[plain OM]

This is a follow up to two recent threads relating to the air conditioner compressor. The first thread is http://www.subaruoutback.org/forums/66-problems-maintenance/23973-red-dust-around-c-compressor-whats-deal.html where the discussion was about red dust apparently coming from the compressor. The second is http://www.subaruoutback.org/forums/66-problems-maintenance/26071-c-intermittent-failure-flaky-relay.html, where a bad AC relay caused poor AC performance.

My car is going through it's fourth summer since new and, for the first time, I noticed some red dust in the area behind the AC compressor. (Photo attached).

Having been involved with the other two discussions, I decided to do some pre-emptive diagnosis. It was pretty short. As I noted in a post in the first link,

According to a Subaru HVAC training manual, the signal to the magnetic clutch solenoid coil "must not drop below 10.5 V for the coil to be effective". [Lower voltage/current to the coil] would lead to slippage, and perhaps a source of the highly noticeable red dust. The problem of low voltage/current could also be caused by something as simple as a bad electrical connection at the compressor or along the way.

I checked continuity in the supply and to ground and it was fine. So I pulled the AC relay, and tested it by applying 12 V to the coil terminals and measuring the resistance across the main switch terminals with a low-Ohms meter. Each time I applied 12 V, the readings differed, just as halfapie reported in the second linked thread.

I also pulled the sub-fan relay, and found the connection to be no more than 0.02 Ohms consistently.

Given the requirement for sufficient current to the clutch, the obvious poor connection being made at the relay, and the absence of red dust until this year, I'm pretty well convinced the red dust is from clutch slippage.

My reasoning is based in part on the fact that there was no dust on the belt or pulley. But more important is the fact that the clutch needs sufficient current to stay solidly engaged. The clutch coil measures 4 Ohms. At 12 Volts, this means it draws about 3 Amps. (This is consistent with the 35 Watt power consumption rating in the FSM.) For the voltage to be only 10.5 Volts (assuming the system voltage is 12 V), the current would be about 2.6 Amps, which isn't much less. It would take less than an Ohm of resistance in series with the coil to reduce the current to this critical point. Even when the system voltage is around 14 V (engine rpm above idle), it would take less than 1.5 Ohms to reduce the current to the 2.6 Amp limit. The measurements on my relay were above this.

I will get a replacement relay and clean up the area in behind as best I can. I suspect that there won't be any deposits afterward, at least until the relay begins to fail again.

Also, as soon as I have the new relay in, I'll crack open the old one to see if I can find exactly what is happening, and will report here.

In the meantime, if you notice red dust behind the compressor, start by checking the AC relay. And, in my case, I will add a periodic relay test to my routines.

Edit: opened the old relay today -- see post #3 below

 

[halfapie]

I agree with this analysis. I did not mention this in my post http://www.subaruoutback.org/forums/66-problems-maintenance/26071-c-intermittent-failure-flaky-relay.html, but I also had lots of red dust around my compressor. There was no noticeable dust at 80k miles, but lots of dust at 85k. I have yet to clean it off though...

It seems that for most people, the AC relay fails before the sub fan relay even though they are identical relays. I guess the AC clicks on/off more frequently than the sub fan, causing more "wear and tear" on the relay?

 

[plain OM]

The defective relay itself:

The contacts on the relay are badly tarnished.

I've attached photos. The third photo shows the inside, with the large energizing coil (measures about 156 Ohms) and the fourth shows what are fairly substantial contacts. However, note the blackened contact surface. The mating contact, which is not visible, was equally black.

I used my relay contact burnisher to try to clean the contact points, and the connection was improved, but not as good as the sub-fan relay. In this case, the best I could do was a fairly repeatable 0.03 Ohms. That's not too much for this circuit. I hope to have a replacement tomorrow and will see what it reads -- perhaps they vary from unit to unit.

I should add that I don't see any transposing of contact material, which is a sure sign of arcing. But I can't see the contacts face on, and there might be signs. In this regard, the relay is opening a very inductive circuit -- the large clutch coil and given the current flow, when the contacts open, the coil will generate a voltage spike. That would usually cause arcing, unless there is something to supress it, which I don't see. But I suspect this might be a reason the AC relay fails before the sub-fan relay of the same model. (Have to look at the specs for the sub-fan motor.)

Edit: see also post #5 below

 

[plain OM]

. . . It seems that for most people, the AC relay fails before the sub fan relay even though they are identical relays. I guess the AC clicks on/off more frequently than the sub fan, causing more "wear and tear" on the relay?

I was thinking about this earlier and the clicking on and off is a possibility, but from a different angle. The fans come on when the compressor comes on, so they work together. And the fans (and their relays) actually might be going through on-off cycles more often than the AC because of their operation when the AC is not even being used.

