Poorly Atari 1040STE

[BrettRogersUK]

Hello Everyone,

I'm getting used to posting on here after all my troubles with my machine.

Firstly, I got the Dual TOS switcher and during extracting the TOS roms a track was lifted. I then discovered that there was another pad that had lifted that I hadn't seen. I installed the dual TOS switcher and put right the tracks that had been broken by way of 'bodge wire' on the underside of the motherboard.

Since then I haven't been able to get the machine to boot up at all. Then my power supply went BANG! so I've got and installed the new PSU from Exxos and new switch for the back. All wired up and everything good, however I'm getting a blank white screen. I'm not getting any blips from the keyboard at all no sound from the thing at all.

W102 - W104 are linked in spaces 2&3 as normal. These have never been changed.

I'm at a loss now as to what to try. I've done some investigative work using schematics and have tested continuity of all address lines and data lines from the TOS Switcher (HI & LO ROMS U102 & U103) to the following IC's:

U100 - MC68000_PLCC
U101 - ST_BLITTER
U200 - MC6850
U202 - YM2149
U203 - MC68901
U300 - ST_DMA (DIP40)
U400 - STE_GSTMCU
U401 - STE_GSTSHIFTER
U509 - U512 - 74LS244
U632 - 74LS244

Everything checks out and there is continuity everywhere from the TOS Switcher board. I've checked the voltages too and they are all ok too. 5.11v on the board.

This machine was working perfectly until I changed the TOS roms. It's like the processor can;t understand the TOS and is stuck in a loop cycle and unable to boot up.

If anyone has anything else they can add that I can have a go at and see if I can resurrect this machine I would be grateful.

Regards,

Brett

[Steve]

I'd offer to look at it for you if I wasn't already snowed under with jobs. It's got to still be issues with the TOS ROM section. Looking at schematics and checking continuity might miss some stuff like ROM chips bring interconnected on the same lines and stuff. Only way to properly diagnose is to completely remove everything from there again and physically & visually check every single pin and trace.

[rubber_jonnie]

A white screen is usually a good indication that the machine isn't reading the ROM, which you seem to have surmised yourself.

From a PSU perspective, the white screen is probably a good thing because it likely means that the CPU is coming up, and the machine is trying to boot, so the exploding PSU probably hasn't damaged anything. Can't be ruled out 100% though, but let's assume it is ok for now.

Given it was working and you fitted the ROM switcher you really need to go back and look at that to be 100% sure that there is no damage either topside or underneath the board that you've missed.

Whilst I agree with the others that removing everything relating to the switcher and starting again is probably the way forward, you risk doing more damage so don't do it for now.

Remember that at the time you first had this problem I pulled my own STE with switcher out and managed to break just a single pin on the switcher (Mine is in sockets), resulting in a white screen? It was only by going back and looking closely I found the broken pin.

I know it is very tedious to go back and look at this, but that is unfortunately what I think you're going to have to do.

I doubt a diag cart would help here either, because often when there is a ROM issue, the cart won't boot either, and won't output to serial either.

If the ROM address and data lines are all good all the way back from both sides of the mainboard to the the socket on the switcher, then start looking at the following:

VCC - Is the ROM getting 5v?
GND - Is ground connected? I had this problem on a Mega ST with a cold solder joint at the GND of one of the ROMS causing it to intermittently fail.
/CE - Is chip enable working? If the ROM is not enabled by this signal being low, then the ROMs will never be able to be read.
/OE - Is output enable working? This is also active low and if it always stays high the ROM won't be able to output any data.

Check all of these for continuity from the mainboard. Not directly underneath, but the first hop away on the trace, to the pin headers of the switcher plus the PLCC socket. A pinout of the 27C4096 ROM will help here too.

In terms of /CE and /OE you're going to struggle to see what's going on without a scope (Sorry I don't remember if you have one or not).

These are common reasons I have encountered for ROMs not being read on STs and giving a white screen, though it is by no means an exhaustive list.

[BrettRogersUK]

If the ROM address and data lines are all good all the way back from both sides of the mainboard to the the socket on the switcher, then start looking at the following:

VCC - Is the ROM getting 5v?
GND - Is ground connected? I had this problem on a Mega ST with a cold solder joint at the GND of one of the ROMS causing it to intermittently fail.
/CE - Is chip enable working? If the ROM is not enabled by this signal being low, then the ROMs will never be able to be read.
/OE - Is output enable working? This is also active low and if it always stays high the ROM won't be able to output any data.

