PWM cooling fan(s) control using ECM
 

I am trying to figure out what would be required to set up Pulse Width Modulated cooling fan control using the E67 ECU that came with my LSA crate motor.

The ECU that GM Performance sells to go with the LSA crate engine (part number 19259293) comes set up for On/Off fan control. I would prefer PWM fan control.

Of course I could find a stand alone PWM fan controller and run it independent of the ECU but I would prefer to have the ECU control the fans using more inputs than just the engine temp. (Such as AC pressure and ambient temperature).

Here is what I am wondering:

Can I use the PWM fan control module (Part number 10377609) that is used in the ZR 1 Corvette which also uses the E67 ECU and drive it with a signal from my ECU just like the ZR 1 Corvette does?

I'm assuming I would need to find the correct pin on the ECU and add a wire. If so, I have no problem with doing that. Does anyone know which pin that is on the E67 ECU?

I think the Trailblazer SS also used the E67 ECU. It uses pin #64 for PWM control of the cooling fan clutch. Might that be the same one the ZR 1 uses to drive the PWM fan control module?

Would the algorithms for PWM cooling fan control already be in my ECU or would they be completely absent? If absent, could I have a tuner grab that subroutine from a ZR 1 Corvette program and install it into my ECU?

Thanks in advance for any words of wisdom.

Simple answer is, yes - you can control a PWM fan controller with GM E67 controller (or E38 or E40).

Wiring is fairly simple - just need to wire PWM output from controller (can find wiring mech's for Corvette online pretty easily).

Obviously calibration needs to be changed as well from what controller was delivered with. Crate engine is disrete output, and you need PWM. There are several variables that need to be changed to make this work properly (including PWM frequency).

Fan output duty cycle can then be calibrated based on these inputs:
-Engine coolant temp - from ECT sensor
-AC pressure (will need to add production pressure sensor and wire as well)
-Intake air temp - IAT sensor in MAF

You can also calibrate 'after run' on fan to keep fan running at lower level after hot shut-down to reduce heat soak.

HP Tuners allows access to all these variables.

Dave

Dave

Dave, Thank you very much for your reply. You've given me a new head of steam to figure out how to accomplish the mission.

If I understand your response it means I need to:

1.Buy the PWM fan control module used on all C6 Corvettes

2. Find a pin out diagram for the E67 ECM like the one that came with my LSA crate engine. Best bet would be the pinout for the C6 ZR1 corvette.

3. Find a wiring diagram for a ZR 1 C6 Corvette that shows the wiring between the E67 ECM and the PWM fan control module.

4. Add a wire from the PWM signal output pin on my E67 ECM and run it to the PWM fan control module.

5. Purchase AC pressure sensor used in the C6 corvette and install in my system.

6. Have my ECM programmed to provide the PWM output signals I prefer to the output pin I added the wire to.

Did I understand correctly?

Thanks again for your willingness to point me in the right direction.

Just add step 5.5 - wire AC pressure sensor to ECM (which will also be on wiring mech for ZR1 Corvette) and you've got it.

Be sure to check origional crate engine harness - they run Fan 1 and Fan 2 output to fan relay - you need Fan 1 to go to PWM control module - and have fan power leads NOT running through relay - just a fuse.

Corvette fan (400 watt) pulls ~32 amps peak, so size fuse, wiring appropriately.

SRX fan (850 watt) we've used on Stielow's latest cars pulls ~65 amps peak.

Dave

Thank you very much for your advice Dave.

I'm curious if you would know this:

Does the software subroutine to drive a cooling fan PWM output signal already exist in the E67 ECM that came with my LSA crate motor such that it just needs to be turned on after we add the wires mentioned above or will we need to first install a program or "operating system" from a ZR1 application then modify that program with a tune for my LSA motor?

I'm wondering if the later approach would also require turning off a lot of other subroutines that the ZR1 application program or "operating system" will be looking for.

Thanks again for sharing your knowledge with the rest of us.
Lou

Algorithm for PWM fan control is already in your software - just need to select 'PWM fan control' instead of 'descrete fan control' - then set calibrations to how you want fan to operate (% duty cycle based on ECT, AC pressure, IAT).

Dave

Question of ignorance here, looking for a little education...

Can you use a PWM fan controller on any cooling fan, or does the fan need to be designed to handle the fast on/off switch of the PWM? In other words, would you burn-out a normal fan if using a PWM controller on it?

I haven't looked, but I'm now wondering if anyone makes an aftermarket PWM fan controller. I think the idea here is interesting, but at least from the GM implementation, it's not just a simple aftermarket conversion...

The high frequency signal goes to the control module - not the fan motor directly. This causes output voltage of controller to be reduced (proportional to PWM signal). So long as fan motor can run at lower voltage levels, should not be an issue.

There are a couple aftermarket PWM fan controllers but I have no experience with them.

Dave

Thanks for your continued sharing of knowledge Dave. It REALLY HELPS.

Since I am using a 2nd Gen Camaro radiator set up made by C&R that has two large fans, I plan to use two of the GM PWM fan control modules.

My thinking is that GM validated those modules for the current draw of just one C6 Corvette fan. Since I have two fans, it seems safest to use two modules vs. "hoping" one module will be able to handle the current draw of two fans.

I know that probably sounds like overkill but hey, I'm an engineer. It's in my blood.

Of course, my plan presumes that I can send the PWM signal from the E67 ECM to both fan modules.

