Completely ridiculous project, but I'm doing it anyway and have a question or two

Yes, it’s goofy, and no one will ever want to duplicate it.

I have a 1953 Olds that was subframes with the engine, transmission, and front chassis out of a 1965 Delta 88. Among a whole host of other problems, this caused an issue with the gear indicator needle.

The car originally had an automatic transmission, but the shift pattern was different from the 1965 transmission. The indicator needle operates by energizing small electromagnets which then attract the needle so that it points at the correct character on the indicator plate, N-D-2-1-R. The current for the electromagnets is supplied from the neutral safety switch, which simply rotates a lever through three contact points. When the first contact is closed, the first electromagnet is energized and the needle points to N. When both the first and second contacts are closed, both the first and second electromagnets are energized and the needle points halfway between them, at the D. When the lever rotates one more position, only the second contact is closed, and so on. A linear, analog means of supplying current to the electromagnets.

The newer transmission uses a different shift pattern, i.e. PRNDSL, and so the electromagnets need to be energized in a different pattern - which isn’t possible given the nature of the rotary contacts.

So I came up with a logic gate approach, to allow the original method of supplying current to remain untouched, but would redistribute the current to the proper electromagnets to cause the needle to move to the proper location. That all works on a breadboard, using 12V CMOS logic chips, and the output is used to activate automotive relays. Three inputs, three outputs, three relays. If two inputs exist, two relays close, activating the proper electromagnets. This portion of the circuit works.

Where I’m running into trouble is that I was using pushbuttons to activate the logic gates, and once I had all the other bugs worked out, I discovered that the pushbutton approach does not correspond to the actual input situation. If I press the first pushbutton, the first relay activates just fine. If I press both the first and second pushbuttons simultaneously, then the first and second relays activate. All according to plan.

But if I press and hold the first pushbutton, and then press and hold the second pushbutton while maintaining the first one, the circuit does not respond by just also activating the second relay, it behaves as if there are two separate, unconnected events. I need to introduce an interrupt somehow, so that the logic circuit sees the situation that is created by adding a second input as a brand-new situation.

I hope that all makes sense.

My thought, as a sheer dilettante with no formal background in electronics, was to add an analog to digital converter before the logic circuit, specifically this one:

https://store.ncd.io/product/mcp3428-4- … interface/

I chose that one because it has four channels, works with 0-20 volts, and has a continuous conversion mode that can be configured to refresh between 15 and 240 samples per second, updating the output buffer with the most recent conversion.

My main question is, would this work? As an old auto mechanic, I don’t like the idea of just throwing parts at a problem until something holds together, but I really don’t know enough theory to be able to judge whether this is a reasonable approach or just laughable. I’m also trying to avoid delicate components as much as possible, because it will all be going into a car, with the attendant temperature extremes, vibration, jolts and bumps, etc. That’s also why I chose 12V components, to avoid having to shifting the voltage down and up.

Any ideas? I can provide whatever info would be useful. Thanks.

I’m interested in helping with this. In fact I signed up for this forum just for your question. I love working with classic cars and electronics, so hopefully I can help. From what information you have shared, I don’t see how an analog to digital converter would help. Can you share a sketch of the logic circuit you came up with? I can think of several ways to do this job, depending on what components you are comfortable working with. Could be done with relays and diode logic, could be done with CMOS or TTL logic gates, or if it were me, I’d use a microcontroller. It would need some protection from automotive voltage spikes on the power and inputs, it would have 3 inputs and 3 outputs, output drivers of your choice (MOSFETS or relays?), and only a few lines of code. I did some preliminary logic tables for this, as well as looked up the wiring diagram for the '65 (it seems to have a conventional reverse light and neutral safety switch), and a big question is: Are you reusing the position switch from the old transmission on the new one? If so, how will you handle 6 possible selections on the new transmission VS. a 5 position switch and indicator? If the switch can handle the mechanical travel, it’s a question of do we ignore park and map the other positions accordingly, or do we ignore low and map both P and N to N?