best way to add the back plane connections?

[John Galt]

What is the best way to Add the 25 connections to the DUE?

are most without the newer main boards just soldering directly to the DUE?

do i have to worry about electrical noise and shielding?

just wondering how others have done it.

Thank you

[John Kennedy]

I soldered wires directly to the Duo in order to interface with the floppy interface. 

I do NOT recommend this approach. 

It worked but it’s a fragile mess. 

I *think* someone made a PCB daughterboard that broke out all the connections. That would be smart. Maybe Adwaterandstir could make something similar for those of us without the systems that have  the connection already onboard?

[John Galt]

i was also worried about replacing the DUE in the future if the need came up and having to resolder everything.

I thought about soldering to the front but that seems a larger mess.

i'm waiting for the new reproduction case and any new PCBS that David makes with the Back plane so i didn't want to start changing out boards on something temporary.

well temporary but not really.

i plan to have the backplane external and connect everything via a short shielded 25 to 25 pin connector.

[Brian Welland]

Hi all,

This is the gerber file for a suitable PCB as suggested above.

--
You received this message because you are subscribed to the Google Groups "Altair-Duino" group.
To unsubscribe from this group and stop receiving emails from it, send an email to altair-duino...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/altair-duino/ceee382f-601f-426d-bf77-034dd8eae9fan%40googlegroups.com.

[Patrick Jackson]

Thanks a ton for the gerber! I'll send this out, and work on a new internal cable. My method of direct soldering has completely failed XD

[John Galt]

Now I can see why people had a hell of a time soldering connections to the due board.

It took me a few hours to make a nice easy to read layout of where to solder the wires as the 8 data lines are not straight forward on the DUE

Also deciding if I should add a power switch to the backplane separately.

[John Kennedy]

Hi - do you have a photo of the breakout PCB attached to an Arduino?

I have ordered some boards from your gerber files, thank you!

[John Galt]

the side with the outline in white gets the DUE plugged in and the 25pin connector. the other side that is blank goes to the front panel board.

it has to because the 25BD connector has to stick out and point backward away from the front panel so you can plug into it with the DUE next to it.

i don't know what pass through connectors are required. i decided to solder to the DUE pins, since it would be another few weeks of getting the pcb made and then sourcing all the components.

the breakout PCB helped me understand what pins i needed to use, as it is very confusing without really laying out where things go.

[John Kennedy]

Yeah took me two goes to get all the right pins soldered. 

It looks so bad I’m having make the PCB to avoid the shame. 

[John Galt]

mine will be hidden inside the main case, i made an external case for the backplane. I already modified my DUE with a bunch of other external wires, figured oh well.

I put the LED Output Register card in the first slot so i could show output from the front panel 
the disk controller has side USB port access.

[John Kennedy]

I got my PCBs back to add a 26 way connector to the system, but I must admit I can't immediately see how to use the board with the existing arrangement of pins and sockets.

A photograph of the PCB in a working system would be very helpful!

[John Galt]

in order to use that pcb you have to source a bunch of the pass through connectors.

i felt it was not going to go well so i went with the direct method of soldering a ribbon cable to 25 pin.
the 26th pin is not used, its listed as just ground but its left open.

the printed side with the white text is where the DUE is plugged in. the black back side goes to the Altair front panel PCB.

the white 26 pin is where the 25 db connector goes in and sticks out the back towards the DUE

here is a PDF i made of the pin outs and where they go on the DUE. 

soldering all 25 connections to the DUE was tedious but it only took a few hours working in stages, i did the misc wires first then grouped the Data connections and then the address connections.

[John Galt]

oh i also took 3.3V, 5.0V, VIN(RAW) and Ground from the main Altair PCB panel on the right hand side for convenience and power connection ease.

[John Kennedy]

Hmm. Looks lIke the 26 pin connector will block access to the USB and power inputs. Would have been better if the connector was on the other side, or at the top or bottom. 

[John Galt]

Depends on the height of the pass through connectors, issue for me is I can’t get components, I figured after I ordered the pcb then I would spend weeks trying to find the right connectors verse just spending a few hours soldering and then I could route everything myself.  I also felt the 25 db pin was too close, some people instead of putting the 25db there have a 25db on a ribbon cable and solder the ribbon cable into that spot to make clearance run the 25db out the back panel. 

