• Please review our updated Terms and Rules here

Sol-20 Tape interface issues

H

Hugo Holden

Veteran Member
Joined
Dec 23, 2015
Messages
4,747
Location
Australia
Recently I have been working on the cassette tape interface on my Sol-20. I put together some notes on it, in the attached article/link.

I have just added a few pages to the end of the article. I had a problem that Processor Technology knew about; the welding together of the miniature relay contacts that control the remote connections to the tape recorders. The problem is due to the charging current of the filter capacitor in parallel with the cassette recorder's motor. PT's solution was to fit a resistor in series with the relay's contacts. However, in some small cassette recorders, this is not ideal. One solution I came up with for my own recorders involves adding a transistor inside the recorder. This may be of some interest to those using the Sol-20 with cassette tape storage, see the last few pages of the article:

http://worldphaco.com/uploads/The_SOL-20_tape.pdf
 
Given that the cassette motor is a heavy inductive load, I'm surprised that no one's suggested a "snubbing" diode across the load:

Theory here

It makes a huge difference, as it "snubs" the reverse EMF spike generated by the coil inductance. Don't use a small-signal diode, but a rectifier capable of handling the large current (e.g. 1N4007 should work here).
 
Chuck(G) said:
Given that the cassette motor is a heavy inductive load, I'm surprised that no one's suggested a "snubbing" diode across the load:
Click to expand...

Chuck,

In the original wiring configuration the snubber diode was not required to snub the voltage spike produces from the magnetic field energy in the motor being released at switch-off because the motor had a 220uF parallel capacitor connected across it at all times, which damps it right out, well past critical damping. The disadvantage though, being that the switch contacts which switch the motor have to be able to support the capacitor's initial high charging current, when the motor is switched on. Mainly though, for these recorders, that was intended to be a switch on a microphone which is much more robust than the contacts in the tiny DIL relays.

In the case of my recorders, it was easier to alter the wiring by inserting the transistor in series with motor's +ve terminal, therefore isolating the 220uF capacitor when the transistor switches off and the motor goes off, and therefore requiring the addition of the 1N4004 diode as a snubber to protect the transistor's collector circuit. Though the BD140 can withstand -80V on its collector, however on the second half cycle of oscillation it could possibly reverse bias the collector-emitter junction, so to be on the safe side switching an inductive load with a transistor, it is always better to always have the diode present.

Hugo.
 
Last edited:
Well, that's a strange design choice--the energy in the motor's windings has to be dissipated somewhere--and using an electrolytic for this seems to be an odd choice, as it would seem to extend the time current is flowing (LC time constant). In this respect, DC motors are worse than relays, as they keep moving after the current is turned off.

Were this my system (and it isn't), I'd install a snubbing diode and lose the capacitor. At least that way, the flyback current has somewhere to go--and you're not trying to charge a capacitor on the "make" part of the cycle. I'd probably also opt for a MOSFET instead of a BJT--they're a lot cheaper now than they used to be 40 years ago.
 
Chuck(G) said:
Well, that's a strange design choice--the energy in the motor's windings has to be dissipated somewhere--and using an electrolytic for this seems to be an odd choice, as it would seem to extend the time current is flowing (LC time constant).
I'd probably also opt for a MOSFET instead of a BJT--they're a lot cheaper now than they used to be 40 years ago.
Click to expand...

Chuck,

In this case it was a good design choice that the manufacturers made to have the capacitor directly across the motor, because the motor is an internal mechanical governor design, and this creates quite a bit of interference with the contacts in there opening & closing and its a tape recorder application so they wanted to keep the external noise down.The capacitor is just a pest though from the point of view of the initial currents when the motor is switched on by small contacts.

I used the BD140 transistor as I had one in my junk box and in this application the saturation voltage drop when the motor is stalled is only around 100mV and less than that in run mode, so it doesn't really need p channel low Rds ON mosfet, but if one was used instead, since it is a 6V system, a logic level gate threshold type (around 2.5V) would be the type to use. The BD140 also has a case with a insulated screw hole only requiring a sil pad insulator under it and easily screwed to the metalwork under an existing screw. The added parts always have to go, or be mounted, somewhere and this made it dead easy, so I chose it for mechanical reasons as well.
 
Chuck,
I see what you were getting at with the last question. I guess the governor contacts don't fail on closing at least, because of the inductance and resistor of the motor windings limiting the current. On opening there could be sparking, but for all we know there is a diode concealed in the motor or maybe just some resistive or capacitive damping. I haven't pulled one of these little motors apart yet.

I don't think the motors in these small cassette players are too wonderful, they have speed irregularities, but it is surprising how well the computer's tone decoder allows for this.
 
I recall using an inexpensive GE portable cassette player with my Altair 8800 (before I got a diskette drive). I used the 300 bps Novation modem with some frequency shift capacitance rigged so that originate=answer frequencies. I controlled the recorder motor with a small relay. Offhand, I don't recall any problems with relay contacts welding, but that was a long time ago.

That cassette setup was replaced with a dual-cassette paper tape emulator (Techtran) which had high-speed search features and could stream at up to 2400 bps, which was pretty fast back in those days. Tape control there was via ASCII characters. Pretty neat for its time--one drive was read-only; the other was read-write.

I still had to toggle in my bootstrap, but it was a lot easier than toggling in whole programs!
 
It is quite surprising how the Sol's cassette tape interface system manages to get away with such cheap small cassette players. Looking at the playback of a 1kHz test tone from these small recorders there is quite a bit of phase & frequency jitter at a low rate of a few Hz, easy to hear and easy to see on the scope. It may well be the accelerations or decelerations of the motor from its mechanical governor, or it could always be old belts. (I have some new ones on order)These errors of course double up on a record & playback. Despite this, the tone decoder system in the Sol appears very good at ignoring these frequency errors.
 
You must log in or register to reply here.
Back
Top