However the refrigerant system has a pressure switch which detects when there's too little, or too much, pressure on the high side. Too low pressure is often caused by a loss of refrigerant, but it can also be caused by a compressor that isn't running as strongly as it should.

Is it possible that if the clutch is slipping due to too low voltage (current) as a result of a flakey relay, the high side pressure might be only marginal such that with small changes in engine speed, or voltage to the clutch, the switch turns the compressor off, but then, because the control panel still is calling for cooling, restores it soon after, perhaps only to go through the same cycle again? This might be also the cause of presumed excessive cycling, as others have reported.

What seems to be evident, however, is that the relay is something to routinely check or just replace. And from what you say about there not being red dust at 80 k but plenty at 85 k,when the problems with the AC were apparent, the connection between the bad relay and the red dust appearance seems even more solid.

 

[plain OM]

Got the replacement relay yesterday and installed. I noticed when I bench-tested before installing that 1) the contact resistance readings were rock solid -- 0.01 Ohms every time, (unlike the old one, even after cleaning the contacts) and 2) the "click" sound was much more "crisp". The sub-fan relay was also like this, so the AC relay really wasn't working properly.

I disassembled the old relay and examined the contacts. Contrary to what I had mentioned earlier, there are signs of arcing, and these are apparent in the attached photo. This is after I had tried burnishing the contact surfaces while the relay was still intact, so the burned area must have been worse before. That would be enough to cause the intermittent connections.

The service manual has specifications for the compressor clutch gap. This varies with the compressor model, but is important for the clutch to be able to engage properly. If a bad relay has caused excessive clutch surface wear, then replacing a defective relay might not resolve a poor AC performance issue because the clutch might still not work properly.

For my 2007 with the Denso scroll compressor, the clutch gap spec. is 0.3 - 0.6 mm.

 

[subarufan]

Having read a few threads about the A/C Relay what does anyone think about buying a NON-OEM relay since it seems the Subaru ones are not lasting long?

Just curious.

 

[plain OM]

I had thought of that, especially because the list price here for the Subaru-supplied part is $27 Cdn! I asked (and showed the old one) at several car part outlets here (incl. NAPA, Carquest) and none had a listing for the relay (whether for AC or sub-fan application.) I also tried to find the specs, but the Mitsuba website wasn't much help. However, that doesn't mean that there isn't one that could be used in its place.

There's quite a few threads in other Subaru-related forums related to relay failure, but none that I read referred to using a non-oem replacement.

The real problem is the high back emf (transient voltage) developed in the clutch coil when the relay opens. The normal way to deal with this would be to install a suitable power diode across the coil with its polarity set to dissipate the unwanted voltage. However, this will also slow the change of the magnetic field, possibly leading to slower clutch engagement or release. This would be counter-productive.

Incidentally, none of the local dealers in my area normally stock the relay, despite the fact that my 07 uses seven of them, and probably the same in all the 05-09s. I suspect the price is a factor. I wonder if U.S. dealers do the same, or if given their lower cost, normally stock them. If so, I'll probably try to pick up a few spares the next time I'm in the U.S.

 

[subarufan]

Yea maybe the relays are the same up to 09 so the patent probably still exists as to why the aftermarket hasnt made any yet.

The A/C relay is $7 bucks. Maybe should order a box of them!!

 

[halfapie]

I actually installed an aftermarket relay in my 05xt. The dealer is too far from me and I didn't want to bother with the trip. The local auto parts store did not have relays listed under Subaru, so I got a 4-prong relay. Since it is a generic 4-prong relay, it does not have the plastic tabs that stick out to fit the grooves in the relay holder.

Anyway, it works fine for me.

 

[plain OM]

so I got a 4-prong relay. Since it is a generic 4-prong relay

Can you provide additional details about the relay, such as brand and model or contact ratings? (Often printed on the case). Where did you buy it, and, of course, price?

 

[highest_vision]

Some insight as to why the a/c relay wears faster than a fan relay - cooling fans should not come on below a certain speed, typically 15-25 mph. Therefore at speed your cooling fans (main and sub) will not operate but your compressor, naturally, will cycle at a similar frequency as idle.
This is an awesome find. Subaru is definitely lacking in this department.
James

 

[ETC]

Doesn't the sub fan come one regardless of temp when the ac compressor is engaged ?

I think the point he's trying to make is that the relay still "works" but, may not be transmitting enough power to fully engage the clutch. The fan relay could in theory do the same thing but still turn the fan ... if the voltage was reduced marginally you would probably not notice the difference in speed/torque.

 

[plain OM]

Some insight as to why the a/c relay wears faster than a fan relay - cooling fans should not come on below a certain speed, typically 15-25 mph. Therefore at speed your cooling fans (main and sub) will not operate but your compressor, naturally, will cycle at a similar frequency as idle.
This is an awesome find. Subaru is definitely lacking in this department.
James

Not quite.

