Board Logo

radiator fan switching by the ECU?
pete wood - February 2nd, 2009 at 09:12 AM

after my thermatic switch died the other day, i'm considering using the ECU to switch the fans. I used to have the copper bulb/capilary/type adjustable switch, but I was never really happy with how that worked and the inherent risk of leaks. :td:

anyway, my qu is as follows.

I've been looking at the wiring diagrams for the ECU and relays and it looks like;

the battery switches the relay on (via a fuse) with the ECU acting as a switched EARTH.

Is this correct?

VWCOOL - February 2nd, 2009 at 09:36 AM

.. that's how it usually works. No big amp load on the ECU that way

ratbug - February 2nd, 2009 at 10:15 AM

Quote:
Originally posted by pete wood

the battery switches the relay on (via a fuse) with the ECU acting as a switched EARTH.

Is this correct?


Yep, thats it, There is a little relay and fuseblock that does the fan control. Its got 4 relays and 2 fuses. Grab this and everything should match up to the ECU.

pete wood - February 2nd, 2009 at 10:18 AM

mmm, chucked that out a few years back. I'll use the normal relays I have.

I just looked at the pinout diagram again. It say

on = 0v
off = battery volts

how does that work? wouldn't you have the fans running all the time when the car isn't turned on then?

ratbug - February 2nd, 2009 at 10:47 AM

Nah, when it is off, you will measure battery volts, but it will be reading it through the relay coil in which the other end is connected to 12v. The ecu works by pulling it down to 0v, so you then have a 12v potential across the coil, turning it on. Plus, you wire the 12v that drives the relay coil to Ignition, so when you switch the power off its definitely got no juice. But wire the power supply to the relay contact that switches the fan on, directly to the battery supply (via a fuse).

Kombicol - February 2nd, 2009 at 11:26 AM

my ECU does a great job of controlling the engine temp on the road.
I also have an over-ride switch on the dash. If I am going to be playing in the sand, I normally turn it on first, as it can get very hot very quickly, and there is a sllight lag with the ecu controlling it.

1303Steve - February 2nd, 2009 at 11:33 AM

Hi

I used the Davies Craig bulb type type switch when my Wasserboxer was using carbs, never had a leak.

I had it wired so that it the fan ran even with the motor switched off, just like most European cars do, because even though the motor is off the water will still circulate with a thermo siphon effect, hot liquid rises, cool liquid falls.

When I went to EFI and had the ECU doing the switching of the fans I still kept the Davies Craig switch wired in for cool downs, I feel that this is an important feature that only seems to be used by European car makers.

VDO make fan switches with common threads that could be used in your set-up.

Steve

pete wood - February 2nd, 2009 at 01:03 PM

the easy option is to just install another one. i'm just exploring the possibility of running it off the ECU. personally, I don't want to screw with the ECU if I can avoid it, but it would be cheaper to do it that way. My DC capiliary isn't leaking either, but it not a very robust system really. I suspect multiple removal/rebuilding sessions have caused the early failure. I do see your point about the fans running when the motor has been switched off though. After a long run in hot conditions it can be a very good thing. Honda did this also on their 4ws preludes.

k_love - February 2nd, 2009 at 02:53 PM

depending on your application you could give the ecu 12v+ on the A/c input
and the ecu will switch the fans to run constantly
this works on a SOHC jdm 1994 ej20

this is a good way to 'override' the switching on and off of the relays

pete wood - February 2nd, 2009 at 06:56 PM

someone has offered me a new switch for free, so I think I'm gonna wire it up for both.

1303Steve - February 2nd, 2009 at 11:14 PM

Hi Peter

When you turn your motor off the coolant still circulates, running the fan until it cools down helps the cool down process.

My biggest error with adjusting the Davies Craig switch was trying to keep the motor too cool, so the fan ran too long.

Steve

ricola - February 3rd, 2009 at 02:22 AM

I'd be concerned about the delay between fans turning on and the coolant returning to the engine, there is a fair amount of additional coolant line in most of our conversions and this would be even worse usign the stock ECU switching temperatures (just short of 100degC from memory)...
Rich

pete wood - February 4th, 2009 at 09:20 AM

i must say, that was my concern re using the ECU. also the guy who stripped my loom for me told me he never used the ECU to switch the fans on as he'd had lots of dramas with it.

GeorgeL - February 4th, 2009 at 03:10 PM

Quote:
Originally posted by pete wood
i must say, that was my concern re using the ECU. also the guy who stripped my loom for me told me he never used the ECU to switch the fans on as he'd had lots of dramas with it.


I wonder how long the delay is?

I can't find any data on Subaru pumps, but aftermarket electric pumps for Japanese car applications advertise a flow rate of 20gpm. A conservative assumption is that the stock pump is capable of half that. 20 feet of 1.5 inch coolant line holds a bit less than two gallons of coolant, so the delay between the fan going on and fan-cooled coolant reaching the engine is about 10-12 seconds.

Considering the overall thermal mass of the coolant, I don't think that this delay is significant.

1303Steve - February 4th, 2009 at 05:01 PM

Hi

I have no experience with using the Subaru ECU for fan switching as yet, but with my WBX bug I had the Autronic ECU switching on the fan, I guess all that would happen is that fans would be turned on before the hotter coolant reaches the radiator.

