video does not always show up, I just get a totally white screen, but I hear the floppy loading.. So the machine is actually running, just with no video.. Bizarre..

This is starting to sound very much like the faulty DMA I had in one of the Megas. Is it worth a try with changing it for another BAD-DMA?