Fixing a Remotely Controlled HVAC Unit Circuit Board (Essentially a Thermostat?)

[Curmudgeon10]

Here are the voltages.  In the picture provided in this thread of a main board wired to a control board, the control board looks identical to mine, and an enlargement of the main board shows what I see as a yellow-orange-brown (last one a little iffy) order, left to right on the control board connector.  Does that along with the voltages indicate a pretty good guess for the order?

[Ian.M]

Nope.  Those voltages indicate probable CMOS logic driving the LEDs, but they aren't internally consistent.  The first line shows the same voltage on each of the other wires with respect to Yellow, so one would expect to see zero volts between any two of the other wires for the rest of the table, unlike what you've got.  Therefore either you made several mistakes measuring them or they are changing as you do so.  Re-read my reply #8, power off, and check if *any* of the wires have continuity to chassis, then power on and re-measure them all for voltage giving both the unloaded voltages and  the loaded with 10K voltages, in the table and filling out both above and below the diagonal so we have two measurements loaded, and two unloaded for each pair so we can more easily spot mistakes or changing signals.

Then we *may* have enough to determine +Power and Ground and to start figuring out the rest, but I've got a nasty suspicion that if it continues to be ambiguous you'll have to get the main board out to trace tracks.

[Swake]

Reverse engineered the board.

Certainties:
- C2 is polarized, this likely indicates GND/Negative.
- J1 connector has pin numbers, these are respected in the schematic.

Uncertainties:
Polarity of LEDs assumed based on polarity of C2. Might be inverse in reality.
IR receiver is unknown. In the Kicad Library many if not all have the pin-out as in the schematic.

[sokoloff]

I saw similar physical topology, but assumed OUT was direct to pin 2 and Vss was (for reasons unknown to me) powering the cap through the resistor.

(Neither makes perfect sense, but reversing OUT and VSS made more sense to me.)

[Swake]

From the measurements in combination with the schematic black is likely GND and on pin5.
Please do what Ian.M suggests: power down the unit, measure resistance between black and chassis and report for next step.
If you are lucky the resistance is near zero and we have found the ground wire.

(as you are at it also measure resistance between chassis and the 4 other wires, it might help out too)

[james_s]

It seems to be that the easiest thing to do is just remove the cover to expose the other board the cable goes to and then note the order of the wires there. It's pretty certain that the order will be the same on both ends of the cable.

[Curmudgeon10]

I'm removing that cover today, one way or the other. 

[Ian.M]

Good.  You should then be able to confirm the theory that the wire order is the same at both ends.  From photos of other Chigo main boards, its almost certain you will find a 7805 regulator IC (TO-220 package, with a small, probably black heatsink).  Its tab is certain to be the 0V rail of the circuit, even if its not grounded to earth ground, so continuity testing from there to your wires will positively identify the 0V/GND wire.  With the regulator part number side up, tab towards you, its output is its leftmost pin, so you can confirm the +5V supply wire by continuity testing. If those two wires are the end pins of the connector, the order probably is the same at both ends, but as we don't want to blow the IR sensor, you should also perform the previously suggested unloaded and loaded (10K)  voltage tests, with respect to the known 0V/GND

[Curmudgeon10]

Closing in on a solution?

Just got the cover off, and was able to extract the Main board just enough to confirm that the wire order on that connector (from bottom of connector as it sits vertically on the board) is Yellow-Red-Orange-Black-Brown.

[Swake]

ok, good.
Can you remove the connector from the main board and plug it in the IR sensor board? If it fits because it is the same connector, I think you then have the order of the colors because I would expect the cable to be 'straight' else it would have been very problematic during assembly stage.

[Curmudgeon10]

Unfortunately, I can't do that.  The connector on the main board is soldered to the board (as was the one on the sensor board), and the wires were fixed to the connector via tiny pins.  In fact, on the sensor board, since I can locate no one in my town who could actually "repin" the wires to it, I had the connector removed and some individual wires soldered directly to the board.

[james_s]

I think it is extremely unlikely that the pinout is different on each end. I don't think I have ever seen a piece of equipment where a cable of that sort wasn't straight through.

[Swake]

On the connector on the sensor board there are pin numbers. I guess the one on the main board has some too.
connect 1 on 1, 2 on 2 and so on.

[Curmudgeon10]

I have a competing theory from another forum where some are proposing a red-yellow-orange-brown-black order. Here is their analysis:

"First, 5 volts is a standard voltage for electronic controls.

Well, just looking at the printed circuit traces, I see a couple of things.