[John Kennedy]

I built the breakout board, and the AltarDuino is still wokrking!

Before I test it with the disk interface card I need to confirm the ribbon cable orientation. 

Q. There is a jumper on the breakout board. What’s the purpose? Open or closed for floppy interface?

[John Galt]

I believe that 2 pin is for a power input I traced it 2 weeks ago but forgot exactly where it hooks up.

I remember one side was ground and I think the other went to raw voltage.

Don’t propagate those 2 pins and don’t use them. 
 

The have regulation on the aurdino side and the backplane side already.

[John Galt]

Also now is the time to install the Rx/Tx modification on the due.

For another serial port.

Use 3.3v for that new serial not 5v learned that.

Then you wire in a 9db pin converter board and

Flash the firmware for the new port and enjoy I use mine for the fabgl ansi terminal as primary

Which is a better vt100 terminal with 64 colors and more graphics capabilities over the Geoff terminal.

[John Kennedy]

I need to get everything working first.. after transitioning from the rats nest of wires to the PCB, everything has stopped working. Two of the pins in the central cluster of six accidentally got shoved into the same socket, and I had to refresh the Arduino to get it working.

The disk drive controller card seems completely dead, and I can't even get 5v to 3.3v at the IC sockets on that card. That was my weekend.

[John Galt]

ouch.

:(

[Patrick Jackson]

John Galt, are you saying that there's another terminal firmware for the included terminal board that has more features than the default setup?

Also, I removed the ratsnest I had for the gpio breakout and I instead installed one of the PCBs that break out the 25 lines.

However, I get the exact same behavior has before, even after double and triple checking my wiring! I get random LEDs on the main Altair board turning on/dimming and I get hot chips on the led register board!

[Patrick Jackson]

Also, its this version of led card:
https://adwaterandstir.com/product/led-output-register/

[Patrick Jackson]

The only chip that I slot into the led register board that DOESN'T cause problems is the 74HC00 chip in the top left. Adding any other chip causes some LEDs to turn on, dim, or turn on only when certain addresses are selected.

[John Kennedy]

Is your Duo flashed with a build of the Altair firmware that includes IO support?

[John Kennedy]

yes, the designer of the board confirmed that those pins are not a jumper but a location for external power.

[John Kennedy]

Very ouch. I've ordered one of the LED IO boards, just to confirm that the breakout board is working. If that's all good, I will either continue diagnosing the floppy driver or get a replacement.

[John Galt]

There is a fixed version for the Geoff Terminal that i put out.

the Geoff Terminals Graphics modes do not have a function to erase one element at a time.

So for example If you draw a pixel you have to clear the entire screen to remove it which is very slow.

I added additional escape codes so you could

   

place a pixel, erase a pixel

place a circle, erase a circle

place a circle fill, erase a circle fill

place a box, erase a box

place a box fill, erase a box fill.

 

Draw a line ESC [Z1;;;;Z 

Draw a box ESC [Z2;;;;Z 

Draw a filled box ESC [Z3;;;;Z 

Draw a circle ESC [Z4;;;Z 

Draw a filled circle ESC [Z5;;;Z   

Erase a line ESC [Z6;;;;Z 

Erase a box ESC [Z7;;;;Z 

Erase a filled box ESC [Z8;;;;Z 

Erase a circle ESC [Z9;;;Z 

Erase a filled circle ESC [Z10;;;Z   

if you have the patch on your terminal it will show my name as a john galt graphics patch on start up.

The problem with the Geoff terminal is VT-100 support is terrible.

the board included with the pro also supports the SD card and the serial port so you want to keep them.

a much better terminal exists in the FABGL VGA32 Ansi Terminal which also has full VT-100 support in addition it supports many other vintage terminals.

the terminal also supports 64 colors characters and graphics including Sprite and sound support.

in order to use it you need to use up a Serial port, the stock Altairduino pro comes with one 9pin Serial port. the USB ports provided are slave ports.

in order to really use them with external devices outside of a laptop or desk top you have to construct a USB server to 9DB, which i did create, however its not perfect and has issues with many usb devices.

the best solution is to Hack the DUE main board and install another serial port using the TX and RX Leds, then wire up ground and 3.3V power to a 9DB converter, update the Firmeware to support the new serial port

you will then have two 9DB serial ports you can map how you like. i installed the FABGL Ansi Terminal on the new second 9DB serial port and mapped it as primary.

now i have color and graphics support along with full VT-100 escape code support. this allows me me to view color pictures on my Altair and go to BBS over my Wifi modem and see the ASCII Art in color.

i can also view HTML websites and see color pictures through IMP 2.45.

