Solution To Flickering Fan Speed 2 Or 3 A/C PART 2

Sounds like it's the connector of the harness if you solder the connections directly to the switch it will fix it I had the same issue as u are describing and it's been months so far so good haven't seen a bit of flickering
 
Just to have the numbers down somewhere, here are the voltages on fan line 1 measured to a chassis ground in my stock P5 at the blower:

Code:
Blower    Wire             Voltage at      AC lit?
Position  color            Blue/Yellow
1           Blue/Yellow     .45              Y
2           Blue           1.04              Y
3           Blue/Red       1.15              N
4           Blue/white      .89              Y

and these are the numbers after this A/C fix

http://www.mazdas247.com/forum/showthread.php?123797885

was installed measured the same way. (I finally got around to doing the second set of measurements today).

Code:
Blower    Wire             Voltage at      AC lit?
Position  color            Blue/Yellow
1           Blue/Yellow     .58              Y
2           Blue           0.97              Y
3           Blue/Red       1.14              Y
4           Blue/white      .84              Y

They didn't change very much and I would guess that the differences that are there could just as well be from temperature differences on the days these were measured as to the new circuit. What this means is that neither the stock sense line circuit nor the fix circuit put very much current into the sense wire going to fan line 1. The voltage on fan line 1 still controls whether the AC is on or off, but it is no longer at all touchy since the intervening circuit presents a very low voltage to the sense line on the the control board in the dash.

While I was at it I measured the voltage drop at the back of the switch from the fan wire 4 to the ground wire (that is, between to the two thickest wires) at the back of the switch. This was 0.14V. I was a little surprised it was that high, but it doesn't indicate a problem, as V=IR, the current is 10A, so R is only 0.014 Ohms, which isn't very much for the metal to metal junctions there. (4 junctions? Wire -> spade-pad -> wiper-wiper -> pad-spade -> Wire). The power dissipation formula is P=IV => 10*0.14=1.4W. So not very much power being dissipated at the switch, and that fits with there being no heat discoloration in the plastic on my fan switch.

Note also that .84V -.14V = 0.7V, exactly the value expected if there is a diode somewhere between the ground wire and the negative battery terminal. (Not that I have ever actually seen such a diode, it just measures like there is one in the circuit.)
 
Pasadena, I like your fix, but I still think we all need to address the source of the problem, which is the increased resistance through the switch and its plug. The A/C light will be steady with a relay, but you could still end up melting your harness, like we've seen a few pictures of already.
 
Pasadena, I like your fix, but I still think we all need to address the source of the problem, which is the increased resistance through the switch and its plug.

I think the plug melting is actually a separate issue from the fan speed 2/3 problem, although it is probably true, as you say, that if the harness is melting it may make the 2/3 problem worse.

My car flickered but there is no discoloration at the switch, and the voltage drop across the switch while fan speed 4 is running is reasonable.

The thing is, if there is discoloration it can mean one of two things - the wiring is getting loose (and probably arcing from time to time), or the blower is drawing too much current. The former can be addressed by crimping the connections a little tighter, but the latter really needs a new blower. Ideally to check the blower current one would put an ammeter into the circuit, but I don't have an ammeter that can handle those sorts of currents

Google for: The Definitive Blower Motor blown/melted resistor/switch/wire fix

This guy had melted wiring and rewired his Jeep to use relays.
 
So here's what I did tonight. I bought a 6 pin connector male and female, and the corresponding pins and connectors from digikey. At such low quantities they are a huge rip off but what are you going to do? $15 in parts to make a 6 pin harness.

Here are the parts list
2-480705-0-ND qty 1
A1456-ND qty 1
A25336-ND qty 5 (must buy 10, so I'll send anyone who wants to do this the other 5)
A25372-ND qty 5 (same as the last)

Then I followed the original fix from this thread somewhat, in cutting off the jacket of the fan switch harness, and soldering wires onto the 5 male spades on the fan switch, except in this case I used new wires. Also instead of covering it in hot glue creating a massive thermal blanket, I just covered each one with a shrink tube. I put the gold male pins from digikey on the short wires I soldered to the switch, soldered them for good measure, and popped them into the new harness. I used 10 awg wire for the ground and #4 pin, and I used 16 awg for the rest.

Then I used a very small screwdriver to remove the female crimps from the harness inside the car, so I don't waste a good harness. Then I replaced the crimps with the gold crimps from digikey, soldered them and popped them into the other side of the new harness.

So now I have a solution that matches this thread but I have a harness that I can disconnect with ease when I want to remove the radio bezel. I am hoping my gold pins provide a better contact than the copper spades of the original harness.

So far so good. It is steady at all speeds. The A/C light does turn off in between 1 and 2, but this makes sense since the switch is only using 2 of the three internal prongs in between speeds. My car has always done this so nothing new, but I was hoping for solid all the time. It could also be that I somewhat fubar'd my switch with the heat gun when I last took it apart.

I still think running a relay off the sense switch is a good idea, although I think I would have used something like a 3v threshold FET instead of the relay so I don't add unnecessary current back there.

Mazda's design here is absolutely horrible, and running the entire blower motor current through a 4 position toggle switch is incredibly lazy. We've had solid state electronics for decades, there is no reason they shouldn't have used relays for the blower motor instead the resistor pack they used, but of course there is nothing we can do about it now. I am kind of getting the picture about horrible electrics in cars. If this is what Mazda does, imagine what is in cars with bad reputations for electrical gremlins.
 
