Page 1 of 2

Suggestions on Plus 4 problem?

Posted: Sat Jan 26, 2019 6:04 am
by Gyro Gearloose
What problem could this screen indicate?

Re: Suggestions on Plus 4 problem?

Posted: Sat Jan 26, 2019 11:47 pm
by Gyro Gearloose
There are 8 stripes there, I find that amusing. 1000 characters / 8 = 3.125. Each stripe is 3 rows tall, and .125*40 = 5. And it looks like those are 5 offset areas across the 40 columns.
I find that oddly regular. Some sort of address decoding problem I'd guess. I don't know enough (nothing, really) about the +4 to say more. I'll have to actually read the schematics more and make measurements...

Re: Suggestions on Plus 4 problem?

Posted: Tue Jan 29, 2019 4:25 pm
by Zippy Zapp
Is this all you get at bootup? I have repaired several Plus/4's and all of them were either the MPU or the TED. I had 3 with bad MPU. They are the 2 flakiest chips on the whole board. You are getting something though, and in all my cases it was basically black screen or a black screen with rapidly flickering pinkish lines.

Re: Suggestions on Plus 4 problem?

Posted: Wed Jan 30, 2019 12:13 am
by Gyro Gearloose
This is all I get, it does not flash or change on reset or power cycling. When reset is pressed, I quickly see a bunch of @ on screen... So there's activity and it's repeatable.

Re: Suggestions on Plus 4 problem?

Posted: Wed Jan 30, 2019 2:27 am
by Gyro Gearloose
My math is a bit off; those 5 zig zag sections mean they are 8 characters across (40/5). The three rows (3x40) are 120 characters, plus the eight... makes 128.
The last stripe is not the same size, so 7 x 128 = 896 characters. 1000-896 = 104. 104 - (2x40) = 24. That makes 16 + 24 = 40. As we can see, something is counting with 7 bits.

Re: Suggestions on Plus 4 problem?

Posted: Tue Feb 05, 2019 1:14 am
by Gyro Gearloose
Well I'm a bit further along. Power supply, reset signal, and oscillator are all OK. Seems logic levels and chip temperatures are all OK. I'm wondering if BASIC or KERNAL ROMs can go bad in these machines?

Re: Suggestions on Plus 4 problem?

Posted: Wed Feb 06, 2019 6:38 pm
by Zippy Zapp
Do you have an EPROM burner? The nice thing about these ROMs is that they are easily replaceable with EPROMs without an adapter. IIRC they are 28 PIN. And they can absolutely go bad. I had one go south and they seem to get very warm too. I replaced with a burned EPROM with a resistor and a switch soldered directly to the pin legs for JiffyDOS and original kernal switcheroo. The nice thing about using EPROMs is they run much cooler.

Re: Suggestions on Plus 4 problem?

Posted: Thu Feb 07, 2019 1:07 am
by Gyro Gearloose
I looked at U21 which is part of the special address decoding of the +4. In typical cheap-ass Commodore fashion they use unbuffered outputs from U21 directly to the expansion ports.
Not even a 10R resistor in series. Joyful.
If anything goes wrong on the expansion port it can destroy U21 but it seems OK so far.
Strangely I don't have an EPROM programmer. I might have to look into that.

Re: Suggestions on Plus 4 problem?

Posted: Thu Feb 07, 2019 3:31 am
by Gyro Gearloose
U20 goes to the connector... same idea though.

Re: Suggestions on Plus 4 problem?

Posted: Fri Feb 15, 2019 4:32 am
by Gyro Gearloose
Well all logic levels seemed good and since I could see real characters on screen during RESET, and I know that the KERNAL ROM also contains the character data, I figured it must be good.
I started doubting the DRAM, they do go bad, usually they get quite hot but none were running hot.
However the bus multiplexer chips U9 and U10 were these MOS branded 8708s which turn out to be nothing else than 74LS257s. I've heard these aren't reliable so I simply swapped one out with a part from my bin.

Image

The keyboard is not hooked up so I can't see if it responds just yet, but that's a weird amount of RAM. It's repeatable so I guess there's some problem with the DRAM address bus.

I only had one LS257 in my bin, so obviously the other 8708 is preventing full access to all the RAM.

I'll find a way to get more LS257s and see what happens then.