I stopped using my Geoff terminal about 3 years ago now.

[John Galt]

====

Also, I removed the ratsnest I had for the gpio breakout and I instead installed one of the PCBs that break out the 25 lines.

However, I get the exact same behavior has before, even after double and triple checking my wiring! I get random LEDs on the main Altair board turning on/dimming and I get hot chips on the led register board!

====

it is possible your DUE was damaged when your wiring was a little off from before.

would have to swap in a known good DUE and see if the behavior is the same. 

start working from there, as you could still have wiring issues somewhere.

in order to keep myself organized i used a 25DB break out board with all the pins labeled for me.

i then double checked continuity between all due and 25DB pins.

the DUE pins will all read continuity which is annoying but you can still check you have the correct wire from the DUE going to the correct DB25 pin.

I also 3d printed a bunch of things for my PRO case to make sure everything was organized 

and i created an external Backplane box with the led registers visible from the front. i can program it as a display as well.

[John Galt]

i wrote a little basic program that blinks the LED registers as a test i used it as a first test to make sure everything was wired correctly.

i can post it later.

[Patrick Jackson]

Thanks for the input (heh)!

I'll swap out the Due, I got a few kicking around...

Also, one of the chips gets super hot, which is a bit upsetting. I may have killed it when I did a direct solder job. I also wonder about the flux messing things up, but I'll do a new Due just in case.

I'll also take a look at the terminal update and a fabGL board! Looks absolutely INCREDIBLE, jpgs on an Altair! (well....kinda XD )

I'll mess around with it tonight

[John Galt]

you might also suspect something is wrong with the LED register card. if you pull it out and a chip on the DUE is still hot then you could suspect the DUE if the chip is colder then could be something on that card.

you could run continuity checks on each line on the backplane to see if something is shorted.

Flux can cause shorts its a good idea so scrub down the flux with IPA and rule it out. 

inspect your boards for extra solder that went flying during your work. a random blob could have fell into a board and caused a short.

i should start using my GITHUB... i posted my version of the PPMVIEWer with dithering support and printer support for Okidata printers.

i stopped working on it 2 years ago because i got side tracked by too many projects.

for the information on viewing HTML

https://github.com/johngalt01/johngalt01.github.io

https://forum.vcfed.org/index.php?threads/viewing-html-on-a-s-100-machine-with-pictures.1245743/

[John Galt]

if you want to roll the lights on the leg register card

10 ' MAKE THE FRONT LIGHTS BLINK
20 '1,2,4,8,16,32,64,128
30 WHILE INKEY$=""
40 OUT 255,1:GOSUB 300
50 OUT 255,2:GOSUB 300
60 OUT 255,4:GOSUB 300
70 OUT 255,8:GOSUB 300
80 OUT 255,16:GOSUB 300
90 OUT 255,32:GOSUB 300
100 OUT 255,64:GOSUB 300
110 OUT 255,128:GOSUB 300
120 WEND
125 OUT 255,0
130 END

[Walt Perko]

Hi, 

I posted a few BASIC programs to do different designs with the LEDs on the expansion backplane LED board.  

http://www.brainless.org/Altair/Repository.html#section4

.

[Patrick Jackson]

Ah, the Due chip doesn't get hot, its the chip below the nand gate that gets hot

[John Galt]

Check inside the P1 connector to make sure no solder blobs got in there.

[Patrick Jackson]

I may need to get a few more of those chips. The solder all looks clean. Everything else works when that card isn't plugged in!

I'll try flashing the due with the io software again. Pretty sure I already did...

[Patrick Jackson]

Same behavior. Let me try to clean the flux again.
Are we sure the Due is bad? It works 100% every other way, leds, sd card, serial ports

[John Galt]

 if everything works except the led register card then it has to be something on that card.

or something conflicting with the card.

