• Please review our updated Terms and Rules here

Martin Eberhard 88-2SIOJP-R100 only ? from AMON 2.8

A

AlbertS80

Experienced Member
Joined
Mar 3, 2022
Messages
50
I have an 88-2SIOJP-R100 (No Front Panel) card that I am having a bit of trouble with.
I am using this in an Altair system.
I get similar results with my Josh Bensadon backplane with a Altair CPU replica,
and with a very early Altair system (a V0 legacy CPU, 8K home brew card, and Martin's card)
I get the normal AMON start-up screen showing the AMON reserved area of memory that seems predictable.
But when I enter any command the only response from AMON is a ?
I am using 9600-N-2 with an FTDI USB to serial adapter.
I have got the same result with my setup with no front panel, and a system with a front panel.
(I lifted the no front panel 4.7K resistor and added a shunt so I could try both)
I did a Deoxit to both UARTS.
I swapped the UARTS
Any ideas would be appreciated.
Here is a bit of the frustration I am experiencing (with an 8K card):

AMON 2.8 by M. Eberhard
RAM: 1F00
>MT 1000 1
?
>du 1000 200
?
.
Thanks - Al
 
That looks like a command not found error.
I suggest putting that front panel to good use.
Write a little test program to receive a key-press and look at the bits. Do they match the ascii value of what you typed?
If not, is there a pattern? Bad bit somewhere, etc.?
Check the card receive circuit. Check levels with a probe or scope, or to save time, just shotgun it - replace all the chips and see if it works.
-J
 
Try this at 0000:

3E 03 D3 10
3E 11 D3 10
DB 10 0F D2 08 00
DB 11
D3 11
32 00 10
C3 08 00

This code will initialize the 2SIO for 9600 baud, and accept a keypress.
That key will then be echoed to the screen and the value stored in memory location 1000h.
Run it, press a key, and likely there will be no output on screen.
Stop the program, set address toggles to 1000hex, hit examine, and what do you see?

-J
 
Hi John,
Thanks for the thoughts.
I don't have the front panel on my home system, so I will have to bank your code until I get to the FP system this weekend.
My current thoughts are:
1. I get output data, so the data buss to the UART is good.
2. I need to scope the RS232 level shifter 1st - lighting up the scope now.
3. I will swap the CS to the two UARTs if #2 in inconclusive and see if the second serial port works any better.

Will let you know findings later.

Thanks again - Al
 
I originally hoped that someone would have said move JPx from 1-2 to 2-3 on this issue, but I get to delve further.
Accomplished #2 tonight.
The receiver to the UART seems to be working
I sent ZZ (0x5A) followed by carriage returns and that seems OK.
I will try my 8 bit logic probe tomorrow (donated to me by AJ the disk data recovery guy)
and see if the UART is outputting the wright or left data - still trying to decide how to sync the scope and 24MHz8CH probe;
maybe ext out of the scope and one channel of the probe and omit one data bit.
I usually like to add a tune to my posts, but not singing yet.
PS - I tried to post this same issue on groups.io AltairComputerClub, but I need to get approved.
anyone tell me the key to that kingdom?
2-ZZ-CR.jpg
 
Yes, if you don't have a front panel, a scope will do.
Looks like the rs-232 side is ok. And you're correct, next to check the bits on the ACIA output side.
Essentially doing what the front panel & program above does.

And just be patient... it takes some time to get approved on the Altair Club.
And the 6850 is called an ACIA. Intel has the UARTs. But they essentially serve the same purpose.
(that little twist of confusion brought to you by marketing people)
 
I had my logic analyzer on the ACIA selects and D0-D3 outputs.
It sure seemed to be outputting the 4 LSBs of "D" "U" and cr.
I was approved on the Altair club overnight and a message poped in from Martin.
I have to hang my head in shame, and smile a bit too.
The switches for EEPROM size were swapped around.
Once I corrected that the card seems to be working fine.
Thanks to Martin and you John for suggestions and prompting.
Johnny and I now have a bit more clarity
 
Doh!
Don't fret, I've been there before making goofs.
The last time was a few days ago when my cat snuck in and changed some front panel toggle switches by rubbing against them!
Kept trying things until I thought to check the switches!

