PDA

View Full Version : Altair Kit #33 Arrives


Pops Racer
12-23-2008, 02:48 AM
Hi all,

Altair Kit Serial #33 has landed in Australia. It arrived yesterday. Silly FedEx - my phone numbers were on the label, yet they could not contact me until Friday last week, after holding the package for duty since Monday last.

All is well that ends well. Couple of observations regarding the packing lists.

88-CPU Kit
1x Axial 100J - 10pf
component is marked 10J - which is 10pf - incorrect text
1x Axial 221J - 20pf
component is marked 20J - which is 20pf - incorrect text
12V Zener diode is marked 1N4742
5.1V Zener diode is marked 1N4733

88-DC kit
1x Axial 221J - 22pf
component is marked 22J - which is 22pf - incorrect text.

Fasteners
10x M3-12 phillps counter sink - missing.
#6-32 1 1/8 - missing
#6-32 1 1/4 - missing

Overall - excellent quality kit and manual. Not too worried about the screws, I can get those locally.

This is my Christmas present - so no working on it until the 25th! I plan to try and follow MITS assembly steps and refer to Grant's photo illustrated manual. It certainly captures the essence and flavor of 70's computing!

Have a Merry Christmas and Happy New Year. :)

Grant Stockly
12-23-2008, 05:32 AM
The 10 M3 counter sink screws were replaced by pan head screws. The packing list must not have been updated.

It does look like one of the parts on the BOM was written wrong. The 22p capacitor should have been listed as 220J, not 221J.

100J is the correct way to specify 10pf. They must have not had space to write all of the zeros. The capacitor code is two significant digits followed by the number of zeros.

You have gone through all of the screws 6-32s? Those are the ones that hold the motherboard down. I checked everything off the list as I packed the parts. I don't know how I could have missed those parts on yours. I'll check with the other people who have received theirs to make sure they have them. Those screws should fit in an envelope, so if anyone is missing them please let me know! :(

Thanks for the comments. I'll update the text on the diodes and capacitor to make it easier to understand.

Pops Racer
12-23-2008, 06:15 AM
Hi Grant,

All good, yep those axial caps are really small and the manufacturer did not bother printing the multiplier (0). I checked a couple of websites (ask Prof. Google!) and 22J is also ok for 22p, but yes they should be marked 220J. No problem! :)

Will recheck the screws tonight - There were screws in a bag attached to the outside of the motherboard envelope.

Thanks again and have a great Christmas & New Year

- Kelly

Pops Racer
12-24-2008, 09:01 PM
Re checked the hardware package. Only things I could not find were the 6-32 1 1/8 and 1 1/4 bolts. Everything else is there. Not sure they are critical components.

Cheers,
Kelly

Geoff Harrison
12-25-2008, 03:49 AM
Congratulations, you're in for a lot of fun over the next few weeks.

I plan to try and follow MITS assembly steps and refer to Grant's photo illustrated manual.


One quick observation: when Grant's instructions disagree with MITS's, make sure you give precedence to Grant's. I just wasted a day tracking down why my CPU wasn't working right and the problem turned out to be that I forgot to follow Grant's suggestion on P12 to replace R42 with a 4.3k instead of MITS's 6.2k. :(

Pops Racer
12-26-2008, 11:04 AM
Hi Geoff,

Yes it has been fun so far. 800 pins were soldered on the motherboard yesterday. I got the 18 slot with 8 slots supplied. It is a beauty!

Today, I worked on the D/C board - just partway through fitting the LEDs into the front panel. Every LED was tested before installing. I think I checked C7 and C8 5 times to make sure they were the right values!

I am definitely cross checking everything against Grants manual! It helps a lot. I like the way that Grant has preserved the board down to the mistake 'C' mark for SC13. ;)

Cheers
Kelly

Pops Racer
12-31-2008, 10:46 PM
Hi all,

Finished my Altair 8800 yesterday! Hurray! It would not run when first powered up, examine did not work, reset caused machine to free run. So I suspected there was a problem with D/C to Motherboard wiring.

Turned the unit over and there was the culprit! Solder bridge between two motherboard pins! :eek:

All good now. Got the 1024K board in and memory tested.

Here's some fun behavior for the 8080. I use the HLT instruction in simulation to stop the sim and let me know that the code is finished. Ha! Doing it on a real 8080 causes the cpu to enter an 'undefined' state. Stumped me for a while.

From the manual:- "Operation: Implementation of the HLT instruction steps the PC to the next instruction address and stops the computer until an interrupt occurs. The HLT instruction should not normally be implemented when a DI instruction has been executed. Since the DI instruction causes the computer to ignore interrupts, the computer will not operate again until the main power switch is turned off and then back on."

