Altair question

[Erik]

I got this email over the weekend. Maybe someone here has some insight?

I believe you were the person I was talking with before. If not, could you direct me to a better source to answer this question?
Just purchased a pristine looking Altair 8800 (not the "a", not the "b", but the original). After replacing a bum capacitor, I have the fan working and the lights are on. As I test the power supplies and the bus voltages, ground seems to be ground, the +16V and the -16V are fine, but the 5V DC is almost 12 V (the connectors on each of the bus slots that are on the opposite end of the ground). I have 2 X 4K static Ram and 1 X 16K static Ram but the memory addresses don't light up. All the data lights are lit, like I have no memory attached. But it may be that none of the boards will like having a 12 V DC where they expect to see a 5V. I went back to the power supplies and found them all delivering the appropriate voltages except for one. The top most power supply in the chassis is delivering 12 Volts instead of the listed 8 volts.

Couple of questions:
1) Is the 50% over listed voltage potentially the reason why I am getting 12 volts on the top most bus connector and not 5?
2) Do you know where the supposed 8 volts from the power supply gets converted to 5 volts on the bus? or is "5" volts really just the 8 volts from the power supply?

 

[Terry Yager]

IIRC, the 7805 regulator (on-board) takes the 8v input and outputs it as 5v. (I could be full of it, OTOH).

<later>

Ok, so I cheated (looked it up). Pins 1 and 51 of the S-100 bus should have un-regulated 8v. (going to the regulator?). Some sources list these pins as being +5v. As near as I can tell, there is no +5v. output from the power supply (on the bus). Again, I could be wrong, YMMV.

When the input voltage is between 7.0 and 7.5 volts, the regulated output of a properly operating 7805 should be between 4.8 and 5.2 volts.

Hope this is of some help. If you need more info on testing the 7805, lemme know, I can go into more detail (requires a voltmeter and variable power supply).

--T

 

[hwhall]

Without seeing the hardware, here's my suspicion: I bet there was supposed to be a supply delivering 8v to a regulator that delivered 5v to the buss (about 3v overhead for a regulator is about right). This is probably the supply that was found to be delivering 12v instead of 8v. The regulator circuit has failed and the 12v vice 8v may be another symptom of that. If cards were left in the buss with 12v on the 5v lines, they may be toasted, resulting in the unresponsive LEDs that were observed.

 

[Terry Yager]

Yes, as I understand it, there is no regulation on the S-100 bus. Each card in the system has it's own regulator(s) on-board. You're prob'ly right about the mem boards being fried, tho if indeed you are getting 12 v. on a 5v. (or 8v.) rail. (With any luck, those circuits have some kinda protection in which a trivial component (like a voltage regulator?) burned out rather than the entire board). Some boards have actual fuses.

--T

 

[Terry Yager]

I checked a couple more sources, and can't seem to find any support for a +5v. signal on the (Altair) S-100 bus. There isn't any 12v. either, so the original poster probably has a problem with the Altair power supply outputting 12v. where it wants 8v. Of course, before IEEE-696 there were several variations of the S-100 bus, depending on the manufacturer. IEEE-696 standardizes (most of) the signals, although it still left certain pins un-defined, IIRC. These "extra" lines were sometimes used by manufacturers for thier own (proprietary) purposes.

Quoting from Herb Johnson's (Dr. S-100) webpage:

I think it is fair to say there are THREE major versions of the S-100 bus.

After the MITS Altair 8800, the company IMSAI produced its "IMSAI 8080"; several of the MITS bus signals were replaced or ignored on the IMSAI bus but the core set remained: two 8-bit unidirectional data lines, 16 address lines, the 8080 status and control signals, and other lines and power (unregulated positive and negative 8 volts, unregulated positive and negative 16 volts). Power was regulated on each S-100 card with individual voltage regulators: +8 became +5 for logic for instance. At about this point the name "S-100 bus" became popular. Examine this document for a list of IMSAI and Altair signals, as well as our bus lists in this document.

Various manufacturers used slight variations of the IMSAI S-100 bus; even MITS varied their bus for later Altair models. Most variations amounted to different signals to accomodate refresh or memory timing. There was a loose standard for managing banked memory (16 bit address = 64Kbyte memory) via I/O ports but there were other schemes as well. Many 8-bit processors were adapted to the S-100 bus in this period, but most systems used the Z80 or, later, the 8088 or 8086.

