How about this? It needs one such circuit per turnout, but does enable an ON-ON SPDT toggle switch to be used both to trigger a pulse (about 0.6 sec with the components shown) and to provide the required direction indication.
How about this? It needs one such circuit per turnout, but does enable an ON-ON SPDT toggle switch to be used both to trigger a pulse (about 0.6 sec with the components shown) and to provide the required direction indication. View attachment 101582
Impressive! Where do you get that program from, what is that?
However, one circuit per turnout? Not economical I think, and centralisation is important to facilitate later extensions, expansions, automation projects,..
One circut would not be too bad once in PCB form. If you need 50 TAs to be conrolled multiplexng can be used or a large I2c bus where each driver can control 16 things. More difficult than it sounds. Nonetheless you can stil multiplex with a high and low side driver. You can multiplex the inputs too. You could store the state in FRAM. The micro does the pulsing too. With two 16 chnanel I2C output chips you can have 250 ish solenoids with 32 driver chips and if you want you can control with I2C and a couple of chips. 16 bits can be reduced to 4 with a priority encoder. So about 8 bits for about 250ish solenoids.
Impressive! Where do you get that program from, what is that?
However, one circuit per turnout? Not economical I think, and centralisation is important to facilitate later extensions, expansions, automation projects,..
Still Requires the Center Off Switches and it will Work but Costs more since you need one of these circuits for each set of Solenoids.
Also the Heat in the Mosfet will Increase as the Gate Voltage Decreases.
alec_t, your proposal requires one pcb with components per turnout: this is not only more expensive, it would also require 3 wires from the control panel to each turnout, same as with a centralised solution (chemelec).
Case 1: SPDT ON-ON switches cost €2.23/10, plus the DPDT pushbuttons @ €3.84/10 versus Case 2: DPDT (ON)-centerOFF-(ON) switches that cost €6.80/10. Case 1 also requires more wiring, so cost still higher. Chemelec, your last proposal is the best.
Except in case 2 I have to provide LED feedback on turnout position because a center-off switch does not show that.
In any case, thank you all for your thinking, Erik
Something I tried to post from my cell earlier in the week didn;t post for whatever reason.
The general idea is to matrix the switches, so a 16 x 16 matrix could control 200+ solenoids (exact number not 255 because of 0,0 is an off condition.
You could design the board to accommodate 16 high side drivers and 16 low side drivers. To save $, don't populate. If you need 16, use a 4 x 4 matrix. If you need more start populating drivers. e.g. 4 x 5 or 5 x 5.
The number of bits can be reduced from 32 to 8 if you use a 4 line to 16 line decoder (not sure if they are made), but I think they are.
One bad thing is that the high current lines are run.
An I2C BUS to I/O port may be able to control them.
Keep the state of the switch in FRAM which can be I2C based. No batteries are required and should have an unlimited lifetime.
Another problem is how to say use the state to activate a crossing light. Well, if you need it, hopefully you can use a limit switch.
In terms of activation, you can multiplex that too, and save on ports. Look up a typical keypad. Analog Devices has an IC that will decode a huge keypad to off load processor work.
Chemelec, I agree with you: your design is just the best and will be used on my layout; all here have contributed so much, and in the end a lot come out of here. This board has given me so much in terms of information, friendship, help, sociability and feedback.. I feel in debt to you all.
Your offer of a pcb I gladly accept, I appreciate it a lot but I would like to know how I can return the favour: pm me anytime, I gladly reply.
Have a good day,
Erik