Train Work - Followup

To followup on my last post on Train work. I probably needed at least one more track controller to wire up the remaining track controls.  But if I am building one, might as well build more as shown here.  That is four more controllers which will expand the solenoid control by 160.  The small one next to it is the LED controller.

The new track controllers have a little more power in the solenoid activation.  This schematic maybe a little hard to read, but here is the explanation.  The I/O Controller (MCP23S18) is a SPI controller whose outputs are either low or open collector.  The NFET on each solenoid control line is OFF when the MCP23S18 output is low or ON (grounded) when the MCP23S18 output is open-collector.  Previously an auxiliary PIC processor controlled these NFETs directly. Thus the I/O pin on the aux PIC would be either OFF (ground) or ON (3.3VDC).  This ON state did not drive the NFET into full saturation (turn on) and thus there was more (though small) resistance between the SOURCE and DRAIN pins than if the GATE voltage was higher. 

 

I used the MCP23S18 only because it was the only part available at the time.  But since this forced me into a pullup resistor on every NFET gate, I decided to take advantage of it.  The Gate resistor is connected to a BIAS line.  Using a PIC I/O pin to control another NFET, the bias line is either 5VDC or GROUND.  Thus on Track Controller power up, one of the first items is to set the BIAS_CTRL I/O pin to high, this causes the BIAS output to GROUND, ensuring all of the solenoid outputs are high and the solenoids are not activated.  The pullup on the GATE of Q54 in the schematic below, also ensures that the BIAS output is GROUND while the PIC processor initialization is in progress.  This prevents any solenoid "chatter" during the initialization process, due to the GATE inputs floating.

What else this does is place 5VDC on the GATE of the individual NFET solenoids when they are activated, instead of 3.3VDC in the previous design.  This drives the NFET further into saturation and thus lowers the SOURCE to DRAIN resistance and provides more current to the solenoid.  I have noticed that some track switches seem to be switch faster with the one new controller I have installed.  I need to finish the rest of these and see if that perception is universal.