Dura Heat GFA40 Propane Space Heater, PC Board

 History

My previous post outlined disassembly of the heater and my basic troubleshooting steps.

The Board

What follows is a look at the PC board.

So here's the PC board mounted at the top of the base in a plastic holder. Just bend one or both tabs on the ends to remove the board.

Here's the backside, image flipped BigClive style so that the tracks line up with the top side.

This is the schematic I created. But there is something wrong here. Design-wise it looks wrong. It simulates wrong. But I can't find the error. Resistance from the top of ZD1 to D5 is 10.2Ω, i.e., R3. Yet when tested live they are essentially identical at 11.9Vdc and maybe 0.25Vac. The cathode of D5 is connected to relay coil+. I should have measured at the top of D1,D3 - maybe next time

Here's a modified schematic that performs as expected. The only issues I see is that flyback current can be handled by Zener diode ZD1 - D5 is redundant.

So here's what you've got: AC line voltage comes in on the blue connector CN1. L1 goes through the fuse to the full wave bridge (D1-D4) and the motor. The motor returns to the neutral lead.

The neutral flows through 10Ω power resistor R1 and the polystyrene capacitive dropper C1. There appears to be an integrated bleed-off resistor R2 soldered between the capacitor leads. The resistance value us estimated, the capacitance value is a guess from what I could sort-of read on the cap.

As mentioned, D1-D4 form a full-wave bridge, with reference ground at the bottom of D2,4, and DC+ at the top of D1.3. E1 is an electrolytic smoothing capacitor (41µF est). Expect this cap (E1) to dry out and lose capacitance or rupture over time. That will produce more AC ripple and the relay coil will suffer.

Now we take the DC voltage and regulate it to 12Vdc via ZD1. ZD1 was tested separately and passed current at 12V when reverse biased - I don't know the diode model. So D1,3,E1+ source voltage through R3, dropping the voltage to 12V at the cathode of  ZD1. The anode is at reference ground.

That 12Vdc is connected to relay (REL1) coil+, coil- is at reference ground. D5 is the shunt or flyback diode that will pass the current sustained by the relay coil's magnetic field when the heater is unplugged and the  DC voltage falls. Without the diode, the voltage would rise (back EMF) to a potentially destructive level as the the inductive coil does its best to sustain the current flow. Check out  how ignition coils and spark plugs work.

So when the heater is plugged in, the blower is presumed to be running. If power fails or the fuse blows, the thermistor is still hot and generating voltage. The in-series over-temp safety switch hasn't opened yet. So the only thing to prevent the propane gas valve from staying open are the relay contacts which open when the DC voltage collapses. That's the singular purpose of this circuit. If the flame goes out, the thermistor voltage will drop and turn off the gas. If the fan stops or something impedes the flow of air, the over-temp switch will open.

Conclusion

I hope this helps you understand and troubleshoot your propane heater.

Warning: The content above is submitted in the context of a DIY repair. There are likely regulatory factors that affect the design and manufacture of these devices. I don't design or sell these contraptions, so I don't know. I just use them and try to fix them when they break.

If anyone has any insight as to where my original schematic went awry, I'd be interested to hear it.

Dura Heat GFA40 Propane Space Heater

Let's say you have something like one of these, a Dura Heat GFA40 propane space heater:

and you want to repair it.

First you remove the 4 screws that hold the bottom plate of the base in and see this.

and this. You have access to a nice PC board with connectors. That's about where the fun ends because to get to the burner plate to access the thermocouple and over-temperature safety switch, you have to:

1. remove 3 screws to remove the rear screen (easy)
2. remove the motor
• remove the (4) screws that hold the motor base, accessible from within the base. The ones in the 10, 4, and 8 o'clock positions, as viewed from the base, are easy. The one in the 2 o'clock position is behind the switch mechanism. I ended up loosening everything so I could get to that screw straight-on so I could loosen it correctly. That  meant removing the 4 screws that hold on the base,  the 6 screws that fasten to the standoffs between the inner and out tubes, the machine screw and nut that hold the capacitor, the 4 screws (accessible through the label under the knob) that hold the valve to the base, and the valve knob. That make everything loosey-goosey enough that I could loosen the last screw for the motor mount and extract the motor and fan blade.

From there I had good access to the burner plate and could remove the thermocouple, over-temp safety switch, and igniter post as needed.

Now I know I could have just tested between the blue wire (next to the yellow one, not the one to the capacitor) and the bare silver earth wire. First for continuity to confirm the safety switch was closed, then for voltage as I heated the tip of the thermocouple.

But I was sure the themocouple was bad because the heater ignited fine and burned until I released the knob and hence the TC bypass. This was after I repositioned the heater that I had trouble. The first run was perfect - it ignited and ran on the first try. Now I'll get the propane tank and test again with an eye on the relay and the yellow lead to the control valve.

Summary

My takeaway from this exercise is that this unit is not particularly serviceable, due to the singular lack of accessibility to that one motor mount screw. The component layout could have been a little better, but it's generally good, and not a mess of dodgy soldering and wire splices. Virtually all of the screws were the short black sheet metal screws. Exceptions were the machine screw and nut that secures the capacitor, and the small silver screws for the safety switch.

Recap

I spent more time troubleshooting the heater and diagramming the PC board (next post).

When I back-probe CN4 the blue lead shows the thermocouple (TC) generates at around 20mV or more after several seconds. The yellow lead which is connected to the blue TC lead via the relay contacts may only show only 4-5mA. I did not see a matching differential voltage across the leads (?!). The resistance across the pins (with cable disconnected) is consistently 200mΩ. The relay's spec for contact resistance is ≤ 100mΩ. I removed and inspected the spring contacts from the connector housing - both looked good.

I resoldered a number of pins, sometimes removing the old solder, not that any solder joints were looking problematic. The heater is still very sensitive to tapping on the base, or removal of the meter probe from the connector. The relay coil is getting 11.9Vdc. So I suspect the relay could be a little better, but the unit is usable as is, with an occasional tap here and there.