If you pull all the cards out of the backplane and just use the LED register card alone, does it have the same issues?

[John Galt]

    74HC04 or MC74HC273A could have been damaged. those use address lines and Data lines. if chip blew internally then it could be selecting more then one LED and that is showing as dim lights and trying to light more then one at the same time.

[Patrick Jackson]

Yeah, I only have the LED register card at all, no other cards. I'll look for chips to order now.

My goal is to make a matrox 256 emulated card so I can try and do CAD work on an 8080 XD

[John Galt]

ok, well a matrox 256 emulated card is not going to work with this backplane. 

the design states that you do not have memory address access with this design.

https://github.com/dhansel/Altair8800-IOBus/tree/main

• No RAM/ROM access

  The I/O bus works only for INP/OUT instructions. It is not possible to intercept memory read/write operations at this point.

in order to have a video card you need direct memory access. the Dazzler setup is currently the only way to do this and it has a max resolution of 128x128 1 bit color. or 64x64 in basically 14 colors, or 32x32 in 14 colors.

currently you have 2 options with a Dazzler, either emulated or a hardware dazzler.

now as far as cad work is concerned. your going to basically make the space ship from Elite. which i have done with the dazzler. i have written a 3d and 2d engine for the dazzler along with some display functions.

still far from 100% complete as i have other things going on. your going to have very little screen space to show anything.

verse you can run a FABGL terminal and you could draw on that, then you have to take a page out of the C64 programming manual and create a simulated screen memory space which with 64KB and less then 54KB of TPA under Cp/M

you could emulate a 320x200 kaypro monitor graphics setup easy would be again 1 bit color, otherwise you need a separate color look up table again like the C64 and your limited to just a few colors of the available 64

all because of the memory limitations. at least on newer Z80 processors you had memory page switching.

anyway if you don't care about output or saving the information then you can just pump escape codes to the terminal and have it draw the ACAD R12 spaceshuttle.

since its a terminal animation is a major problem you don't have multiple video pages to access and so you would need to erase and redraw the screen when you reposition X,Y,Z, well thats how ACAD used to roll.

getting back to your issue. 

it is obviously half working since when you change the data sent to the port the leds change... you may want to again use a continuity tester and check your connections. remove all the chips and start checking.

keep in mind its possible you did not fully solder a VIA connection from one side of the board to the other. i use a set of magnification glasses and i look right in at all my solder connections. just soldering the back side of the board solid 

does not mean your fully connected on the opposite side. many of the connections are also buried under the sockets and caps and so its hard to see.

this also assumes you have the back plane 100% correct as well.

 

[John Galt]

i should say it might be possible to modify the hardware dazzler and create a 256x256 version maybe with continuous memory instead of the Quadrants setup used in the I and II versions.

however that is a whole other can of worms with PIC programming and what ever the limitations of implementing the current dazzler through the USB port was.

same with implementing a Matrox Alt-256 s-100 card.

[Patrick Jackson]

I'm gonna have to get back to you in the morning (4 hours or so XD ) before I head to work, I'll rescrub the due, backplane, and card with alcohol and do a power continuity test and then try each pin. Failing all that, new chips will come in a couple weeks.

[Chris Davis]

I believe I did have an LED card where one of the chips was overheating.  It was a long time ago, but if I remember correctly it was because I had the wrong chip in the wrong socket.

Here is the version I sell on my site, so if that's where you got it, compare the placement of chips to David's design: https://github.com/dhansel/Altair8800-IOBus/tree/40cab48cd537027da8362cdc8aa4c64a79aec559/01-led-output-register

[Chris Davis]

That was the wrong link.  Here is the correct one:
https://github.com/dhansel/Altair8800-IOBus/tree/a597c041477d7179d8b6ca24c0cc832d4bb852d3/01-led-output-register

[John Kennedy]

I just finished building the LED board to make sure that I had things wired up (the drive controller board I was using was not working).

It doesn’t work. In fact, the 5v pin on the LED board is only 3.3 volts.

I am at a loss. Is the version of the AltairDuino I have somehow got a different design that prevents the breakout board from working?

