Well I have moved along with this board (and at the same time the Bally A3 solenoid driver board). I have determined that the best way forward for both these boards is to bring them into the 21st century and use FETs for outputs driven by optocouplers which makes the output transistors fully isolated from the driving circuitry and eliminating problems upstream should a FET short.
Probably the main deciding factor was when I had a long hard look at the driver board I have here:
Using darlingtons (which were the best thing available at the time this board was made) suddenly seemed a pretty stupid idea!
FETs have some excellent traits such as (specs are for the ones I am using) 'on resistance' of just 0.044 ohms. This means that, when switched correctly, they will radiate almost zero heat.
Here is the trick. Some other replacement board manufacturers have seemingly overlooked a very important part of the specs for FETs and the TTL that drives them. In order to switch on FULLY there is a very wide range of gate voltages specified by the manufacturer. Depending on the production batch of the FETs the "fully conducting" gate voltage can need to be as high as 3 volts (for logic level FETs).
Now the problem part. The driver chip is TTL (a 74LS154) and the specs for that part clearly show that an output high is "typically" 3V4, which is JUST enough to saturate a FET on the high end of its specs BUT the minimum "high" is just 2V7 and this is NOT enough in every situation!
You simply cannot design electronics this way - I ALWAYS allow as much margin for variances in parts as possible. So what to do about this dicky drive situation?? After a little head scratching the solution became clear. Use opto couplers. They can be driven easily by the 74LS145 and the output of them is fully electrically isolated from the input and so can be connected to the same supply as the output driver FET. In the case of Hankin boards, 24V.
This means we would now have a gate voltage somewhere up around 20V - in fact, we don't even need to use logic level FETs, almost any FET would work OK with this circuit. Here is my circuit scratching:
It seemed like a great idea on paper so I thought I would breadboard it to confirm all was good:
That was succesful so I moved on to the layout. I wanted to lay the transistors flat on the board - I just do not like dozens of transistors standing up waiting to be bent, bashed and broken off the board. I had already done something similar with the Dr. Who mini-playfield driver board. I pinned them to the board with a plastic clip to hold them in place:
So, scrap the design work to date on the Hankin board and start again. Added the required number of optocouplers and changed the orientation of the FETs as they seem to fit better vertically on the board:
A bit of wiring and fiddling to go but I think the end result will be a board that will last better than the original and it uses ALL parts that are still manufactured today so repairs, if they are ever required, should be OK well into the future.
More to come............
