POSTing:
https://youtu.be/G2uS0BaOtpM

OK, 4 SIPPs but only 640K?

Some pics of tests.
Is a 286 20MHZzonly 9 points?

Commander Keen is smooth.
https://youtu.be/_dKaZTZ3Lzc

Wolfenstein is a slide show.

Wow....I didnt expect you get it running! Excellent work...!

Predator99 wrote on 2023-10-15, 07:05:

Wow....I didnt expect you get it running! Excellent work...!

Thanks for the support... I proved you wrong!! 🤣
Old stuff, it is what it is.

Basically here is what happened,
I covered some areas with solder mask and put it on the wrong trace, soldered a wire on the good trace that was next to the broken one... and the broken trace ended under the ISA Bus VLSI chip, worst case it would have meant desoldering it again. Honestly I'd have a long wire go directly to the corresponding lead. I spotted the mistake with continuity tests; all leads tested one by one (45 minutes).
Such a silly mistake made me go almost crazy when in fact I was already done at first try.

Now I have a few touch ups and some solder masking job to do and it'll be ready for use.

I'm planning on replacing the SIPPs with normal SIMMs, more RAM available. What is strange is that 4 SIPPs are only seen as 640KB. 160KB sticks? Odd... 🤣
I'll look for a better BIOS as the one in is an old black and white (can't remember the name), with the three options (boot, settings, diagnostic). I'll try some modbin, other boards have the same chipset and better bioses...
Turbo switch doesn't work, makes no difference, probably BIOS related? or maybe missing connection to chipset? I can live without it though.
That's for another round.

It really was instructive, I learnt a lot and I'll avoid mistakes in the future. Even my soldering skills got better.
All in all I can now go much faster.

Whew. I can relate as cellphone repair tech in microsoldering. I have to do 18 times installing wires on *every* Samsung A7 port repair because all the pads are ripped. Bad design basically. And other broken pads on any HDMI port repair, Thankfully, that does not happen often. Same with iphone 7 and 7plus broken four pads under audio IC, but not too bad to do.

Good thing the Samsung A7 2017 is dying away as a too old phone so now very rarely.

Cheers,

pentiumspeed wrote on 2023-10-15, 21:28:

Whew. I can relate as cellphone repair tech in microsoldering. I have to do 18 times installing wires on *every* Samsung A7 port repair because all the pads are ripped. Bad design basically. And other broken pads on any HDMI port repair, Thankfully, that does not happen often. Same with iphone 7 and 7plus broken four pads under audio IC, but not too bad to do.

Good thing the Samsung A7 2017 is dying away as a too old phone so now very rarely.

Cheers,

I don't envy you, not a small bit.
I saw many videos and it is painful to see the amount of time lost on that.

Next time I'll use a sharpie to pinpoint the traces or holes. It's that simple... 🤣

Today I desoldered the keyboard controller. Beneath was the other side of the hole where the TURBO button's trace was ending.
Job went smooth but on the back I still broke a trace (repaired). 40 pins ain't no joke 😀

Under it are a bunch of res, caps + 2 ICs:

TITO13DS LS00
https://pdf1.alldatasheet.com/datasheet-pdf/v … I/SN74LS00.html

DM7406M
https://pdf1.alldatasheet.com/datasheet-pdf/v … LD/DM7406M.html

Turbo ends to TITO and signal exits as per scheme and does lead to "Turbo" on the second VLSI
https://pdf.datasheetcatalog.com/datasheets/r … l/499829_DS.pdf

Turbo switch does nothing, no increase or decrease at power on. Capped or not it stays the same.
I'll test different BIOSes to check whether one has more settings, maybe it is disabled by default but the old BIOS hasn't the option.

Edit: and there is still the issue of RAM benig seen as 640KB and not the whole 1MB.

Unless I'm wrong.

Maybe the BIOS is still the kind that needs the ram size to be specified through another .exe file?

Edit: issue addressed here
286 BIOS - 640KB only but 1MB installed (TOPCAT chipset)

I tracked down the Turbo signal to a 74LS00 under the keyboard controller.
https://pdf1.alldatasheet.com/datasheet-pdf/v … I/SN74LS00.html

Turbo is active high for max frequency.
Pulled low to have clock divided by a certain value.

LS00 is always 0.15V as is Turbo voltage.

The Turbo isn't responding when button is pressed, but the clock is correctly reported to be 20MHz AT.
Cpu benched 9.x points in Sysinfo, that seems correct.

If it is pulled low but is @ 20MHz isn't this impossible?

I ordered a replacement to see if it was the LS00. If it doesn't solve the issue I'm just gonna assume something else is dead and live with it.

Thanks for reading my ramblings.

Is your button working? Come across a couple of dead buttons in the last year.... I think it's because they are higher voltage switches that self clean themselves with a decent voltage on, but at 5V nope. Possibly hooking them up to an inductive load and 20V or so and hammering them on and off for a bit would improve function.

BitWrangler wrote on 2023-11-04, 23:16:

Is your button working? Come across a couple of dead buttons in the last year.... I think it's because they are higher voltage switches that self clean themselves with a decent voltage on, but at 5V nope. Possibly hooking them up to an inductive load and 20V or so and hammering them on and off for a bit would improve function.

Button works, I first tried with a jumper to be sure.
I'm working on the mobo on my bench.

In pic that's what I measured. To me LS00 must be replaced.

Edit, voltages on the 3 connecteed pins, varies dropping slowly to 0.23 (values around it).

With keyb controller installed (to measure I had to remove it but holes through board allow readings - my mistake) voltages drop to 0.03V to TURBO.

UV mask is much easier to handle and make sure you have UV protection goggles. Epoxy is too hard to handle and real messy.

Cheers,

I replaced the LS00. Still no Turbo.
Cpu starts @20MHz and stays at 20 MHz.

I have 2 suspects:

1. Turbo switch is indeed not signalling correctly
2. Chipset isn't working

Before I do something stupid, if Turbo isn't switched byt keyboard can I conclude that it is not working?

If the keyboard controller is plugged, V to Turbo is +4.79.
If it isn't V to Turbo is +0.05V
I could try to solder a wire to a switch and then to G but it sounds useless and pointless to pull it low this way.

Unless anyone has a solution I'd call it here as it is running at full speed. No problem there and I already spammed enough 😀
It was nice bothering you all.