The only factory rear engine, front radiator car I can think of is the 1.9 & 2.1 WBX Vanagons (apart from water-cooled Porsches), these didn't use the ECU for fan control, they had a switch in the radiator, maybe we should copy what the VW engineers did?

Steve

ricola - February 5th, 2009 at 06:16 AM

It's not just the flow rate that would lead to a delay in cooled coolant getting back to the engine, but the fact that there is a lot more coolant to cool down in a converted car. Also, with the subaru switching temps being so high, there is less margin before you start boiling over. Personally I have my fans coming on a lot earlier than subaru, probably 85-90 degrees...

GeorgeL - February 7th, 2009 at 01:32 AM

Quote:
Originally posted by ricola
It's not just the flow rate that would lead to a delay in cooled coolant getting back to the engine, but the fact that there is a lot more coolant to cool down in a converted car. Also, with the subaru switching temps being so high, there is less margin before you start boiling over. Personally I have my fans coming on a lot earlier than subaru, probably 85-90 degrees...


A good point. A lot rests on how much you depend upon your fan for normal cooling. In a conventional car the fans are used only for the odd instances where ram-air flow is insufficient. In that case the turn-on temp can be relatively high. In the case where you are using the fan for normal cooling it should be on at a lower temperature Otherwise you'll end up with the fan sensor controlling engine temperature rather than the thermostat. Aftermarket ECUs make it easy to set the temperature, but if you're stuck with the Subaru ECU you'll have to use an independent system.

Depending upon electric fans for normal cooling is a bit dicy. Vehicles that do that (big buses, for example) usually drive their fans from the engine.

pete wood - February 7th, 2009 at 08:36 AM

Quote:
Originally posted by ricola
It's not just the flow rate that would lead to a delay in cooled coolant getting back to the engine, but the fact that there is a lot more coolant to cool down in a converted car. Also, with the subaru switching temps being so high, there is less margin before you start boiling over. Personally I have my fans coming on a lot earlier than subaru, probably 85-90 degrees...


mate, 85deg is way too low. 92degC is what I'd consider the bottom of fan switching. 88-90 is a good regular operating temp. if the factory feels it's ok to run the motor up to 100deg then it's probably fine. course you probably wouldn't want it running at 100deg all the time everyday, but now and then is fine. I had mine up round 100 last friday and yesterday and the car is fine. In fact, when I first had the car on the road it sat on 100deg on the freeway. Not ideal, but I haven't had head gasket or overheating issues....and I have the DOHC Ej25, famous for headgasket issues.

re using the ECU and flow rates, I agree with you. the ECU is programed to run the fans on a setup with LESS THAN HALF the coolent of a rear engine/front rad setup. The overall thermal mass of 10-12litres of coolent is significantly more than that of 4-5litres of coolent. So the fans are actually working harder/longer to have an impact. This agrees with what my V6kombi owning mate told me. His radiator(not front mounted)/engine only holds about 3-4litres of coolent. His fans run for 2mins and cool everything completely. Mine take a lot longer, so getting them on a little earlier (93deg instead of 95) is a good thing. Other thing is, the ECU only turns on one fan at 95. We need both fans coming on straight away with double the thermal mass so the seperate temp switch makes more sense there too. Having said all that, the nature of metal piping and such means that front rad/rear engine setups shed a lot more heat on the move than you might think.

Either way fans are only going to be a backup. The ECU switches the fans off at 100km/h (speed sensor) and with good reason too, fans stop the airflow instead of increasing it. And lets face it, if the rad can't cool the car with 100km/h of airflow you've got a bigger problem than the fans not working properly.

BTW, I drove my car in 42deg heat all afternoon yesterday. I sat on 89deg except for coming up just over 97 when I hooned it up a big twisty dirt road that was all hillclimb. Flicked the fans on and kept moving and it was fine.

Brad - February 7th, 2009 at 09:08 AM

I have used ECU switching on every conversion I have done and never had an issue. In the begining I used thermal switches as I had cioncerns but after quite a bit of testing it worked out better to use the ECU to control it and made the system "cleaner".

The ECU in an EJ 22 switches the fan on at about 92 deg and off at 85 deg. Due to the way the sip cooling system works this is fine.

The capacity in coolant does not really cause any issues from what I can see and as the temp prob is on teh outside of the engine it turns on as soon as the engine temp is where it needs to be. With the increased capacity it just means it turns on and off and the coolant moves around. BUT once it is at temp the variation is not that great.

I would use the ECU over an external control.

ricola - February 8th, 2009 at 01:17 AM

85-90 was only an approximation, we all know how inaccurate a lot of aftermarket and especially OEM temperature senders and gauges can be!
What I normally do is run the system up to temperature and wait until the cooling system has run out of sufficiently cool coolant and the thermostat is wide open and the temp starts to rise, then adjust the fan switch to come on a little after that. My aim is to avoid the fan being on all the time but not result in too much variation of the running temperature. Also, a lot of the subaru performance guys use cooler than stock thermostats, as cooler running temps (obviously within reason) will give more performance, hotter running temps give better economy.

pete wood - February 8th, 2009 at 04:38 PM

the stock thermostat in the newer cars is only 78deg as it is...and they want to go cooler than that?! :crazy: well there you go I guess