Check all of these for continuity from the mainboard. Not directly underneath, but the first hop away on the trace, to the pin headers of the switcher plus the PLCC socket. A pinout of the 27C4096 ROM will help here too.

In terms of /CE and /OE you're going to struggle to see what's going on without a scope (Sorry I don't remember if you have one or not).

Hi,

I only have a cheap oscilloscope that I can use to see if I'm getting anything. It should at least tell me if there is any sort if signal at all.

I was just thinking last night if a diag cart would be of any use and since they are rare it wouldn't help anyway. I can't find one anyway.

The VCC and ground pins are showing 5v on the HI / LO roms but without the pinout for the PLCC ROM I hadn't checked. I didn't want to remove the sticker just in case. So I will check these today.

Thanks again for the suggestions. I will check these today.

Regards,

Brett

[rubber_jonnie]

Hi,

I only have a cheap oscilloscope that I can use to see if I'm getting anything. It should at least tell me if there is any sort if signal at all.

I was just thinking last night if a diag cart would be of any use and since they are rare it wouldn't help anyway. I can't find one anyway.

The VCC and ground pins are showing 5v on the HI / LO roms but without the pinout for the PLCC ROM I hadn't checked. I didn't want to remove the sticker just in case. So I will check these today.

Thanks again for the suggestions. I will check these today.

Regards,

Brett

By cheap, what do you mean? If it's one like the eBay model I described here: https://www.exxosforum.co.uk/forum/viewt ... =91&t=4911 then you may find it hard to see what's going on with /CE and /OE, but give it a go.

By "The VCC and ground pins are showing 5v on the HI / LO roms" I presume you meant that VCC was 5v and GND was 0V? Just checking to be clear, GND from the ROMs should just buzz out to any other GND on the board and shouldn't read 5V. The PLCC ROM does have have two GND pins, and whilst I'm not sure they are both used in the switcher, I know they are both required for programming the ROMs.

As for the diag cart, they are bloody useful, but not so good in this case. I've had quite a few occasions where I had a white screen and the diag cart wouldn't work, even over serial.

Good luck and if you can get some pictures of what /CE and /OE are doing on the 'scope then that would be helpful. I can always scope my own STE if you need a comparison.

Just take your time, you'll get there.

[BrettRogersUK]

By cheap, what do you mean? If it's one like the eBay model I described here: https://www.exxosforum.co.uk/forum/viewt ... =91&t=4911 then you may find it hard to see what's going on with /CE and /OE, but give it a go.

By "The VCC and ground pins are showing 5v on the HI / LO roms" I presume you meant that VCC was 5v and GND was 0V? Just checking to be clear, GND from the ROMs should just buzz out to any other GND on the board and shouldn't read 5V. The PLCC ROM does have have two GND pins, and whilst I'm not sure they are both used in the switcher, I know they are both required for programming the ROMs.

As for the diag cart, they are bloody useful, but not so good in this case. I've had quite a few occasions where I had a white screen and the diag cart wouldn't work, even over serial.

Good luck and if you can get some pictures of what /CE and /OE are doing on the 'scope then that would be helpful. I can always scope my own STE if you need a comparison.

Just take your time, you'll get there.

Hi,

Unfortunately yes, I have the red one. :oops: :oops: :oops: :oops: :oops: This is the only time I've ever needed one so couldn't really justify the cost of a proper one as it won't get used much at all. It'll just be an expensive ornament only coming out on the odd occasion or just sitting on a shelf not doing anything.

I will give it a go though. I've never used a scope before so are there any tips at all? How do I probe the CE and OE pins? Do I just select a suitable ground point and use the probe just touching the respective pin?

Yes, sorry the VCC is showing 5v not VCC and GND. Like I say, I haven't checked until now the ROM chip pins to see if they was getting anything. I didn't want to remove the sticker on top of the ROM to see what it was in case I had to send it back or anything.

I think I don't really have a choice but to unsolder the ROM switcher so that I can examine the pads and tracks under it and also examine the pins connected to the ROM switcher. However, as I have tested all the address and data lines on the top of the switcher the various parts I would think that means that all the pins on the switcher are intact?? I'm not just relying on the solder joints on the underside of the board for continuity.

Would it be worth getting some ROM sockets and fitting those?

Thanks for your help, much appreciated.

Brett

[BrettRogersUK]

Right then,

Having checked the rom for voltage I can say that it is receiving 5v.

The CE pin is measuring 0v and the OE pin is measuring 5.17v. Which is the same voltage as the power supply is kicking out on the 5v line on the board.