The cooling fans do, and should, come on at slower speeds; that's when they're needed. At higher speeds, there's often enough air moving into and through the engine compartment to keep things cool. The attached chart is for 2007 (2005-9). Earlier versions are a bit different; they switch one fan on, then both, but the idea is the same. The fans (or at least one) always come on with the AC at slower speeds.

If the fans always come on with the AC, then, based on this alone, the sub-fan relay should wear at least as fast as the AC relay. Moreover, the fan relay handles more than twice the current of the AC relay; the fan motor is rated at 90 Watts on my 07, whereas the AC clutch coil uses 35 Watts, so the effect of current, if any, should be greater on the fan relay. Finally, if the fan runs whenever the AC is on, then it might actually run more often than the AC because there are times when the AC isn't being used, but fan cooling of the engine coolant is needed.

Later today I hope to make some additional measurements on the coil and fan motor to see if the idea of back emf (transient voltage) from the coil could be a factor.

EDIT: Added second chart showing sequential fan operation.

 

[bigdadi]

Plain OM -

That's how I found this thread by 'advance' search of 'red dust' in title...

Today I just noticed the red dust when I discussed with Subie service manager re where my 2007 PCV valve is. I asked him what the red thing is and he said "I'm not sure". then he told me the part with red dust is the A/C compressor.

My A/C is running fine so far and I did have the defrost and def/lower 'on' a lot during winter to keep the windshield from foggy. Can you point me to the location of that defective relay is. For about $7 to $10 in US I want to change it right away and see if that red thing coming back. I have the Gold plus extended warranty but it is ded $100 I don't want to use Ext Warranty if it is simple job that I can do or the total cost is < $100.

The Compressor oil is DENSO OIL 8 ... is this in red color?

Since you have 2007 OBW also can you show me where the PCV valve location as I want to put seafoam to the intake from the PCV valve to do some internal carbon cleaning. The brake booster vacuum hose is easy to locate but it goes to a specific cylinder, not to all cylinders, so I don't think it is recommendeed to use that vacuum hose.

 

[Ruperts Trooper]

Plain OM -

That's how I found this thread by 'advance' search of 'red dust' in title...

Today I just noticed the red dust when I discussed with Subie service manager re where my 2007 PCV valve is. I asked him what the red thing is and he said "I'm not sure". then he told me the part with red dust is the A/C compressor.

My A/C is running fine so far and I did have the defrost and def/lower 'on' a lot during winter to keep the windshield from foggy. Can you point me to the location of that defective relay is. For about $7 to $10 in US I want to change it right away and see if that red thing coming back. I have the Gold plus extended warranty but it is ded $100 I don't want to use Ext Warranty if it is simple job that I can do or the total cost is < $100.

The Compressor oil is DENSO OIL 8 ... is this in red color?

Since you have 2007 OBW also can you show me where the PCV valve location as I want to put seafoam to the intake from the PCV valve to do some internal carbon cleaning. The brake booster vacuum hose is easy to locate but it goes to a specific cylinder, not to all cylinders, so I don't think it is recommendeed to use that vacuum hose.

The relay is among all the relays/fuses under the cover by the battery - the legend on the inside of the cover shows which relay it it.

 

[bigdadi]

Ruperts Trooper

Thank's a bunch ... I will check it out.

 

[highest_vision]

My apologies. I mistakenly wrote that the fans will never come on, but I intended to write they function as you have stated. On at low speeds, off at higher.
I also know that, after looking at your charts and my previous experience that the fans will not cycle near as much as the a/c compressor relay. I do not know what operating temperature for your MY car is but mine is very low (aprox 180*). To get to the point of fan operation at speed would require considerable heat load. I also know that I had no fans whatsoever for three months and possibly longer. The only problems I experienced were at a standstill. Nothing at cruise, in the sun, in 85-90* temperature climbing maybe a 7* 2 mile hill.
The a/c compressor, however, will cycle at a fixed rate (in a mechanical system) regardless of where you set the temperature. This will vary by model of compressor and system design and I haven't done any a/c work on a subaru to factually tell you what they will do. One cycle a minute would be a fair guess. This cycle rate will remain relatively flat regardless of engine or vehicle speed.
I also noted that your second chart indicates you have a single, two-speed fan on your car rather than a dual one speed setup. If you consider the draw of the fan at low speed (versus the 90 watt high) this would increase the life expectancy of the relay even more so. The exception to this would be if the relay is modulated to create the low speed rather than fed a lower voltage. This seems unlikely because you stated the compressor and fan relays were the same.
James

Not quite.

