Troubleshooting reverse polarity SPDT circuit

Thread Starter

basline

Joined Aug 16, 2012
20
Hi All,

I'm floundering a bit with trying to troubleshoot a problem with the circuit before. I design it about a year, and thought I tested everything during the design but obviously didn't as the board isn't working correctly. The goal was to have a circuit where a single pin from my micro-controller (labeled Hot_cold below) can control two SPDT relays with an optocoupler for protection, and I would use that pin to control the polarity of DC power going to a peltier device (so a mostly resistive load). I would also have LEDs wired into the circuit to indicate which polarity the circuit was in (blue for cooling, red for heating).

My goal was when the pin goes high, the red LED would come on, and the relays would contact the NC position (so no power flowing to the relays) and the blue LED would be off. When the pin when low, the blue LED would come on, the coils on the relays energize moving them to the NO contacts and the red LED would go off.

The way the attached schematic actually works is the blue LED is always on, but the red LED behaves as expected relative to the micro-controller pin voltage. The relays always stay in the NC position and it looks like the collector-emitter voltage on the darlington transistor (Q2 in the schematic, its part number is KSP13BU) is about 6.5V o r 0V.

I'm stymied as to what is wrong with the circuit. I've forgotten alot about this project at this point. I'm not sure if this is about the design, or if I have chosen the wrong parts for either for resistors or for some of the transistors. I used eagle to layout my PCB, so I'm pretty sure the issue isn't with my PCB layout. I'm also not married to the micro-controller pin high/low heat/cool convention, but I would like to make this work with a single micro-controller pin.

I know there are a bunch of reverse polarity posts, but I was hoping I might be able to salvage this design if possible but getting help with this circuit rather than scrapping and starting over.

thanksHeat-Cool relays image.png
 

Thread Starter

basline

Joined Aug 16, 2012
20
Dennis,

It appears as though I needed to reference my PCB design more closely. I thought I had Q2 voltages, but they were Q1 voltages. Here are the voltages from collect to emitter on all the transistors:

..........pin is low ............. pin is high (heating LED on)
Q1 ...... 6.4 ........................... 0
Q2 ...... 7.7..........................6.8
Q3 ......10.7 ..........................0
 

Thread Starter

basline

Joined Aug 16, 2012
20
Measure the base of Q2 when pin is low. With 6.4V at the collector of Q1, Q2 should be saturated.
On Q2, it looks like base to emitter is 3.8 volts when the pin is low, about 0.2 volts when the pin is high.

What little intuition I was developing on this stuff is now gone. Does that even make sense?
 

dl324

Joined Mar 30, 2015
16,839
Looks like a problem with Q2. With 6.4V on the collector of Q1, Q2 Vbe should be about 1.4V.

For future reference, voltage measurements should be taken with respect to ground unless specified otherwise.
 

Thread Starter

basline

Joined Aug 16, 2012
20
If I replace Q2, or move to another darlington transistor will that address the the fact that the cooling LED is on constantly? Or is that driven by a design problem?

I think I have a couple spare KSP13BU somewhere. I will follow up once I've had a chance to replace the current Q2 with a new part.

Thanks!
 

dl324

Joined Mar 30, 2015
16,839
Don't see anything wrong with the circuit. What is the total current you expect in Q2? What is the beta?

As test, you could try driving the base of Q2 with a 1k resistor to 12V. That would increase the base current.
 
Top