OSI 600D Superboard Build

[rubber_jonnie]

OK, so C6 replaced with the correct value but no difference.

I also pulled U62 in case that was bad, It wasn't, and also tried with one leg of C6 lifted, but still no output on Pin 6, so replacement CA3130's are ordered.

Provided I then get output from that I can then start to chase down where the output gets to and if R57 works to adjust the input as expected.

[rubber_jonnie]

So I think my CA3130 was a fake, here are the new ones on the left, old one on the right:

IMG20240425104130.jpg

All 3 of the new ones work just fine, and I get a nice signal where I should at pin 6 of and over at pin 5 of U69:

IMG20240425115252.jpg

The blue trace is at the tape in currently coming from the aux out whilst I do a save, and the yellow trace is pin 5 of U69. The pulse width at U69 is within the range of 500uS to 630uS according to the service manual after adjusting R57.

I am able to save however I don't know if saving is working 100% as I could have cassette and tape player issues, so whilst I get what sounds good when I save, when I try to load it back in it's garbled. Nearly there though and I can test with another cassette deck.

[stephen_usher]

I wonder what chip the original was. Maybe a 555 timer.

[rubber_jonnie]

I wonder what chip the original was. Maybe a 555 timer.

Wouldn't surprise me.

[rubber_jonnie]

OK, so I have made some progress thanks to help over at osiweb.org.

I was given a small basic routine as follows:

10 SAVE
20 POKE 61440, 21
30 PRINT "U";
40 POKE 14,0
50 GOTO 30.

What this odd little routine does is similar to the test from the service manual, but it causes a 150Hz signal to be sent from the 6850.

Then by jumpering from pin 9 of the expansion connector to pin 10, this 'save' signal then gets routed straight into the cassette in port, and it can be followed all the way through back to the 6850.

What I ended up having to do was swap U62, U63 and U69 for various different makes, and I finally found a combination that gave me this:

IMG20240610223507.jpg

The yellow trace is the output from the 6850 and the blue trace is the input into the 6850. The shift is normal according to the folks over at osiweb.

I'm now getting partial loads and things are starting to almost work and I'm fairly confident that the save part is good as I am almost able to load a complete program from tape that I saved previously.

All of the ICs I removed and swapped about seem to be working just fine according to the TL866, but clearly there are more timing issues between different ICs that mean finding a 'set' that works well together is trial and error.

Progress at least though.

[Phil_SB]

On the matter of the serial interface - I built the cassette port interface from the schematic on a PCB to be able to use it for another project. I kept having it work for a while and then the input stopped. Eventually I found that the CA3130 was failing to operate consistently. This device is a FET input op-amp and in this circuit as published has no way for any (albeit minute) bias current (or capacitor interaction with the protection diodes) to be discharged. This will accumulate at the input and effectively change the input voltage until it cannot repond to the ac input. (Leaky capacitor or board surface may have avoided this on the original board)
Fitting a 1Megohm resistor between pins 2 and 3 of the CA3130 seemed to cure the problem.
(I have found a similar problem to this on several industrial installations - you always need a 'grid leak' - old valve reference!)
Hope that is of some use.
Phil

[rubber_jonnie]

On the matter of the serial interface - I built the cassette port interface from the schematic on a PCB to be able to use it for another project. I kept having it work for a while and then the input stopped. Eventually I found that the CA3130 was failing to operate consistently. This device is a FET input op-amp and in this circuit as published has no way for any (albeit minute) bias current (or capacitor interaction with the protection diodes) to be discharged. This will accumulate at the input and effectively change the input voltage until it cannot repond to the ac input. (Leaky capacitor or board surface may have avoided this on the original board)
Fitting a 1Megohm resistor between pins 2 and 3 of the CA3130 seemed to cure the problem.
(I have found a similar problem to this on several industrial installations - you always need a 'grid leak' - old valve reference!)
Hope that is of some use.
Phil

interesting, thank you. I need to get back to this machine as I keep diverting but it would be interesting to see if that makes a difference, though I know of other builders for who it just works.

[rubber_jonnie]

OK, so it's been a while since I posted about this machine, but it's a new year so it's time to get the 610 expansion board built.

IMG20250326104622.jpg

It's a lot of soldering for sure but it should give the OSI a good memory boost. Not only that but it'll provide the OSI with a floppy disk option.

The small board to the right of the 610 is the data separator board I'd previously mentioned, which allows the FDD connection. It's mostly built bar some missing ICs that are on order and some resistors/caps that I can install once I get the BoM from somewhere and work out what's what.

You can ignore the other board, that's the PPC expansion board.

I'm gonna need a big box of sockets for that 610 board, better get counting!

[HigashiJun]

Good to have some news from this project.

Good luck with all that soldering.

All the hassle should be worth it.

[rubber_jonnie]

Good to have some news from this project.

Good luck with all that soldering.

All the hassle should be worth it.

Thanks. Good job I recently stocked up on solder!! The main issue is no solder mask, so you have to be very careful not to bridge anything. I need to get some standoffs too, so I can stack the boards.

I had quite a few of the sockets, have now ordered what is missing. I expect in terms of the other parts I'll have those in stock, so logic, resistors, caps etc and probably many of them are the same as on the OSI 600D computer itself.