• Please review our updated Terms and Rules here

Fixing a Sol-20

Nama

Veteran Member
Joined
May 22, 2009
Messages
851
Location
New Zealand
Hi everyone,
I'm a little reluctant to post this as I've had some rude private emails from a particular S-100 user when I posted a question on one of the S-100 forums...but anyway, here goes nothing...hopefully someone here has some experience fixing these and can point me in the right direction.

I am presently trying to fix a SOL-20.
I have removed all the extra cards and am now just trying to get the base unit booting. It boots up to garbage if switch 1-1 is in the on position. If it is in the off position I am getting a screen full of repeating square characters followed by '9' characters. Reading the SOL-20 build manual I have found a section talking about the video RAM, reading this I can assumed that this pattern actually indicates that the machine is at least accessing the video RAM and the video RAM seems to be working fine.

I then removed the onboard 8k of user RAM and have one by one replaced the video RAM with it. Occasionally the screen pattern would change in certain areas which tells me that the specific chip I just transposed is probably bad. Ultimately I found 3 bad RAM chips this way. I have just ordered some new RAM from ebay and it should be here in a week. Bad RAM is certainly one reason the machine is not booting.

Now with all the user RAM removed and switch 1-1 still in the off position I was now expecting to see either a blank screen, or a screen full of singular characters...something consistent at least. What I did get was a screen with lines of nothing and other lines of grey box characters, nothing too remarkable here, however there are also a few seemingly random characters on the screen. I say seemingly because there position actually does not change on subsequent restarts. The only thing that does seem to change is the character that is in the top left corner where I believe the cursor should appear. It's always in the same position, only the character changes. The other random characters, 'w's and '(' are always the same and in the same position. Take a look here:

http://web.me.com/lord_philip/other_computers/SOL-20.html

Anyway, I'm hoping the new RAM fixes things, but I have my doubts. If not then next I'll try to read the SOLOS ROM and see if it's still healthy.

Cheers

Nama
 
Last edited:
If a user is harassing you in private please let the Mods/Admins know. It's easy for us to catch public flames but we won't know about PMs unless you tell us.

We will deal with it if we know.

Good luck with your Sol. I love these machines and know you'll get it going. :)
 
You may find a lack of folks with sol-20 specific experience but hopefully someone smart will chime in. Are those screens always consistent? Have you tried reseating any socketed chips including the CPU? I'm not sure if this is the same but from other threads usually the machine is initialized and the RAM has random content which would be on the screen until the CPU clears that and then reads/loads whatever boot rom to memory and jumps to that code. Not sure where a good site for troubleshooting would be, there's http://www.sol20.org/links.html for links to others who may have some experience. Possible even cctalk if you can get on there.
 
A Sol-20. Wow, that puts my OS Challenger 1 in the shade when it comes to vintageness. Good luck with it Phil!

Tez
 
Last edited:
The screen with the all 909090's is familiar to me; you need to find a way to confirm that you have good RAM, and that SOLOS is loading in the CXXX Hex range properly. Your video is probably OK.

Also make sure the voltages in key places is correct. Where are you located? There are a number of SOL people out there who may be able to help. It's quite a thrill when you finally get to the > prompt. Good luck, keep us posted.

Bill
 
WOW!
I didn't expect anyone to reply at all, and already thanks to everyone (especially Chuck(G) and billdeg) I have a troubleshooting guide, and some insight into how to move forward.

I have new RAM on it's way. I already know that I have 3 bad RAM chips so that certainly won't help. I've done a quick check for the voltages too, but not enough. I'll check again in more detail.

I'm located in Tokyo, so unfortunately I'm probably quite far from most of you.

Tezza, yeah, basically I was thinking about getting either an OSI like yours or a SOL-20. In the end I opted for the older vintage as it's probably the last old computer I'm going to buy for a while, so I thought I'd make this fix special.

Cheers and thanks for the replies.
I'll keep you posted.
 
Last edited:
If a user is harassing you in private please let the Mods/Admins know. It's easy for us to catch public flames but we won't know about PMs unless you tell us.

We will deal with it if we know.

Good luck with your Sol. I love these machines and know you'll get it going. :)

Hi Eric, unfortunately the user was not from the VCF forums.

Thanks anyway.
 
Are those screens always consistent?

Yes, mostly they stay unchanged. The character in the top left corner seems to change, but that's the only one i've noticed.

Have you tried reseating any socketed chips including the CPU?

