CBM 3032 and datassette

[Papalapa]

I'm enjoying my new PET but there are some jobs to do yet. I'm waiting for the SD2PET Future but at the moment I'm trying to load some program from the Datassette. Datassette was not used for years and I have had to replace the belts by a new ones.

I recorded a WAV file from the computer to the C2N but the PET do not detected the program and still SEARCHING all the time. After trying with my Commodore 64 and VIC-20 I realized that it was not properly working. The guilty was the IC LA6324 and once replaced the issue was solved with the Commodore 64 and VIC-20. The problem is that I cannot load files recorded with my Sony cassette in the PET. I cleaned heads and adjusted azimuth unsuccesfully.

I used three different PCs and when I record a file on the Sony in WAV format, I play the tape and I can hear the sound so, the audio signal es being recorded on tape. I wrote a small program in the PET, saved to tape with the Datassette and in this case I can load the program on the PET without problems.

Then, all seems to indicate that by some reason the Sony cassette recorded is compatible with the Commodore 64 and VIC-20 but not with the PET ¿?

Due to this, I'd like to know if there's a way to load a PET program in the Commodore 64 or VIC-20 (I have the Pi1541 to do this) and then record the program in the Datassette to load in the PET later.

Thank you

 

[daver2]

The cassette interface for the PET is 100% incompatible with anything else!

The interfaces are completely different, but you could get lucky of course...

Dave

 

[dave_m]

Due to this, I'd like to know if there's a way to load a PET program in the Commodore 64 or VIC-20 (I have the Pi1541 to do this) and then record the program in the Datassette to load in the PET later.

There is a big Catch 22 here. You can load a PET .prg file into a C64 by using a hex editor in a PC to change the first two bytes of the file from 01 04 to 01 08. This will fix the load address to $0801 that the C64 uses. This will allow the file to load correctly in the C64. However when you SAVE this file, it will have the wrong address for the PET to run the program. However it will load the file into the PET at $0801, but the PET wants it at $0401. To fix this, you would need an Extended Machine Language Monitor program loaded into the PET first. Then, after the wanted PET program is LOADed, you can use the Transfer command to copy the program to start at $0401. Then SAVE it to the C2N.

The Transfer command has the syntax of
.T 0801 0C01 0401
where 0801 is the starting address of the data, 0C01 is the end of the data and 0401 is the starting Transfer address. This requires you to know how big the file is. The hex editor can tell you this.

Extended Monitors include mikromon and BATPRO (for 4032/8032) and others. Some require you to program a 2532 EPROM and install it in the spare ROM sockets. See http://www.zimmers.net/anonftp/pub/cbm/firmware/computers/pet/other/index.html

The SD2PET will make life much easier.

 

[Papalapa]

The cassette interface for the PET is 100% incompatible with anything else!

The interfaces are completely different, but you could get lucky of course...

Dave

I don't understand. Do you mean that the C2N (1530) is not compatible with the PET ? If so, why I can save and load programs from the PET ? What is the datassette model used by the PET then ? If the interface is compeletely different, how can I be lucky ? ¿?

 

[daver2]

It is all down to the voltage levels (and frequency response) that are used for playback and record.

Your 'ear' is an imperfect instrument for telling whether something is going to work or not on a different machine.

You mention a "Sony cassette". Am I correct in assuming this is a commercial cassette player/recorder?

Dave

 

[daver2]

Ignore what I originally stated. Old age and poverty!

I was on my phone, so working from (faulty) memory.

I am on my 'real' computer now and I have just checked all of the pinouts - and they are the same across the VIC-20, Commodore 64 and the PET.

The cassette read and cassette write signals on the PET (I checked the schematic for the 8032) are wired directly to PIA and VIA pins - and are hence TTL digital voltage levels.

Dave

 

[Papalapa]

Ignore what I originally stated. Old age and poverty!

I was on my phone, so working from (faulty) memory.

I am on my 'real' computer now and I have just checked all of the pinouts - and they are the same across the VIC-20, Commodore 64 and the PET.

The cassette read and cassette write signals on the PET (I checked the schematic for the 8032) are wired directly to PIA and VIA pins - and are hence TTL digital voltage levels.

Dave

Dont worry about the age, I'm 61 and I know what do you mean

Many thanks to spend your time with this, it's clear now for me.

 

[Hugo Holden]

The Datasette internal design is quite different to a commercial cassette player.

Basically they hit the record head with logic pulses directly out of a differential drive formed by two logic gates. Their circuit diagram suggests the inductance of the head partially differentiates the logic pulse to give a recorded positive going then negative going voltage that crosses zero. When the incoming logic level is low, the head is polarized in one direction, and when high in the opposite direction. So on the tape there would be opposing dipoles straddling the signal transitions and those representing the signal to be recovered. Also the Erase head simply has a DC bias applied, no AC bias is required as there is no question of record playback non-linearity.

On the playback side of things, the pulses on the tape are amplified up by some gain stages then, passed into an OP (IC2-2) amp wired as a zero crossing detector with some hysteresis R17, and then via some logic buffers out.

(I think there was a thread at one point which there was a remark about whether the polarity of those recorded or recovered pulses could be important as a compatibility issue too, though I think it was noticed somewhere that two variants of the Datasette it appeared to have opposite polarity recording at the head with respect to the playback, so likely it was not an issue, but I cannot recall if that question was fully resolved or not. At least in systems that ultimately detect zero crossings, there should be no overall effect, hopefully someone will chime in again on that issue)

So the whole affair is a 5V logic digital signal in, and 5V signal out system, unlike the audio player which is designed to record low level audio and return it to its output in a linear amplified version of the same.

Though in the commercial unit, if the volume control is turned right up, usually the audio output stage will clip, but it won't clip to a definite close to 5V logic level like the Datasette.

I would expect that recordings made on the Datasette are only compatible with a Datasette, and that commercial recordings of any audio made in standard commercial tape recorders, wave files etc would not play back to any use in a Datasette/PET input system because of the recorded level, but, I might be wrong about that. I figured that probably all the commercial tapes for PET's, must have been recorded on banks of actual Datasettes or at least used the same Head drive circuit, not commercial audio cassette recorders, but I could be wrong about that.

https://www.zimmers.net/anonftp/pub/cbm/schematics/datassette/c2n.gif

In vintage computer systems that use commercial cassette recorders, good example the SOL-20, the signal out of the player is heavily processed, including with AGC gain amplifiers, zero crossing detectors and also phase locked loops to help stabilize the data before it is ultimately fed to the UART. Though, there were much more simple interfaces say with the AIM-65.

 

[Hugo Holden]

....a couple more notes about that Datasette schematic, IC3 is drawn as a non inverting buffer, but clearly the IC type must be an inverting buffer. And although IC2-2 changes its output state from zero to 5V, depending on whether its AC coupled input voltage is above or below the signal's own zero baseline, it is not a true zero crossing circuit which produces a narrow pulse at zero crossing, it is more of a high gain signal clipper circuit with positive feedback to help square the signal up.