The cooling fans do, and should, come on at slower speeds; that's when they're needed. At higher speeds, there's often enough air moving into and through the engine compartment to keep things cool. The attached chart is for 2007 (2005-9). Earlier versions are a bit different; they switch one fan on, then both, but the idea is the same. The fans (or at least one) always come on with the AC at slower speeds.

If the fans always come on with the AC, then, based on this alone, the sub-fan relay should wear at least as fast as the AC relay. Moreover, the fan relay handles more than twice the current of the AC relay; the fan motor is rated at 90 Watts on my 07, whereas the AC clutch coil uses 35 Watts, so the effect of current, if any, should be greater on the fan relay. Finally, if the fan runs whenever the AC is on, then it might actually run more often than the AC because there are times when the AC isn't being used, but fan cooling of the engine coolant is needed.

Later today I hope to make some additional measurements on the coil and fan motor to see if the idea of back emf (transient voltage) from the coil could be a factor.

EDIT: Added second chart showing sequential fan operation.

 

[plain OM]

James:

Your example of the cooling capacity of air passing over the engine at higher speed without any fans is right on the mark. If we had a light on the dashboard indicating when the fans were running, we'd see that they would rarely come on. This was one of the reasons for changing from belt-driven fans to electric. The belt driven system acted as a load on the engine at all speeds, even when not needed, whereas the electric can be more easily controlled. Less load = greater gas mileage.

I also noted that your second chart indicates you have a single, two-speed fan on your car rather than a dual one speed setup. If you consider the draw of the fan at low speed (versus the 90 watt high) this would increase the life expectancy of the relay even more so. The exception to this would be if the relay is modulated to create the low speed rather than fed a lower voltage. This seems unlikely because you stated the compressor and fan relays were the same.

No, my 07 has two fans. In this configuration (chart marked 2007), both fans are either on, or off, at the same time. The relays (there's 3) switch the connections to the fans -- for slow speed they're in series, and for high speed they're in parallel.

In the other chart ("marked sequential control") which is for some earlier models (the chart is from circa 2000), the ECM will turn on one fan, and if needed, add the second.

I believe the sequential fan control (one on, or both on) was used on cars with the horizontal engine radiator. Water flowing across from one side to the other would always pass in front of at least one fan when it was on, and two if necessary for greater air volume. However, with a vertical flow radiator and fans horizontally side-by-side, water going down the tubes is either in front of one fan or the other. Consequently, if only one fan was on, the water in the tubes in front of that fan would be cooled, but the water going down in front of the other fan would not. So it was necessary with the vertical flow radiator to have both fans on. This way all the vertical tubes would be exposed to moving air from the fans -- the difference being that the fan speeds, and therefore air volume, can be changed.

As far as relay damage goes, I'm fairly certain the basic current load of the compressor or fans is not the cause. The relay contacts are quite substantial, and compared to others I have in my shop for which I have capacity ratings, are probably capable of handling much higher currents than a fan motor or compressor clutch would require.

Relay contacts are rated in terms of both current capacity and voltage. The relay used in my 07 for the sub-fan and for the AC clutch coil probably has a rating of 15 or 20 Amps at 12 V. So it should not be damaged in any way when working in a circuit with these levels.

However, this rating applies when the relay is operating in an resistive circuit, whereas the clutch coil and the fan motor are inductive by design. Inductive circuits, and especially those with electromagnets, which the motors and coil are, can generate high voltages when the current to them is turned off. Tests reports I've read suggest that the voltage can reach well into the hundreds of volts in what is otherwise a 12 V circuit. It is this high voltage spike that can cause arcing across the relay contacts as they open; the arcing burns the contacts and reduces their ability to make a good connection.

(I believe, however, that the fan motor, even though it normally draws more current than the clutch coil, does not generate the same degree of voltage transient as the coil does, and this is why the clutch relay fails whereas the sub-fan relay seems to have a longer life. However, I have yet to confirm this difference. Perhaps others here with greater motor engineering knowledge could help.)

 

[mrk610]

Does anybody have the part # for the relay . Got interested and looked at mine i also have a load of red dust. Car is an 06 with just over 50 k .
Thanks In advance

 

[plain OM]

Sorry, I should have put it in when I reported on taking it apart. For my 07, and I believe for the 2005-9, it's 82501AG05B, but you should verify this with your parts supplier as Subaru often changes parts, even within a MY.

I believe a build-up of the red dust over a relatively short time can be an indicator of a failing AC clutch relay. In my case there was a build-up from nothing in less than a few months (car in it's 4th summer). I've since replaced the relay and cleaned off the red dust residue. Now I'm monitoring to see if the dust reappears.

By the way, I found compressed air along with an old toothbrush effective in removing the dust from the metal and plastic parts.