[John Kennedy]

oh wait a minute.. now I am worried that the ribbon cable connector that comes from the breakout board does not present the pins in the same order as the LED board expects..

Has anyone actually used a breakout board to connect directly to LED or disk boards? I may have assumed too much here.

[John Galt]

No the breakout board is correct I used it to help make my pdf of where all the connections go

I used the placement for my wires I soldered directly to the due pcb.

Take a continuity tester and start tracing everything make sure your pins all match up from the pdf I posted 

Check you getting full contact from the top of your pcb stack all the way to the front panel board.this was a fear of mine in making a multi stack of pin connections.

Use a magnifying glass and make sure no bridges anywhere to be sure.

 My Altair kit was from 2020. I have Also modified my original pcb in a few ways to add things.

On your backplane board are you using the external 5v regulator? 

Have you flashed the firmware on the due with the new Altair ino and enabled the backplane?

[John Kennedy]

Oh thank goodness. Thought I was going mad. Resolved the issue.

The ribbon cable I was using - or possibly I soldered the breakout pins on the board in different orientation - was physically inverting the pins. Swapped things around using a second ribbon cable in-line, and the LED card is working. Now I can try the disk drive card.

[John Kennedy]

The disk card is partly working - I can switch on/off the motor, but it never detects the presence of a disk.

It’s possible I’ve damaged a chip with all my tinkering, so I’ll keep testing.

[John Galt]

the issue you ran into with the pins on the 25db was the reason i used a 25db breakout box with the pins all numbered. i wanted to make sure i didn't reverse anything as it is really easy to do that as you ran into.

it makes things a little messy but you clean it up with some zip ties and its all hidden inside the box anyway.

[John Kennedy]

Note: Now that I have things mostly working, I have a few spare breakout PCBs (the board that plugs between the Arduino and the older AltairDuino board to make accessing expansion options easier than soldering wires to the Arduino). Anyone got any other PCBs to swap? (Ideally expansion boards other than the disk drive and LED board). 

[John Galt]

I would recommend the rtc card.

I’m putting together the speech card. As the next project.

[Patrick Jackson]

Once these chips come in, I want the LED card on the front, the speech card/sound card, the rtc card, and a software defined card (pi pico). RTC would be AWESOME for real RNG and such!
I think a custom graphics card would be fun too, possibly on the software defined card. More like a vector graphics card, no memory access, just port accesses

[John Galt]

some points:

===

I think a custom graphics card would be fun too, possibly on the software defined card. More like a vector graphics card, no memory access, just port accesses

===

you already have this without the backplane. the Geoff terminal already supports vector graphics.

you can install a ESP32 Vga32 Ansi terminal and get 64 colors and even more vector graphics. this card also supports tons of native terminals allows additional applications for other Cp/M machines to work on your altair.

without direct memory access your just looking at a terminal setup with one character at a time over serial communications.

----

===

Once these chips come in, I want the LED card on the front, the speech card/sound card

===

The speech card is not a sound card, the SP0256A-Al2 is a chip that holds speech Phonics, the design for this backplane does not use a CTS256A-AL2 processor chip. 

a speech chip like this will spit out pieces of human speech that have to be strung together with over 430 sets of rules. the CTS256A-AL2 is a pretty rare chip but it handles those 430+ rules for you

and puts the setup on a serial com port interface. this allows you to send a text sentence over serial where the CTS encodes TEXT to Phonics and then sends output to the SP0246 synthesizer.

for Cp/m no-one every made a Text to phonics translator (i'm working on one) there are online translators and some code around on GitHub but non of it is really applicable.

you can send pre-converted Text to Phonics to the chip, the original manual for the chip includes a small vocabulary, this is also the same limitation as the SpeechJet  chip that was all over a few years ago.

its complex but moving on...

because you already have a sound card if you again use a ESP32 VGA32 Ansi terminal.

it supports musical notes via escape codes and people have made music players using it.

the Ansi terminal also supports mouse if you wish to start creating applications that use the mouse on the altair.

