exxos's Atari, Amiga & retro forum

I've repaired a few of these lately and the older issue machines are so much easier to repair, why? I don't know but they seem to be more tolerant. The rev 4S is the one I hate the most.

I did try a few equivalent transistors for the clock signal but ended up fitting a NOS original and then it booted right away. TR4 & 5 are also very troublesome on those boards but I'm guessing you've found that out already :lol:

TR4 is a normal failure.

Another place (More Fun Making It's Discord) suggested swapping either TR1 or TR2 for TR3 and using a BC54x transistor in its place. Did that and got a clock. The machine started up to a white border and black paper. Piggy backing a 4164 chip on each 4132 chip gave a difference result on IC18, so desoldered that and now it's a yellow border. :-)

Well, I removed al the upper RAM chips so I could test them. Three of the eight bad.

The machine still doesn't boot fully, with a black paper and white border. However, this could well be because I'm using an incompatible early ULA.

The other issue I had was the Hakko FR301 wasn't sucking properly. Indeed, it was partly blowing!

I opened the pump up to find that the flapper valve was partly gummed up with flux, which had got through the filter, and also bits of the filter which had come away. I need to get some spares.

Anyway, after cleaning things up it seems to be a lot better. Hopefully the bits of filter haven't got further into the pump.

stephen_usher wrote: 21 Aug 2024 22:20 Well, I removed al the upper RAM chips so I could test them. Three of the eight bad.

The machine still doesn't boot fully, with a black paper and white border. However, this could well be because I'm using an incompatible early ULA.

The other issue I had was the Hakko FR301 wasn't sucking properly. Indeed, it was partly blowing!

I opened the pump up to find that the flapper value was partly gummed up with flux, which had got through the filter, and also bits of the filter which had come away. I need to get some spares.

Anyway, after cleaning things up it seems to be a lot better. Hopefully the bits of filter haven't got further into the pump.

Those odds sound about right for the RAM...

New Hakko filters arrived today courtesy of Amazon so the FR-301 is back in action.

Pulled IC3 and IC4, 74LS157s, from the Spectrum and they test fine. Cleared out the upper RAM holes and fitted sockets, which also arrived from RS today.

Discovered that the D1 RAM chip is reading high when all the others are low when the system initialises. I hadn't noticed because the display on the Samsung rubbish LCD telly is so poor and dark and so I thought that they were jail bars. Shorting D0 to ground made the lines go away. Doesn't matter which RAM I use either. So I'm stumped.

But I'm wondering now if D1 is the data line the ROM uses during the RAM test and maybe "blue" line is the first iteration of the RAM test and the system is crashing due to the incorrect ULA being in place.

@stephen_usher humour me with this but swap the 175s over, I've had some oddities testing those lately.

It's also worth socketing the CPU and rpm to try eliminate the D1 issue. I had similar recently but it escaped me what the issue was but I thought it was ram chips? Those 4132 chips don't necessarily play nice with 4164s timing wise.

P s. I cheat and use a diagart with the retroleum duag software on it.

PhilC wrote: 22 Aug 2024 21:45 @stephen_usher humour me with this but swap the 175s over, I've had some oddities testing those lately.

It's also worth socketing the CPU and rpm to try eliminate the D1 issue. I had similar recently but it escaped me what the issue was but I thought it was ram chips? Those 4132 chips don't necessarily play nice with 4164s timing wise.

P s. I cheat and use a diagart with the retroleum duag software on it.

The CPU is socketed, all the upper memory is out. I can desolder the ROM this weekend.

I'll just go an pop some fresh 74LS157s in just for you. :-)

P.S. No change. :-)

Ah hadn't realised you'd get the high ram out.

Notuch else to check as the 157s do the ram multiplexing as you know, so you could probably take them out and see what happens with D1?

The good thing about the external duag cart is that it should boot even without a rom on the motherboard.

Those spectrums are simple but annoying to fix at times, almost too easy you over think it :lol:

PhilC wrote: 22 Aug 2024 22:02 Ah hadn't realised you'd get the high ram out.

Notuch else to check as the 157s do the ram multiplexing as you know, so you could probably take them out and see what happens with D1?

The good thing about the external duag cart is that it should boot even without a rom on the motherboard.

Those spectrums are simple but annoying to fix at times, almost too easy you over think it :lol:

Without the 157s the CPU can't write to lower RAM to initialise it, which it's doing.

The only other thing I can see on the data lines is the ULA. Or of course the pull-ups and the rom as you know.

Has that board got a reset switch?