So now I have a solution that matches this thread but I have a harness that I can disconnect with ease when I want to remove the radio bezel. I am hoping my gold pins provide a better contact than the copper spades of the original harness.

Being able to detach without reaching for the soldering iron is definitely an improvement on the original. As for better contact, well, just measure the voltage from the fan speed 4 wire to the ground wire while the blower is running at that speed. If it's significantly less than 0.14V (what mine measured, and it's stock) then your hopes have been fulfilled. I wouldn't expect gold to give a better contact when new than copper, but since gold doesn't corrode, it tends to keep working near its optimum longer.
 
I didn't pull it out and stick the multimeter on it yet today, since it's 130 in the garage during the day, but I drove around a lot and it was steady on 2 and 3. It turns off in between numbers on the way up but not on the way down. Yeah, my worry is that my new connectors aren't any better and I just pushed the problem downstream, but we'll see.
 
Last edited:
It turns off in between numbers on the way up but not on the way down.

That's probably nothing to worry about. It may tell us something about the mechanics of how the wiper moves across the pads inside the switch, but it isn't relevant to overall function.
 
Since overheating at the plug at the back of the fan switch is either caused by a bad blower (too much current) or a connection/wiring failure in/near the switch (too much resistance) it would be good to have an easy way to distinguish between the two. Measuring the voltage across the switch will tell you if the latter condition is present, but won't say anything about the former. I had thought that to measure the blower current one would need to put an ammeter in line with it, but that's not strictly true. All one needs to do is measure the voltage across the blower resistor in fan position 1, between the fan speed 4 and 1 wires, and then divide that by the resistance measured between the same two points (which can be done safely when the fan is off, since in that state the blower resistor isn't in a closed circuit.) Possibly the current problem depends on fan speed, in which case measure the voltage/resistance between lines 4 and 3. The one thing you can't do without a high current ammeter is to measure the current at fan speed 4, since there is no fixed resistive drop in the system to use as a reference. (I don't think the resistance across the switch will be stable enough to measure this as it will vary with whatever grit/oil/grease happens to be between the wiper and the pads on any given test.)

Also it is probable that there are two different blower failure scenarios. One where the blower takes a huge amount of current to start, and another where it draws more all the time. Both should be measurable with a fast enough voltmeter. Note also that the blower resistor values will likely change as it warms up, so the current measurement would only be accurate in the first few seconds.
 
i did this today!

6098056330_2fa975f3db_z.jpg
 
i did this today!

6098056330_2fa975f3db_z.jpg

There was a reason the stock connector has nonconducting plastic between those spades. If the big black wire on top is not shorted to blue/yellow now (hard to tell in the picture) it certainly will be once some dust or other crud builds up in the tiny space between them. Or it might arc. There really should be some insulation (other than air!) around those connectors.

Also, while it may just be the picture, that looks like a cold solder joint on the top big black wire.
 
i'm not finished. the op used hot glue, which i plan on doing but i didnt have any at the time
 
i'm not finished. the op used hot glue, which i plan on doing but i didnt have any at the time

That will work.

After looking at that top thick black wire more I think you definitely want to redo that one. Not only does it look like it has cold solder joint, but the clamp is on the spade at an angle. At least it looks that way in the picture.
 
Quick Question, How in hell do you remeve the Radio Center console ?? I tried to many time but no luck, the damn Wires that are attach to the AC controls dont give me enought room to do s*** --- any tips....... Thanks.
 
Quick Question, How in hell do you remeve the Radio Center console ?? I tried to many time but no luck, the damn Wires that are attach to the AC controls dont give me enought room to do s*** --- any tips....... Thanks.

1. Detach the climate control cables from their posts on the white plastic wheels. There is one on each side under the dash. To get to the one on the passenger side first remove the glove box (squeeze on one side and it will drop down).
2. About 6" back from the end there is a black plastic clamp holding those cables, pull them out of the clamp. (Carefully, that sucker has some very sharp edges.)
3. Tie a string to the end of each climate control cable so that you can be sure to route it back the way it was originally.
4. Aim the vents up.
5. Stick your hands in the vents and pull down and away from the console.
6. Detach the emergency flasher cable.
7. Pull the console out a little farther and detach the climate control wires.
8. Pull the console with the climate control unit out the rest of the way.
9. Detach the strings and tie them off somewhere, so that the path they mark is maintained.

On an automatic if the shifter is in park it will get in the way. Chock the wheels, put on the parking brake, and move it as far away
from the console as it will go.

Reassembly is the reverse, and you can use the strings you left in to help pull the cables through. Without the strings you will probably have to stick your hand into some tight places where there are lots of sharp metal edges.
 
Not quite the flicker problem, but my fan speeds 1-3 stopped working completely this morning. It will still work on four. Is this a resistor issue, or is it likely the switch?
 
Not quite the flicker problem, but my fan speeds 1-3 stopped working completely this morning. It will still work on four. Is this a resistor issue, or is it likely the switch?

Most likely the resistor. 4 bypasses the resistor and 1-3 go through it. Very easy to get to the resistor to check. (Under the glovebox.)
 
Back