Thursday, November 19, 2026

Developed a complementary Schmitt Trigger device for signals from the 1130 to the Virtual 2315 Cartridge Facility

THE ORIGINAL SLEDGEHAMMER PRODUCED PERFECT SIGNALS - I WAS INSPIRED

The Signal Conditioner board I built uses a Schmitt Trigger design with very wide hysteresis so that bouncing signals, undershoots and other consequences of the difficult to control signal path are hammered into submission, giving me clean signals that feed into the IBM 1130 logic circuits. I am sitting around in my home watching the incessant rain, seeing most of the local roads flooded around me so that traveling outside is foolhardy. I had plenty of time on my hands.

I thought about the tweaking and adjusting I have been doing for signals going the other direction - produced by the 1130 logic circuits and received in the FPGA of the Virtual 2315 Cartridge Facility (V2315CF). One devilish complication is that the V2315CF operates in two modes. One mode makes use of the internal 2310 disk drive of the 1130 to move the disk arm and to produce the sector and index marker pulses that delineate the four sectors/slices as the disk rotates. The other mode completely emulates the 2310, generating the timing pulses and pretending to move the disk arm with fidelity to the timing of the real disk drive. 

This means that sometimes a signal is produced by Solid Logic Technology (SLT) circuits in the 2310 and routed to the 1130 SLT circuits, with the V2315CF simply watching the signals so that it can stay in sync with where the disk arm has moved. Other times, the signal is generated by the FPGA and sent to the 1130. 

Since SLT logic is a version of Diode Transistor Logic with lots of peculiarities, interfacing it with modern technology like the 74LVC244 receiver chips and 75451 open collector drivers is not straightforward. Then, the impedance of the signal paths is a huge mess, consisting of many different segments of travel with their own impedance and signal transfer characteristics. There are twisted wire pairs, ribbon cables, wire warp wiring, traces on various boards, pins and connectors. 

A NEW SLEDGEHAMMER COULD RESOLVE SIGNALS IN THE REVERSE DIRECTION

I could use a similar design to deal with the signals coming into the V2315CF. The hysteresis would allow me to recognize the state of signals and produce a modern clean signal for the FPGA to sample. In particular, when the V2315CF is in its real mode, making use of the 2310 disk drive to create many signals, I simply bridge the 2310 signal to the outgoing wires to the 1130, but also watch them in the V2315CF for the purposes of staying synchronized. 

Thus, the SLT logic of the 1130 produces signals that the SLT logic of the 2310 are designed to receive properly, and the new board could clean those up to work well with the FPGA. All the signal path characteristics between the 2310 and the 1130 are already handled by the IBM designers, so that will work as it did before the V2315CF was inserted into the path. 

In the virtual mode, where the 2310 is not producing the signals, the 75451 chip and terminator resistors in the V2315CF are responsible for detecting the signals coming from the 1130, without the 2310 circuits connected to them. This changes the impedance and other factors, which made it a challenge to find termination values that worked for both modes. 

Since the V2315CF needs the incoming signals to work properly in either mode, the sledgehammer design can handle either wiring path and produce exactly the same signals going into the FPGA. There are a few minor changes from the sledgehammer (Signal Conditioner) board I used for signals from FPGA to 1130.

CHANGES FOR THE REVERSE DIRECTION SIGNAL CONDITIONING

The original Signal Conditioning board creates a 0 or 3V signal compatible with SLT input gates. The new board, which I will call a Signal Driver board, has to produce signals that are compatible with a 74LVS244 receiver chip. I also modified the thresholds so that it switches on when the signal reaches 2.2V on the way up and won't switch back off until the signal drops below 0.65V on the way down towards zero. 

Changing the thresholds requires a different reference voltage, which meant a different chip that produces 1.6V instead of the 2.5V used in the original board. Then the two resistors on the op amp are changed to achieve the thresholds. The op amp swings dramatically between the 0 and 5V rails at the thresholds. 

The op amp output is buffered by a MOSFET so that the op amp is not loaded down at all. The MOSFET then drives an open collector transistor circuit. The terminator resistors establish the pullup for that open collector output. 

Finally, there are seven signals that come from the 1130 that I need to handle with the board, but the original Signal Conditioner is a five channel board. I expanded to seven channels.

WHIPPED UP A PCB DESIGN AND SENT IT OFF TO THE FAB

I completed the Signal Driver board and uploaded it to JLCPCB.com where it is being produced. I also created a Digikey order for the extra components to go on the board. I did order extras of some of the parts for the Signal Conditioner board, since the price was cheaper that way. However, others such as the unique resistor values for the Signal Driver board had to be ordered. I also needed the new voltage reference chip and the seven pin headers to accommodate the additional channels. 



No comments:

Post a Comment