OK, after studying the circuits and looking at the notes from Marty Goodman, I *think* I understand what the TPDD2 cable actually does. I'll have to build up a circuit to test, but I do believe I understand the point of the components on the circuit board on the cable are doing.
According to the notes, these 3 components had a marking of "14" on them. According to a datasheet/cross reference I found, these indicate that they are PNP "digital transistors". The internal schematic of these things is two resistors of 10kohm each, one resistor is connected from the base to the emitter, and the other connected from the base to one terminal of the case. This seems consistent with what Marty reports, for the connections from the signals FROM the TPDD to the cable, he indicates a resistance of 22kohm. Close enough. The TPDD doesn't have RS232 line drivers on it, so the signals that it generates are TTL level signals: 0V and 5V. RS-232 requires signals of +12V and -12V nominally for the different levels, but will accept any signal as good from -18V to -3V and +3V and +18V as good signals. the "dead zone" from -3 to +3V is indeterminate. So, for TTL, this is a problem for the 0V level, but not for the +5V level. If you look at the schematic for the Tandy 102, you will see that for each of the three inputs, RXD, CTS, and DSR that come from the TPDD, there are two resistors, 6.2K and 18K connected in series from each input on the connector to Vee (-5V), with the input to the buffer taken between the two resistors. The connection to the driver is a high-impedence input, so this will appear as a voltage divider when the TPDD 2 is driving the signal high (5V). This should appear as around +7.5V to the receiver input, well outside the "dead zone", so we are good there. Now we come to the 0V signal level from the TPDD 2. I believe that the base (+ resistor) of the transistors in the cable are connected to ground, the emitters are connected to the signals coming FROM the TPDD 2. The collector is connected to the input pin on the Tandy 100/102. At the 5V level, the transistor will be saturated and switched ON, so ~4.5V-5V will appear at the input to the 100/102. At the 0V level, the transistor will be switched OFF and go into a high impedence state, and this will appear to the input on the 100/102 as no connection. The internal circuitry of the 100/102 will therefore effectively have the input of the receiver connected to Vee through the 18k resistor, interpreting the signal as -5V. I'll have to build a circuit to test, but I believe that's the intent of circuitry in the cable. On Tue, Jan 12, 2016 at 4:12 PM, Geoffrey Oltmans <[email protected]> wrote: > Yep, that makes sense. The inputs to the TPDD should already take care of > this on the input side... they are clamped to the Voutput low/high of the > buffer. On the flipside, the outputs on the TPDD have diode protection to > prevent damage if something is mis-wired, but no way to generate the normal > +/- 12V that RS-232 requires. IIRC, +5V is not much trouble since it would > be outside the dead zone of around +/-3V, but 0V is a problem... so that > level needs to be translated. > > > > On Tue, Jan 12, 2016 at 2:04 PM, John R. Hogerhuis <[email protected]> > wrote: > >> On Tue, Jan 12, 2016 at 11:40 AM, Geoffrey Oltmans <[email protected]> >> wrote: >> > That's useful, and I must confess I was looking at the wrong part of the >> > schematic! ;) >> > >> > I suspected that it must not be a straight through cable, because that >> would >> > be too easy. From his description it sounds like he ran across three >> > transistors, SOT-23 packages I'm guessing. According to a cross >> reference >> > most likely a PNP transistor with two built-in resistors, similar to >> what is >> > shown in the schematic on those same lines. >> > >> > I'll have to try and construct a cable, but the next hurdle is that I >> have >> > no disks! >> > >> > >> >> I might be wrong but I don't think anyone has successfully created a >> TPDD cable other than one that used to be sold on EBay and that guy >> didn't publish his design. >> >> The theory is to sap power from flow control lines to convert from TTL >> levels supported by the TPDD to the minimum RS232 levels necessary for >> the Model T. >> >> -- John. >> > >
