Skip to main content

Posts

Lessons Learned - False Voltage Readings from Multimeters with Low Batteries

Recent posts

MOSFET Driver Board PCB

Red PCBs do look sharp.  The MOSFET Driver Board PCBs came in and overall they turned out well.  Compared to previous boards I have ordered from Seeedstudio, the tin on the pads is a bit uneven and the inkjet for the silkscreen must have got gummed up.  You can see that the silkscreen is a bit messed on the white rectangle I left as a space to write on.  Small gripes for hobby PCBs and overall I'm quite happy how they turned out. MOSFET Driver Board PCB

Lessons Learned - High Side Optocouplers

Optocouplers can be used on the high side without issue. Optocouplers let us isolate two separate power systems.  In the case of my pinball project, I am using a computer PSU to power the logic, lights, and lower voltage items like motors etc; and I am using a somewhat suspect 48V power supply for the solenoids.  To keep these systems separated, I used optocouplers on my mosfet driver board. When designing the board, I was concerned that my transistor like optocouplers would not perform on the high side of the mosfet gate.  You cannot easily use a NPN transitor on the high side (Emitter Load).  You would normally use a PNP transistor.  The test circuit in the data sheet also showed the optocoupler in a low-side (Collector Load) configuration.  Why did I not use a high-side configuration?  Well, say if something went wrong with my optocoupler, I want my mosfets to default to the off position, ie. ground the gate.  With a high-side configuration they would default on due to the pu

MOSFET Driver Board

Final version?  I hope so.  I have gone back and redesigned my solenoid driver board several times.  In previous versions I attempted to offload some of the logic from the microcontroller.  For example, I used a Schmitt-Trigger [74HC14] with an RC network to provide a single pulse to the solenoid.  Also, I added in a PWM signal to set the average voltage 'seen' by the solenoid.  This worked but was very limiting to have the pulse length hardwired in and it was difficult to adjust without more expensive precision potentiometers. I chose to simplify my circuit to make it a general MOSFET driver for flexibility.  Since I have more than enough space inside my pinball cabinet, it was not a problem to have the extra boards.  The boards were designed to have 4-channels since by default I would get a minimum of 10 boards for the same price from Seedstudio.  Also, the cheapest board option was restricted to 10cm x 10cm otherwise the price balloons quickly. Driving 48V from a separa

ATX Breakout Board

Last November, I fried my 'lab power supply' that I made from a computer PSU.  I was testing my solenoids and the stress was just too much for the PSU.  The mistake I made was rapidly turning on and off my solenoid which likely resulted in voltage transients (spikes).  Since the PSU does have a current limiter, it was likely the constant voltage spikes that caused the damage.  There are various methods to suppress voltage transients such as by-pass capacitors and zener diodes.  In particular solenoids and other inductive loads experience these voltage spikes, hence my no longer functioning bench power supply. ATX Breakout Board I could have just rebuilt the same power supply by adding binding posts etc to the PSU housing but after searching the internet an ATX power supply breakout board seemed a popular option.  The main advantage compared to my old power supply was the addition of fuses on the outputs.  The extra safety this provides is perfect for bench top testing

Solenoid Driver Board with less Vias

Managed to get it down to three vias and the routing is a bit better.

Solenoid Driver Board Layout Preview

Here is an image of my solenoid driver board.  It is 99% autorouted but I did make the traces for the positive inputs for the solenoids much larger for the 10 amps peak they might see.  The current is very brief so the traces are probably larger than they need to be.  The traces for the solenoid inputs are 100 mils, the rest at 16 mils and I have the entire top and bottoms of the board as ground.  There are five vias but I do not think I can avoid them.  I also probably need to line the ICs in a way that is more pleasing to the eye and I have to add the silk screen layer to label everything.