help. any "old school" H9/H-8-5 experts out there?

[glenn.f...@gmail.com]

I was hoping to demonstrate the H9 terminal on my “pioneer” setup (ca. 1977/78) at VCF.  This would have been the heathkit terminal offering at the time (assuming you couldn’t afford an LA-36 DECwriter!).

 

I have my mostly “restored” H9 from last  winter

https://photos.app.goo.gl/c35iwjgdYuwrbzVRA

 

but I’m having trouble with serial communication. I have tentatively diagnosed this as some kind of loading issue on the serial port, but I’m certainly no expert.  I am trying to use this with the H8-5 serial port (which was the normal configuration at the time).

 

I have it set for EIA (RS-232) output and have checked all of the jumper configurations and believe I’m in sync with the manuals.

 

Here are three traces of the signal, all showing the letter ASCII “K”.  the first trace was taken with nothing connected to the terminal:

https://photos.app.goo.gl/vdj13p6mYfmNTZQ4A

 

the signal reaches +10.1V for “Space” (Logic 0) and -11.3V for “Mark” (Logic 1). Looks good.

 

When I connect the terminal (via null modem) to my PC and run TeraTerm I see the following trace:

https://photos.app.goo.gl/owRKz4B2hRnbg5h89

 

The “Space” level is pulled down now to +4.9V, but that is still within EIA specs and I am able to communicate two ways between the TeraTerm window and the H9.  Note however I must set my  TeraTerm configuration to 7 bits, no parity, one stop to get this to work.  This is odd because the I/O board in the H9 is definitely jumpered for 8-bit operation, but as I look at the waveform indeed it appears to NOT be sending the most significant bit.  For reference here’s what a “K” should look like:

https://upload.wikimedia.org/wikipedia/commons/b/b0/Rs232_oscilloscope_trace.svg

 

note in my traces the MSB “0” is missing?

 

Anyway, for the final trace here is what happens when the H9 is connected to the H8 via the H-8-5 serial port:

https://photos.app.goo.gl/pVEbgiL285oAijky6

 

note that the “Space” level is now only +2.5V, which is below spec (must be between 3 and 25).  Using the simple test procedure in the H -8-5 manual I get random jibberish when I try and read from the port. Also HDOS boots fail because I can’t even send the “B” for Boot!

 

FYI here is what the output section of the H9 looks like:

https://photos.app.goo.gl/KsSWJjoctiwU32oc6

 

Does anyone have any thoughts on this or have you possibly experienced this?  It looks like one could play with the resistor values to control the voltage levels?  But why should I have to?  the board is built to original specs?

 

I have actually tried this with two different H9 I/O boards (same results) and three H-8-5 boards. Do the LEDs in the 4N26s degrade in output over time? (they would all have to have degraded because I’ve tried many boards.)

 

I’m stumped…

 

One possibility is to set  the H9 to TTL output mode and then put TTL-to-EIA driver chips in, but that’s probably not something I would put on my plate now.

 

Tx for any thoughts/input…  if I can’t get this to work I probably won’t bother lugging it to VCF…

 

• Glenn

 

 

[Glenn Roberts]

P.s. the H8/H-8-5 works fine with my H19… so the issue would seem to lie with the H9

Sent from my iPad

On Sep 28, 2021, at 10:03 PM, glenn.f...@gmail.com wrote:



[Joseph Travis]

Hi Glenn,

The minimum RS-232 level is +/- 3vdc.  Judging from your o'scope photos, I would be looking at the AY-3-1015 USART and it's configuration.  Since it is all done manually, I'd double check the I/O board  jumpers and look for solder bridges or opens on the bottom of the board.  However before you dive into that, connect the RS-232 TX (EIA Serial Out) to RX (EIA Serial In), type at the keyboard and verify you're displaying what you're typing (online, half duplex).

Regards,

Joe Travis n6ypc

--
You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/26868FDF-E7B4-4D2D-84E6-4DE2CC24449C%40gmail.com.

[Norberto Collado]

Change jumper to use position 1-3 to use 1K resistor instead for pull-up.

 

Norberto

--

You received this message because you are subscribed to the Google Groups "SEBHC" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sebhc+un...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/081201d7b4d6%2433a56f80%249af04e80%24%40gmail.com.

[glenn.f...@gmail.com]

Norberto, Joe: thanks for the suggestions!.

 

The H9 is now working!  But it turns out that the RS232 level thing was a red herring.  I went back and carefully examined the cabling. I was using Heath-supplied cables to connect the H8 and H9 but perhaps one of the cables was originally wired for a different configuration? (DECwriter?).  the standard cabling requires three connections (Rx, Tx and GND) but this cable had a fourth connector installed.  It turns out that was providing a spurious input which worked its way all the way to the UART on the H-8-5.

 

Removing the mysterious fourth connection solved the problem and the H9 is now working.  The keyboard shroud must be installed just right or the keys bind on that so I’ll have to look at that.  Also I have some kind of “ghost” cursor on the screen but I think I’ve narrowed that down to a  handful of chips on one of the boards.

 

Anyway we’re “go” for the H9, which will make this a much more authentic “pioneer” (1977/78) exhibit.

 

I’ll post pictures soon.

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/66D59059-787F-4A87-AB05-98E60BA2E053%40koyado.com.

[Les Bird]

Glenn,

I love your H9 and your H8 exhibit. Looks so authentic!

When my dad brought home the H9 kit and built it I was blown away by how fast it was, at 600 baud (ha!). Then we cranked it up to 9600 baud (I think). Some time later he modified it with the 24 line mod (we never did the lowercase mod). After playing around with it I found a hack to do (sort of) direct cursor addressing on it. If you lock down the (I think AUTO CARRY) button and put the cursor at the beginning of the line (CR) and then do a backspace the cursor will carry up to the previous line. Doing a CR puts the cursor at the beginning of that line. Doing another backspace and CR will move the cursor up to the previous line again. Using this technique and keeping track of where the cursor should be I was able to make a light cycle game on it. I remember writing a nice assembly language library that I used to do all this for me so I could re-use it in other games.

Of course when we finally got the H19 the H9 got stored and forgotten for decades. I still have it but it's in pretty bad shape with a good amount of corrosion underneath where all the electronics are. :(

Still, lots of fond memories with the H9. Part of history. I would love to get another working H9 someday...

Les

[Glenn Roberts]

You were quite the “boy genius!”

 

I have an H9 from Ed Aumiller (“rescue” last summer).  I think he did every upgrade imaginable on this thing but it’s a rats nest of wire wrap and point-to-point solder patches.  Not sure I’ll ever get that working. but I think he did the 24 line mod and upper/lower case.  Looks like maybe an auto-repeat circuit.  But the stuff is all just hanging loose in the chassis.  If I ever work it it would be to try and restore the original functionality.  I did turn it on and I think the CRT and high voltage power are good so it would probably mostly be a matter of debugging the digital electronics.  Maybe if it’s a long winter…

 

The keyboard on this thing is heavily worn.  It was apparently a real workhorse back in the day…

To view this discussion on the web visit https://groups.google.com/d/msgid/sebhc/a991b0a6-0bd6-4472-8011-3241590f81efn%40googlegroups.com.