----

The RTC is a great piece of kit.

the RTC that comes with the Altair-duino just pulses based on your settings. this can be setup on the configurations menu

this clock is not accurate you will find the closes rate it offers are either slightly too slow or much too fast to accurately time a second. 

thus any clock you make will be WAY off even after a few minutes of operation. it also requires a system interrupt and inserting your code around the O/S which is a pain because a warm boot or loading a program will mess it up as there is 

no good room to keep things protected. 

in the end you end up with a clock worse then the Commodore using Jiffys.

The backplane RTC board is the best as it avoids all this and keeps accurate time on a battery back up. the clock does tent to corrupts itself every few days unless you leave your system on all the time.

on Cold boot it is best to reset the time. similar to when you booted MS-dos each day.

this clock is accessed via the serial port and so you don't have to worry about a program or o/s messing it up.

however it is of limited use directly with Cp/m as it does not support clock functions. but its great for your own program development and also as a random number generator.

i put out a few RTC clock applications to get people started.

its also good if you want to accurately time a program or your code but its limited to Seconds as the resolution can not be made smaller. 

a jiffy type clock can be good to 1/100 of a second.

in short you can enjoy graphics and sound now without waiting. i would recommend the Tx/rx hack for the Arduino DUE which will get you another serial port to use then plug in a Ansi Terminal as your primary and you can enjoy HTML graphics and pictures on your screen

[John Kennedy]

“The side with the outline in white gets the DUE plugged in and the 25 pin connector”

That isn’t possible with the PCB I have. The only it can go together is in the image below (yes, I’m soldering up a second breakout board)

[John Galt]

I guess the way the board is designed that is correct.

hopefully you get it working.

[John Kennedy]

Has anyone actually used this breakout board?

[John Galt]

I did not use this breakout board directly. I used the PCB layout to confirm where to solder the wires to the DUE board directly.

the breakout board is wired correctly to the DUE. 

[Mike Tailor]

As far as the Gerber is concerned, it seems to be the same source as I realized half a year ago.
But I soldered the adapter board from above, which I find more stable.

https://groups.google.com/g/altair-duino/c/inLKss1sNag

[John Galt]

johntk that setup you had with the wires was scary looking.

my approach was i used ribbon cable and organized the wires so that address and data were on separate ribbon cables. then a 3rd ribbon cable handled the secondary signals. and a 4th ribbon cable handled the power and ground.

i then used a 25DB breakout box with labeled screw terminals and setup an external 25DB connector from the back of the altair pro case. 

i made a box for the I/O Bus and had the 25DB connector come out the back. then used a 25DB to 25DB rainbow ribbon cable(for looks) to make the connection between the back of the PRO and the I/O backplane.

that kept everything very neat and safe from accidental shorts.

[John Kennedy]

I found the original creator of the PCB. He hasn't actually used it either - it was for a friend.

They used it like this - which is NOT how I expected it to be mounted. Looks like it is soldered directly to the Arduino.

Anyway, I think it's possible that the floppy disk drives and/or cable I am trying to use at fault. After all, the LED expansion card works.

Time to find a known-good drive/cable combo somewhere.

[John Galt]

this is the cable i used.

https://www.amazon.com/dp/B00FZ3IL80

it has the correct twist for drive 1

it also has both style of connectors for both drive 1 and 2.

so you can use the 5.25" or 3.5"

[John Kennedy]

Thanks - just got it in order to rule that out.

[Mike Tailor]

Yes, exactly... and that friend was me. I wanted to expand my old Altair revision with the bus.
I showed it here:  https://groups.google.com/g/altair-duino/c/inLKss1sNag
Personally, I didn't find the intermediate adapters that practical, and simply soldering them onto the Arduino is much easier and an Arduino doesn't cost the world.

[John Kennedy]

So the PCB just sits on the base of the Arduino, and you drop solder through the PCB holes or something to make contact with each pin?

[Mike Tailor]

Yes.. that's how I did it... originally you were supposed to mount it in between, but I find that very complicated..
You should just think about the side connector beforehand.
There are only a few soldering points... I've already shown you.

[John Kennedy]

Thanks - I'm trying that cable now. Can you possibly confirm the jumpers used on your TEAC 55 drive? Occurred to me I might have experimented with them and not set them back.

[John Galt]

Mine are all defaults

If you look on eBay any drive you can spot the jumpers those are defaults

The only issue is when you have a second drive getting them to match can be a pain so you can read disks in both drives.

