2.2 Booster in MegaST4

[Bikerbob]

So pin 13 wired up to switch.. same flaky display issues.. in TOS 1.04 from dualtos - floppy continues to boot even though screen is bad.

L46 is a Ferrite bead inductor.. why its going to that pin?? I dont know.. all I know is I have seen it before in other MegaSTs. But in MOST I have seen L46 is just missing.

when I run the Diag Cart (exxos ver) I get I6 Testing Bus Error Handling. I dont know what startup codes from the cart mean.. is there any kind of manual anywhere? Does this give an idea where to go?

James

[exxos]

Did you remove the motherboard ROMs ?

I6 error is nothing to worry about.

In 8Mhz mode you might be able to disconnect the 16Mhz wire and see if it improves. Maybe even try without the buffer board.

You can also remove the DUALTOS ROM, refit the motherboard ROMs and try TOS104 see if that changes anything.

Though if you are not working in 8MHz mode, you must have a bad connection somewhere. But also I said earlier, The resistor packs may or may not be a factor in this problem you have. I'm not sure if anyone even tried a IMP shifter. Trying one of the more standard ones may help.

Unfortunately mega machines are literally one of the worst machines to work with. It is why on the later booster designs I quoted I would not be supporting MEGA ST installs anymore.

[Bikerbob]

The Gal on the buffer board took some good damage(there when I got it). Wondering if that is a possible issue? I assume the Code is proprietary Exxos?

I could dump what I have on that Gal see if it matches what you have? But I also assume if it was corrupt it just would not work.

buffer1.jpg

[Bikerbob]

Did you remove the motherboard ROMs ?

I6 error is nothing to worry about.

In 8Mhz mode you might be able to disconnect the 16Mhz wire and see if it improves. Maybe even try without the buffer board.

You can also remove the DUALTOS ROM, refit the motherboard ROMs and try TOS104 see if that changes anything.

Though if you are not working in 8MHz mode, you must have a bad connection somewhere. But also I said earlier, The resistor packs may or may not be a factor in this problem you have. I'm not sure if anyone even tried a IMP shifter. Trying one of the more standard ones may help.

Unfortunately mega machines are literally one of the worst machines to work with. It is why on the later booster designs I quoted I would not be supporting MEGA ST installs anymore.

Yeah - I know Exxos - the woes with the megaST - I am dealing with it for my PAK68/3 machine.. I did remove the roms. I will try it with 16mhz wire disconnected. and or also remove the buffer(I have not tried that.. assumed with MEGAST it would be needed)

James

[exxos]

Yeah - I know Exxos - the woes with the megaST - I am dealing with it for my PAK68/3 machine.. I did remove the roms. I will try it with 16mhz wire disconnected. and or also remove the buffer(I have not tried that.. assumed with MEGAST it would be needed)

The 16Mhz line is very flaky on the MEGA ST. It runs literally runs from the top to the bottom of the board to the MMU. It picks up a lot of interference and is just inherently unstable. This is why the buffer board was added to get a new 16Mhz signal to drive the booster. Though I don't think anyone has actually tested the IMP chip in all this. In particular with the diode dropping the voltage. Bit of a mixed bag setup to trying diagnose. In particular with so many different revisions of the motherboard doing different things. You really need to get a scope on the signals to see what is really going on with it all.

EDIT:

The IMP *might* be able to drive the booster without the buffer board. But is irrelevant until you get it stable in 8MHz mode. If it is still unstable then I would suggest continuity checking from the top of the booster CPU socket to something like the MMU/GLUE address bus and databus and other signals make sure you have not got anything broken. Failing that, change the pullups. Other than that I'm afraid I'm out of ideas because in 8MHz mode, it literally acts as a DIP to PLCC CPU adapter, nothing more. Well other than address decoding of ROM.

Also stated on my website chipset itself could be causing problems.

Capture.PNG

Though these tests I think were with stock machines without the pullups changed. As your board as all IMP chips. It could be a double-edged problem in your case.

Sustainability is down to the bus pullups along with the brand of chipset. Mix in more confusion with different motherboard revisions. Swapping the chipset might be a simple thing to try before looking towards changing the pullups.

[Bikerbob]

Thanks.. I wont put a tonne more energy into this because exactly what you say. I read the thread on the V2.2 and dealing with the megaST over the years.

I had suggested to the fellow I will try and get the machine back up and running STOCK.. which I did.. and if I can get the booster working then I will.

I know the IMP chipset might have been an issue - but the Blitter was newer so that was good. He had built the bufferboard wrong with header pins and not socket pins - so I had to remove those from the bufferboard and that was a job - not all the pads survived the removal, and when I put the new socket pins in I had to bodge some to make sure they were making a good connection. They were tested, they are all connected, but its just one more thing.

Anyway.. Thanks for the suggestions.

The best machine for this booster to go into is a STFM I assume?

James

[exxos]

The best machine for this booster to go into is a STFM I assume?

The 70789 is the most stable.

Though don't you have a H5 you could try the booster in ? at least then you will know that is functioning correctly or not. If you want to do that i'll figure out how to fit it in. If you can get it working on the H5 then you know the problem is inherently with the MEGA. if it does not work on the H5, then you know the problem is on the booster (at least anyway). Either way something would be ruled out.

[Bikerbob]

You really need to get a scope on the signals to see what is really going on with it all.

I have a good digital scope, what I need to do now is learn how to use it. I really want to learn how to use it in regards to Atari's. Hate learning from Youtube, but might be the best way to learn how to actually use it.

I do enjoy fixing things though :) crazy I know.

James

[Bikerbob]

The best machine for this booster to go into is a STFM I assume?

The 70789 is the most stable.

Though don't you have a H5 you could try the booster in ? at least then you will know that is functioning correctly or not. If you want to do that i'll figure out how to fit it in. If you can get it working on the H5 then you know the problem is inherently with the MEGA. if it does not work on the H5, then you know the problem is on the booster (at least anyway). Either way something would be ruled out.

That is a good idea.. So in the H5 I would remove the CPU - the dual tos.. no need for the buffer. Great idea nothing more stable than that.

Then I can say - your booster is still good.. just not in this megaST.

James

[exxos]

That is a good idea.. So in the H5 I would remove the CPU - the dual tos.. no need for the booster. Great idea nothing more stable than that.

Then I can say - your booster is still good.. just not in this megaST.

Yep.

Remove the H5 CPU and ROM.
Plug the booster into a 5V slot.
Set the booster to TOS206.
Connect the 16MHz wire direct to the shifter pin 39 on the H5.
The BG wire (which you currently have soldered to the blitter pad) needs fitting to a different 5V port on pin 11.

That should be all you need to get working then.