    The thicker traces on red and black suggest these are Positive and Negative (often called Common/GND when you're dealing in electronics, but it's definitely NOT AC power safety ground).
    The conventional color choice for + is Red...... and the preference for - is Black ... and the polarity of the capacitor strongly suggests that was done here.
    The mystery orange/brown wires both do the same thing. They go to a "device of some kind", probably a switch or light, and the other side of the device is black (- or GND)? That would be very conventional.
    And here is the takeaway from that: if you exhanged brown and orange, most likely all that would happen is the lights or switches would be exchanged; e.g. it would light up ERROR and not ON instead of the other way 'round like it's supposed to.


The 3-prong thingie in the middle looks an awful lot like an infrared sensor and the schematic seems to match up.


I think the schematic is this. (https://electronics.stackexchange.com/questions/649949/what-is-this-component-and-did-i-make-this-schematic-correctly) "

I did measure resistance from each wire to chassis ground and came up with all open circuits.  The multi meter did record reistance between several points on the sensor board, so I know its operating correctly.

[james_s]

I would not expect chassis ground and DC ground to be the same thing. The other poster that mentioned the 3 terminal regulator had a good point, the tab is almost certainly tied to DC ground.

[Curmudgeon10]

As I reported in another post, I did measure resistance but apparently I only used main power ground, not DC ground, so the measurements just showed open circuits.  Not sure how to find the "DC Ground."

I have exposed the main board, and can pull it out a bit, but it is (I guess) tied in to the wire bundle that exists the rear of the board through the wall unit casing, and then on to the outside unit.

[sokoloff]

The DC ground is probably the black or brown wire. It is the negative side of the linear regulator. That’s probably tied to the negative side of an electrolytic capacitor in the power supply area.

[Ian.M]

If you show us photos of the main board in the unit, we can probably point out a DC ground (0V) to you so you can continuity test from there to the wire ends.

[Curmudgeon10]

This is the best I can do with a picture of the main board.  I can only draw it from the housing about an inch or so.

[Ian.M]

That's not a lot to work with as we cant get to the regulator.  However there are two accessible points that have a high probability of being grounds - the negative leg^* of an electrolytic capacitor next to the off-board signal connectors, and the metal frame of the tactile switch next to them.   If there is continuity between those two points the probability becomes a near certainty.

In all cases, test for continuity from the capacitor leg to the wire ends, and from the switch frame to the wire ends, and tell us what you find.

* For a through hole Electrolytic capacitor, the negative leg is the one next to the contrasting stripe on the printed wrapper round its can. Other types of polarised capacitor have different marking conventions.

[Curmudgeon10]

OK, thanks.  Will do.  Results in a little bit.

[james_s]

Why not just wire it straight through and try it out? It's very unlikely that it won't just work at that point and if something is wired wrong it is unlikely to do any catastrophic damage, the only vulnerable part on that PCB is the IR receiver and those are cheap and readily available.

[Curmudgeon10]

Why not just wire it straight through and try it out? It's very unlikely that it won't just work at that point and if something is wired wrong it is unlikely to do any catastrophic damage, the only vulnerable part on that PCB is the IR receiver and those are cheap and readily available.

I'm only being super cautious because I was advised earlier to be super cautious.

I measured the resistance between one leg of the capacitor (my probe could only reach one leg due to interference from other components), and the wires, as well as the metal frame of the switch and the wires.  All such measurements showed values except the brown wire, which went to zero.

That would seem to confirm that the brown wire is DC ground.  If the fattest trace on the sensor PCB board is ground, then brown would go to that pin (on the bottom), and the other wires, ascending, would be Black, Orange, Red Yellow per the order found on the main board connector, discussed earlier.

Its starting to sound like I could solder this thing up and hope that it works.  Any other tests or precautions recommended?

[Ian.M]

Check the yellow wire is +5V with respect to the confirmed Brown 0V/Gnd, and it doesn't droop when you connect a resistive load (any resistor within a factor of three of 1K should do) to verify that's the +Supply.  If it passes that test, then IMHO you can connect them in the 'straight through' order with confidence that it wont cause damage, and with good odds it will 'just work'.

[Curmudgeon10]

  The Minisplit Lives!!  Had to throw together a temporary work surface so I could solder the wires 4 ft above the floor, and I'm not going to be conducting any tutorials any time soon on soldering techniques for tiny wires, but everything works.  All that remains is to reinstall the various skins/covers that were removed to determine the wiring order.

Many thanks for the professionals who hung in here on this thread and helped me figure this out.