4 drives even harder to setup.

[John Galt]

[Tom Lake]

If you're using a straight-through cable, then once you set the rotation speed, head loads/unloads, and door LED jumpers*, then all you need to change on individual drives is the jumper marked DS0 through DS3. This jumper selects which drive it is, 0 through 3. 

Using four drives is not hard at all if you use the same brand and model for each drive.

* Settings found in the FDC+ manual, section 3.

On Sunday, August 18, 2024 at 12:17:41 AM UTC-4 John Galt wrote:

[John Galt]

“Using four drives is not hard at all if you use the same brand and model for each drive.”

Of the four drives I setup.

3 red led

1 green led

1 extra older red led

None of the drives excepted disks from each other.

All required adjusting the track pcb forward and backwards until it was able to read a disk from another machine.

This too I me hours to balance the drives 

1 red led drive refused to adjust no mater what I did, that drive ended up in my msdos machine

And works fine.

The io busdisk controller didn’t like it.

When using the straight cable I found my drives would acknowledge a disk insert once then refuse 

To change disks unless I powered down the drive this was after messing with the jumpers

Only 3.5” disk drives I tried were ok with this the 5.25” drives had a fit unless I used the twist cable.

Just my experience with this.

My current setup uses just 2 drives however one of the modified drives has been taken to a second machine so I can share disks.

I suspect trading disks with people will not work as everyone’s drives will be slightly different as I experienced.

[John Kennedy]

Thank you - I appreciate you taking the time to photograph your jumpers!

I had some success when I connected a 3.5" drive, but not with the 5.25" drives I have. Hope to pick up another 5.25" and try that. Before giving up :-)

[John Galt]

Just to make sure did you use the 5.25" drives with the cable that has the twist in it?

the 3.5" drives do not care about the twist, but the 5.25" need them to work properly. even if you change the jumpers if your trying to compensate for a non-twisted cable it will not work right with the 55G drive.

the drive will not detect a disk insert and may continuously spin the drive motor with the activity light on.

you can change the jumper for disk detect and force it to recognize a disk, and it will but then you can not change disks the buffer on the disk will not change unless you power down the drive and restart it with the new disk.

a lot of weirdness is due to not using the twisted cable for the first drive. hooking up the second drive with the untwisted cable can work if you setup to have a hard drive on A: and B: then C: will not work but D: will when setup in the configuration menu for no drives mapped for the C: and D: drives under cp/m. you can't map one real and one virtual drive it has to be either both not mapped or set to see a virtual disk. 

[John Kennedy]

The long national nightmare is over: my AltairDuino just booted CP/M from a 5.25” floppy disk, using the backplane adaptor.

In the end the issue was the floppy drive itself! I found another drive today, and everything worked first time.

I’ve no idea why the several other drives that I have stopped working. Bad luck (or worse) I guess.

Thanks to all, and John Galt in particular, for helping me.

Now, this does mean I have a second (and therefore spare) fully functioning Floppy Disk interface card, as I built a second thinking it was the issue (and not the forty year old floppy drive).

[John Galt]

you know i had one drive where the stepper motor was all sticky from age and it didn't work right at first.

this was when i was testing the old drives on my PC while constructing everything for the Altair duino

i ended up oiling and cleaning the heads and moving the stepper around by hand and it came back to life.

i don't know if the failed drivers are mechanical or electrical but can't hurt to keep messing with them.

also deoxit is a god send. you pull the jumpers and give everything a lube with the stuff push the jumpers back on and spray the edge connector just to get all the contacts happy again.

it would be cool to get a 4 disk system now that you have it working for the first time.

thanks to david i finally got FLOP2PC to work so now i can backup my Disk images from floppy to my PC.

had to change the ASM file a little and recompile.

[John Kennedy]

I thoroughly cleaned the four (!) 5.25 drives I know have, re-set jumpers, used the new, proven cable from Amazon. Three of the drives now mostly work, one is just unresponsive.

Next project is to get the full-sized IMSAI CPA front panel working, and then the (now known to work) DazzlerII card working (I think it's a RAM card issue).

That's the thing with S-100: there's always something new (old) to work on!