Best of luck with the machine!
 
I'm gonna jump in since I'm also having issues with my 88-2SIOJP card relating to the ROM chip that I have not resolved.
For context there was damage to the card, but at least most of it seems to be fixed.

Right now I see this behavior:
  • If I have a ROM chip installed and enabled, RAM is disabled
    • If the ROM chip is removed or disabled via DIP switches, RAM works
    • If the entire card is disabled, RAM works
  • Taping over phantom an ISTSBN pins changes nothing, so neither of these seem to be being pulled low by the card
    • Both of these chip disable methods are disabled anyway
  • If I remove the ROM chip and run the code snipped @johnx993 mentioned earlier, serial works
    • So this tells me my serial interface is in fact functional
I believe then that something to do with the ROM is disabling RAM and thus AMON can't run properly.
I have checked the jumper and dip switch settings a billion times and everything seems right. So perhaps I'm missing something.

Does anyone have any ideas? Thanks!
 
I think Martin mentioned that addressing could be tricky.
Triple check the switches. And was it working before the accident (to rule out things like switches installed backwards, etc.)?
Manually test addressing via the front panel. RAM and ROM could be addressed to the same place.
Try depositing/ examining test values at the supposed ROM address with ROM removed. Can you write to it?
And try examining values at the supposed RAM address with ROM installed, is it code from the ROM?
 
Okay so this is interesting:

With no ROM: RAM behaves as normal. 0x0000 to 0xF7FF is writable
0xF800 to 0xFFFF is not.

With ROM:
ROM appears at 0xF800-0xFFFF
But for 0x0000 to 0xF7FF some bits of RAM actually are writeable, but others are not.

Maybe there's a stuck addressing bit somewhere and ROM is appearing in multiple places do to it.

Switches are definitely all set correctly. I never was able to verify operation prior to the incident. I actually did find some switches I used in place of jumpers behaved backwards compared to expected which caused issues. But they've since been replaced with regular jumpers. All DIP switches behave as I expect verified with a multimeter.

Anyway, I think this gives me a better idea of where to look! Thank you!

EDIT:
If I take RAM out completely, address bit 15 can be set to either 0 or 1 on the front panel, and the data is identical. Therefore, that is our culprit.
 
Last edited:
I seem to have root caused my issue. Since it seems to be damage rather than a board issue, I wrote my response in the original thread. But basically, U17 which handles the location of the ROM in memory is producing a weak output and is always being interpreted as high. This in turn prevents U12 from ever going high and disabling ROM.
 
@AlbertS80 Did you solve your issues?

One of the things I got tripped up by was the 88-2SIOJP like the 88-2SIO and the Turnkey board use a form of inverted signaling on their TTL interface compared to what an FTDI USB Serial adapter expect for say a Arduino, Raspberry Pi, or ESP32 might use. The FTDI and modern 5v and 3.3v TTL interfaces use positive logic signaling levels where a Mark is a positive voltage and a Space is 0v. So an idle transmit line sits at 5v or 3.3v.

The S-100 boards mentioned when using the TTL interface invert the TX out of the UART (like a 1488 would) so the TX idles at 0v and Spaces are +5v. This also lets the receive side for both TTL and RS-232 use the same hardware to do the 1489 accepting 0v as a Mark.

You can tell the FTDI chip to change polarity using FT Prog.
 
Hi Spiette, My problem was just a mix-up of switches.
Post 7 is my confession to issue.
Thanks for adding your notes for others that may have issues not related to my OE.
Al
 
AlbertS80 said:
My problem was just a mix-up of switches.
Click to expand...

That caused me a bit of head scratching and I hope I'll remember down the road why they work the way they do.

Here's how my board wound up, switches and connectors. It's replacing a Turnkey board in my Attache that talks to the video board so port 0 gets the Molex M156 connector and port 1 the 26 pin ribbon cable R100 treatment to a female DB-25 on the back.


finished.jpg
 
You must log in or register to reply here.
Back
Top