The default power up state is interrupts disabled. (INTE LED is OFF), so the HLT really freezes the computer :D - But in the Arrowhead Tips section (page 5) it says, "To reset the computer you have to hold the stop switch raised while raising the reset switch. Release the reset switch first. (No, we don't know why, but it's traditional!)"

I guess it is a quirk of the D/C design and the 8080 CPU (Looks like assert PRDY and toggle PRESET causes the CPU to resume normal operation?)

Anyways - heaps of fun just mucking around with some of the basic instructions. Tried reading the sense switches with IN A,377 and then storing the data somewhere. Works really well. Good thing the jumper was added to 'read' the sense switches! :)

Moving on to the 88-4MCS and 2SIO using only original MITS documentation. :cool:

Kudos to Grant for this amazing kit.

Cheers,
Kelly

Grant Stockly
01-01-2009, 10:09 PM
If you have an AM radio, you may want to consider trying this program out.

http://www.digibarn.com/collections/weirdstuff/altair-sheetmusic/index.html

Look at the bottom for the "sheet music" scans.

Pops Racer
01-02-2009, 11:54 PM
If you have an AM radio, you may want to consider trying this program out.

http://www.digibarn.com/collections/weirdstuff/altair-sheetmusic/index.html

Look at the bottom for the "sheet music" scans.

Oh yeah! :D I did that yesterday - toggled in both songs and got an old cheap am radio - very cool. :cool:

Now trying to figure out how to connect 2SIO to PC. I suspect both circuits are DTE and so must be 'crossed-over' (null-modem) TX -> RX, etc. Do you use DTR and RTS?

Cheers
Kelly

Grant Stockly
01-03-2009, 12:28 AM
I didn't wire any of the handshaking lines on mine. I think hardware handshaking has to be enabled in software.

Here are notes on the 2SIO from a while back.

http://www.stockly.com/forums/showthread.php?t=507&highlight=2SIO

Geoff Harrison
01-03-2009, 12:36 AM
Now trying to figure out how to connect 2SIO to PC. I suspect both circuits are DTE and so must be 'crossed-over' (null-modem) TX -> RX, etc. Do you use DTR and RTS?


If you wire the port as per the MITS manual you end up with an odd hybrid DTE/DCE port. Tx and Rx are set up as though it is a DCE, but it uses the control lines as though it is a DTE. The 6850 requires CTS and DCD to be asserted, so you either have to tie them high in your cable, or give up on hardware handshaking and tie pins 23 & 24 to ground on the 6850 (remember, the signals are inverted on the board), in which case a straight cable will work fine using just the data lines and ground.

Geoff.

Pops Racer
01-03-2009, 04:34 AM
If you wire the port as per the MITS manual you end up with an odd hybrid DTE/DCE port. Tx and Rx are set up as though it is a DCE, but it uses the control lines as though it is a DTE. The 6850 requires CTS and DCD to be asserted, so you either have to tie them high in your cable, or give up on hardware handshaking and tie pins 23 & 24 to ground on the 6850 (remember, the signals are inverted on the board), in which case a straight cable will work fine using just the data lines and ground.

Geoff.

Cool - I have wired it the way that Grant descibes and when I run the echo test I get continuous spaces on the terminal and nothing shows when I type.

Have the board strapped for I/O 20/21 octal. I tried changing the input from port 21 to input from sense switches (port 377) and voila! the transmit works great.

The receive does not work - When single stepping, the IN 020 gets a 377 octal - why would the input from the control register return all 1s?

Pops Racer
01-03-2009, 07:05 AM
I did some digging to find out why the IN 020 was giving me 377 (all ones).

I thought there might be a problem with the 8T97 buffers and the SINP signal. So out with the cheapo 10MHz scope (that was a good investment for 150 bucks) and start probing. Hmmm - SINP working on IC S - all good.

But the OUT 021 was working perfectly. hmmm - some overlap on the 8T97 chips between Data In and Data Out - maybe it was the chip enable coming out of IC P (7400) - lets go check pin 8 on IC P - :o :eek: The only pin on the whole board that I did not solder!!! arrrgh! :mad:

Out with the card - solder pin 8 on IC P. Card back in, hook up computer to serial. Develop more finger calluses toggling in the echo program. Voila! It works! Now on to loading Basic. I've got 5K installed, what will happen with 4K basic? I guess the memory space is set by the loader? Or is is hardcoded in the basic tape?

Thanks for all the help :D :cool:

Cheers
Kelly

Grant Stockly
01-03-2009, 10:41 PM
Did you have any issues with mounting the motherboard? Where are the pictures? :)