The OE pin should be active low so should be reading same as CE pin 0v, am I correct in thinking this? If so where does the OE pin get its signal from and how can I make it pull to GND?

[rubber_jonnie]

Hi,

Unfortunately yes, I have the red one. :oops: :oops: :oops: :oops: :oops: This is the only time I've ever needed one so couldn't really justify the cost of a proper one as it won't get used much at all. It'll just be an expensive ornament only coming out on the odd occasion or just sitting on a shelf not doing anything.

Well, TBH that was the first one I got, so no shame there :) Sadly whilst a good soldering exercise (Mine was a kit), as an oscilloscope it isn't up to much.
I totally get not wanting to spend the money on a 'proper' 'scope, I held off for a long time, but with having so many machines it turned out to be essential. Trouble is, a bad scope can send you up the wrong path. It's a tough choice, especially if you're on a budget, but something like the Hantek USB model I mentioned is a better starting point, even if a little more expensive than the eBay specials.

I will give it a go though. I've never used a scope before so are there any tips at all? How do I probe the CE and OE pins? Do I just select a suitable ground point and use the probe just touching the respective pin?

Yep, the probe does needed to be connected to ground on the machine, then touch the probe to what you want to measure.

Yes, sorry the VCC is showing 5v not VCC and GND. Like I say, I haven't checked until now the ROM chip pins to see if they was getting anything. I didn't want to remove the sticker on top of the ROM to see what it was in case I had to send it back or anything.

Cool, just wanted to be sure we were on the same page re VCC. Regarding the ROM, the sticker doesn't specifically do anything, like cover a window on the ROM, it's just a QC label to say it's tested, removing it won't break the ROM, but you should be able to access all of the legs of the ROM without removing it, and it you are just continuity testing, then you can use a pin to extend your multimeter probes, it won't affect that sort of test.

I think I don't really have a choice but to unsolder the ROM switcher so that I can examine the pads and tracks under it and also examine the pins connected to the ROM switcher. However, as I have tested all the address and data lines on the top of the switcher the various parts I would think that means that all the pins on the switcher are intact?? I'm not just relying on the solder joints on the underside of the board for continuity.

Would it be worth getting some ROM sockets and fitting those?

Thanks for your help, much appreciated.

Brett

If the address and data lines go everywhere they should, then that means it could be something like /CE or /OE. Check everything has continuity from one hop away from the ROM sockets to the PLCC socket or you might miss a trick.

One of the other considerations is damage to the vias that join pads from the top to the bottom of the mainboard. If one is cracked across the middle, and solder is only joining the pin to the pad on the bottom of the board and not to the top as well, then that may be an issue

The main risk to removing everything is that you run the risk of further damage to the mainboard in doing so, hence why I suggested a few other things first. If you still have the original ROMs, and you do end up removing the switcher, then I would definitely go with ROM sockets and test with the original ROMs. Although by no means perfect, you can plug the switcher into ROM sockets, and it is in fact what I did with my both my STEs.

My money would be on a bad connection somewhere in the area you've worked on. BTW you should also check W102, W103 & W104 if you have changed them, a bad solder joint or other connectivity issue there may be preventing proper operation.

[rubber_jonnie]

@BrettRogersUK One final thing. You need to check all of the connections for shorts. All it takes is a whisker of solder and 2 pins that shouldn't be connected are.

Just use your multimeter in continuity mode for this. Start at pin 1, walk down each remaining pin and check for shorts, then move to pin 2 until you have checked all of them. Tedious? For sure. Necessary? Sadly yes.

Also do this for the PLCC socket.

[BrettRogersUK]

@BrettRogersUK One final thing. You need to check all of the connections for shorts. All it takes is a whisker of solder and 2 pins that shouldn't be connected are.

Just use your multimeter in continuity mode for this. Start at pin 1, walk down each remaining pin and check for shorts, then move to pin 2 until you have checked all of them. Tedious? For sure. Necessary? Sadly yes.

Also do this for the PLCC socket.

Well all checked for shorts and none found. Which is good I guess. I have checked with my 'cheapy scope' and can see the signal on the /OE pin is going high from switch on. The /CE line isn't moving. When I check the /CE line with a multimeter the reading is 0v. When I check the /OE pin with a multimeter I get a reading of 5.17v. the /OE pin on the ROM is supposed to be active low but it's showing a voltage and when scoped I can see the pin is always HIGH.

I have checked the RESET pin on the MC68000 too and found that this pin is running high too from switch on, the XHALT pin as well is running high too.

So what else would I need to check from here.

Regards,

Brett