Yeah, the first thing I tried was to reseat all the chips to some degree. All the RAM has been out and back in multiple times now.

Thanks
 
Some small progress. While I was waiting for the new RAM to arrive I thought i'd try putting the 64k RAM card (actually only containing 48k RAM) back in and see what effect that would have. Before doing so, I first checked that one of it's 16k banks was set to the 0000 address space. On power up I now get a blank screen with a flashing cursor in the top left corner, however there is no > prompt showing, just the flashing cursor. The keyboard in unresponsive, but I'm not sure if that because all the foam pads are dead, or because the machine still isn't in a state to receive anything from the keyboard yet. Not having the > prompt makes me think it's the latter. Anyway, baby steps forward.

Philip
 
You may be running into a bus conflict with a 64K RAM board. I'm not familiar with the Sol-20 in particular, but many of the S-100 boards I have will still turn on the bus transceivers even if the bank being addressed is not populated with RAM chips. This results in multiple devices placing data on the address bus, and that will usually prevent both devices from working.

In general, not getting a monitor prompt means that the monitor is not correctly loading, so the machine is probably at the latter state you suspected it was in.
 
Thanks for the reply,
I don't have any user RAM installed on the motherboard at present, so the only user RAM is on the card...Do you think this make any difference to your bus conflict hypothesis?

Next I think I may look at reading the SOLOS-1 ROM and see if it's healthy. Like you, I've been trying to find reference to the pinouts for the 9216 but haven't found anything conclusive.
 
WOW!
I didn't expect anyone to reply at all, and already thanks to everyone (especially Chuck(G) and billdeg)
??? Why would you be surprised? Let's just hope we can help you get this one solved a little more quickly than the PET - two years is a bit long ;-)
 
Thanks for the reply,
I've been trying to find reference to the pinouts for the 9216 but haven't found anything conclusive.

I was just looking at the personality module schematic on page 222 of the SOL manual here:

http://www.hartetechnologies.com/manuals/Processor Technology/SOL-20 Manual/SOL-20 All.pdf

It does mention 2708, 9216B and 8316E can be used (No mention of 2716). The personality module seems to have shipped with either two 2708's installed or one 9216. There is reference to this on page 82.

"One of two kits, using the same PC board: 2708-0 or 2708-1 may be supplied.
The 2708-0 version uses one 9216 masked ROM which has no window
on top of the IC package. The 2708-1 version uses two 2708 EPROM’s
which have windows."

This does seem to add some credibility to it possibly being 2716 compatible.


EDIT: I did some searching for 8316 ROMs, and this one does indeed seem compatible with the 2716. I found a few references to this such as:

http://www.vectorlist.org/Rasterlist/2002/08/0013.html
 
Last edited:
??? Why would you be surprised? Let's just hope we can help you get this one solved a little more quickly than the PET - two years is a bit long ;-)

No use rushing these things when there is so much fun to be had ;-)
 
No use rushing these things when there is so much fun to be had ;-)
Not to mention all the nice people you meet along the way...

I was just sorry to see that you'd apparently somehow gotten the impression that "S-100 folks" aren't every bit as friendly and ready to help (if they can) as any other area of this hobby.

Good luck!
 
Not to mention all the nice people you meet along the way...

I was just sorry to see that you'd apparently somehow gotten the impression that "S-100 folks" aren't every bit as friendly and ready to help (if they can) as any other area of this hobby.

Good luck!

That's true. Actually I didn't expect the S-100 folks to be any less friendly and helpful as the rest of the retro crowd. I was just a little wary of posting any stupid noob questions for fear of hate mail.
 
I tried reading the SOLOS 9216 ROM as a 2716, but all I got was FFs (and strangely enough the second time I tried to read it I got all FF's and just one location with an EF. Probably just a bit of programmer weirdness).
I also tried burning a 2716 EPROM with the SOLOS bin data (sourced from the SOL-20 windows emulator) and the machine did not boot to the flashing cursor, just some random characters that are actually exactly the same as when the personality module is not even installed. This makes me think that the machine is at least loading something from the SOLOS ROM. So no real progress on the fix today

I did also spend a while looking for information on the pinouts for the 9216 and have not been able to find them. I have however found reference to other ROMs that are said to be compatible.

Check out my page here for details:
http://web.me.com/lord_philip/other_computers/SOL-20.html#grid

Again, I'm not sure this is conclusive evidence.
 
Last edited:
Back
Top