ST536 STE EDITION

[exxos]

Cool! Hoping they work well :)

Me too!! Otherwise I'll have a most perplexing problem on my hands! :lol: I'll have to resort to the old chicken sacrifice thing next..

I suspect the ground point being on the plcc socket at the bottom of the board isn't great. It means noise from the CPU passes the PLD and clock to get to GND. On the TF536 with the dip CPU, the GND is more on the middle left of the board. So there's likely less noise around the PLD then. I suspect as small of a issue it may be, it's likely added a few mV in the wrong places aggravating the problem. It's a bit of a guess though. I mean the thing has a board wide ground plane so it's not like much more could be done anyway.

I've added in more HF bypass caps around the osc as well on the next respin. Plus the osc will be run on a higher voltage which should counteract any gnd bounce issues..

[exxos]

@Steve @Badwolf @coonsgm

This is how the patch board fits..

Notice the board is soldered to the cap, it holds it in place, and its also the GND connection.

I removed the old OSC, but you probably don't need to, looks like there might be room to leave it in place. Its output is isolated by the jumper anyway..

IMG_2921.JPG

TOS206 seems to reset, but I can't remember if thats relating to BLTFIX / MAPROM or something else.. So I just booted EMUTOS and running tentacle now.. (EDIT: its ran all the way though, so will leave it running GB6 next)

The regulator gives 3.76V output (3.6V reg) Not bad for a guess really anyway! (was aiming for 3.80V) 430mV boost *should* be enough to get the clock stable.


@PhilC its not going to be simple to fit on the ST536... The board only has 5V 0V and clock output. So it can be hacked on still somehow.. but I wonder if its worth even bothering....

[PhilC]

@exxos I haven't had the issues you mentioned on the ST536 to be honest. It doesn't work in a Mega ST but once I've got my Mega rebuilt I've a couple of plans to sort that. So probably not worth it for the ST536 for now in its current form.

[exxos]

@exxos I haven't had the issues you mentioned on the ST536 to be honest. It doesn't work in a Mega ST but once I've got my Mega rebuilt I've a couple of plans to sort that. So probably not worth it for the ST536 for now in its current form.

I think the issues only started when I change the oscillator from what the original TF536 used.. if there was a huge problem, 100's of TF536 people would be complaining ...

[PhilC]

@exxos well don't change them :P :chairsmack: :chairsmack: :chairsmack:

Will let you know how I get on with the Mega ST once I find where I've put the motherboard this time.

[exxos]

@exxos well don't change them :P :chairsmack: :chairsmack: :chairsmack:

Couldn't get the original ones anymore :( There was no difference according to the datasheets , other than the new ones had a bit faster rise and fall time. But its output must have been higher than the new ones.. even though the datasheet says not.

I guess it be useful if you could measure the ST536 OSC output see what that is...

Will let you know how I get on with the Mega ST once I find where I've put the motherboard this time.

:thumbup:

[PhilC]

Couldn't get the original ones anymore :( There was no difference according to the datasheets , other than the new ones had a bit faster rise and fall time. But its output must have been higher than the new ones.. even though the datasheet says not.

I guess it be useful if you could measure the ST536 OSC output see what that is...

Ok with the Oscillator being underneath the board, what's the easiest point to pick up the clock signal from the top of the board? I presume it goes directly to the ram?

[exxos]

So there has been an improvement with the clock patch. Before , I couldn't even get to the desktop hardly with the simulated 8MHz clock. Now it seems reliable in doing so.

I've also now found there seems to be a odd TTram bug which seems to be linked to the clock switching somehow. So either a tiny bit more in 8mhz or 50mhz seems to have done the trick.

Not really sure what's going on. Probably won't get time to look deeper. But it could be as the PLD is full, the compiler could be skewing some timings somewhere which need compensation in the logic. But it's only a guess.

The clock patch boards have just been posted out. I'll send updated firmware to you guys later...

[exxos]

Ok with the Oscillator being underneath the board, what's the easiest point to pick up the clock signal from the top of the board? I presume it goes directly to the ram?

The osc goes to the PLD... So removing the osc is the only way... But I don't think there's any need on the ST536 to change the osc... But my later firmware may help you...

[PhilC]

Ok with the Oscillator being underneath the board, what's the easiest point to pick up the clock signal from the top of the board? I presume it goes directly to the ram?

The osc goes to the PLD... So removing the osc is the only way... But I don't think there's any need on the ST536 to change the osc... But my later firmware may help you...

Happy to give it a try, I'm just setting up the new board once I've downloaded the PI software as I can't be bothered to reinstall all the Xilinx stuff on my new PC just so I can use the Impact app.