Then Compupro and others developed a faster bus, good to 10 MHZ, provided 24 address lines and an option for EITHER 2-8 bit data lines or one 16-bit bidirectional data lines. Other features include master/slave arbitration and bus mastering arbitration. This became an IEEE standard called IEEE-696 in the mid 1980's, about the time the S-100 bus began to decline in use. But the standard was in use by Compupro before that. They and others produced cards with 68020 processors and 80386 processors (or multiple processors) that out-performed the IBM-PC's of the era, up until roughly the late 1980's. We document a number of these later busses in the bus signal lists below.

Herb also has several different pin-outs on the same page:

http://njcc.com/~hjohnson/s100bus.html

None of them have 5- or 12-volts on the bus, although some cards used 12v. These usually converted the 16-volt into the 12 needed, thru a 7812 voltage regulator on-board. I don't think there should be 12v. going into the bus anywhere.

The "official" (IEE-696) standard may be found here (in .pdf format):

http://www.imsai.net/download/IEEE_696_1983.pdf

--T

 

[Terry Yager]

IEE-696 defines 9 "power" lines, all un-regulated:

2 x +8v.
1 x +16v.
1 x -16v.
5 x 0v. (grounds)

It also has 3 "NDEF" (not defined) lines, which may be used by the manufacturer in any way they see fit (at the cost of standardization, but they should be selectable by the user) and four more "reserved" lines, which are not to be used (yeah, right) for anything.

Of course, the Altair bus pre-dates 696, but I'd still look real hard at the Altair's P.S. for the fault.

--T

 

[alltare]

Re: Altair question

This sounds like the same computer that's being discussed over on the Yahoo group. Rather than repeat everything that's been said there, I would refer you to that site.

Steve
===============

I got this email over the weekend. Maybe someone here has some insight?

I believe you were the person I was talking with before. If not, could you direct me to a better source to answer this question?
Just purchased a pristine looking Altair 8800 (not the "a", not the "b", but the original). After replacing a bum capacitor, I have the fan working and the lights are on. As I test the power supplies and the bus voltages, ground seems to be ground, the +16V and the -16V are fine, but the 5V DC is almost 12 V (the connectors on each of the bus slots that are on the opposite end of the ground). I have 2 X 4K static Ram and 1 X 16K static Ram but the memory addresses don't light up. All the data lights are lit, like I have no memory attached. But it may be that none of the boards will like having a 12 V DC where they expect to see a 5V. I went back to the power supplies and found them all delivering the appropriate voltages except for one. The top most power supply in the chassis is delivering 12 Volts instead of the listed 8 volts.

Couple of questions:
1) Is the 50% over listed voltage potentially the reason why I am getting 12 volts on the top most bus connector and not 5?
2) Do you know where the supposed 8 volts from the power supply gets converted to 5 volts on the bus? or is "5" volts really just the 8 volts from the power supply?

 

[Terry Yager]

Re: Altair question

This sounds like the same computer that's being discussed over on the Yahoo group. Rather than repeat everything that's been said there, I would refer you to that site.

Steve
===============

Mebbe it would be more efficient to go to the Yahoo site and refer the original poster to this site, rather than re-typing everything to that group. Even if the post didn't come from the same person, the Yahoo poster might find this thread useful.

--T

 

[alltare]

Re: Altair question

Yes, I told him about this site a week or two ago. That's probably how Erik got the original message from him. He's probably reading these messages. Who knows? (Are you there, Doug?). I mentioned the other site mainly for others who might be interested in the problem.

It seems to be strictly a power supply malfunction, not an oddball buss. It sounds like someone might have miswired it.

Steve
===============
[/quote]

Mebbe it would be more efficient to go to the Yahoo site and refer the original poster to this site, rather than re-typing everything to that group. Even if the post didn't come from the same person, the Yahoo poster might find this thread useful.

--T[/quote]

 

[Terry Yager]

Re: Altair question

I mentioned the other site mainly for others who might be interested in the problem.

Steve
===============

Unfortunatly, what you didn't mention was the URL or even the name of the other group. Yahoo's got a lot of groups. (Is there an S-100-centric one that I'm not aware of?)

--T

EDIT: Never mind, I found the right group.

http://groups.yahoo.com/group/altaircomputerclub

Tnx for the pointer, I think I actually learned something from reading that discussion. (I most certainly gained some insight into my own power supply problem, which has been nagging me for a while).

--T

 

[alltare]

Re: Altair question

Sorry, it's the Altair Computer Club at
http://groups.yahoo.com/group/altaircomputerclub/

Steve
===============
[/quote]

Unfortunatly, what you didn't mention was the URL or even the name of the other group. Yahoo's got a lot of groups. (Is there an S-100-centric one that I'm not aware of?)

--T[/quote]