Do you suppose that will be a problem?
Would you happen to know what the frequency and voltage amplitude of the PWM signal from the E67 ECM is?

Oddly enough, the toughest part of my plan appears to be sourcing the connectors that plug into the GM Fan control module. Best
I can tell, the only way you can buy those connectors from GM is as part of an entire C6 fan assembly whch lists for over $400!

If anyone knows where those connectors can be purchased separately I would appreciate them giving me a hint or clue.

I may have to connect wires to the module tabs and then pot the end of the module, effectively converting it to a pigtail version.

Thanks again Dave.

I'm glad I started reading this thread! Very very interesting, and far superior to what I did on my last project which was to use a relay to override and run the fans with pressure monitoring via a trinary switch.

Just to clarify (for me, anyway :) ) : the E67 in the LSA crate motor offering from GM (pn 19259293) can monitor the A/C line pressure and turn on the fans? Does it do that with PWM support (just running the fan enough to keep the high side pressure to certain number)? Or does it do that by running the fans full speed until the high side pressure drops below a threshold?

And all the while, it's keeping track of the coolant temp and running the fans appropriately for that too?

Sweet!

If you have crimpers...should be able to just buy metripak or weatherpak connector and make your own harness. Waytek is one source for such items.

No need for 2 fan control modules unless current demands from dual fans exceed capaicty for single module. I've run dual Spal fans from production Corvette fan module without issue (I will say that I have yet to ever use an aftermarket fan that comes close to performance of the factory designed and developed fans).

PWM output from ECM has to match what fan module is looking for. If you are using fan module from Corvette and match the calibrations in controller to a Corvette, will all work fine. 128 Hz is what Corvette fan controller is looking for.



Yes - AC pressure calibration just drives fan high enough to keep pressure within boundries set in cal. It's a table relating AC pressure in kPa to fan output %.

Fan duty cycle is controlled by 5 inputs: ECT (engine coolant temp), AC pressure, IAT (intake air temp), EOT (engine oil temo), TOT (trans oil temp). Highest request wins (in other words if AC pressure is requesting 50% DC and ECT is only 30%, 50% is what is commanded).

Most installations do not monitor oil temp (decrete input into ECM) and trans oil temp is either from descrete input of trans temp into ECM, or over CAN bus from TCM (transmission control module).

If you use PCS (Powertrain Control Systems) trans controller for automatic trans, trans oil temp (TOT) is broadcast over CAN messages and will work with crate engine controller kit to drive fan output.

Typically, ECT and AC will be primary drivers for fan DC - unless you have an unbalanced system from cooling perspective.

Dave

We are still trying to figure out how to handle integrating the Vintage A/C system cooling fan requirements into our overall cooling strategy. This is like that Whack-a-Mole game. Every time you think you have a handle on how to make everything place nice together something else pops up.

Oh well, that's what makes it fun right????

First let me remind you that today I am just a newbie trying to figure this all out for the first time. Dave, who thankfully has chimed in on this thread, has numerous successful installs under his belt including Stielow's latest cars so he is MUCH more credible than me. Bottom line, don't follow my lead until you see the final results. I will keep you posted as we learn.

Kevin and Trevor of V8TV are keeping me honest and out of trouble too.

Getting back to your questions about how the ECM requests cooling fan speeds and how we hope to integrate the Vintage Air system into our cooling fan strategy:

Here is what I have been able to find regarding how GM sets up the cooling strategy in the ECM for PWM control of cooling fans:

Engine coolant temp. strategy:
Cooling fan starts at approx. 204 Deg. F (at 10% duty cycle) and steadily increases with temp up to a max of 90% duty cycle (which they consider full on or high speed) at 235 Deg. F and higher.

A/C Pressure strategy:
Cooling fan starts when A/C pressure is approx. 160 psi and reaches high speed at 360 psi.

Engine Oil temp strategy:
Oil temps above approx. 302 Deg. F will trigger high speed fan on.

Transmission Oil temp strategy:
Trans oil temps above 270 Deg. F will trigger high speed fan on.

Vehicle shut off strategy:
If engine coolant temp at key off is greater than 235 Deg. F or the A/C pressure is greater than 249 psi, the cooling fan duty cycle is set to 50%. If the coolant temp drops below 230 Deg. F, and the A/C pressure drops below 241 psi, the fan will shut off. Otherwise, the fan will automatically shut off after 2 minutes regardless of temperatures.

When multiple cooling fan requests are received the ECM uses the highest cooling fan speed of all requests.

Now, according to Vintage Air,

We need to find out if the fan algorithms can be tuned to request full speed fans at 260 PSI instead of the GM setting of 360 PSI because if pressure in the Vintage Air system is allowed to go much higher than 275 psi, the ability of the system to absorb heat is greatly reduced.

Absent of that, they would be concerned that we will not have efficient cooling because the GM A/C strategy is designed for a fixed orifice tube expansion with a variable displacement compressor, as opposed to the thermostatic expansion valve system Vintage Air employs.

The Vintage air system does have an on/off cooling fan control. It calls for fans on at 254 psi and has a safety shutoff for the system at 406 psi in case there is a fan failure or a blockage develops in the system.

This means we need to check into a few things as follows:

1. We need to make sure that we can turn off the ECM algorithms for any temperature sensors we might not use without causing the ECM to get confused or angry with us. For instance, we do not presently plan to sense the trans. temp or the oil temp through the ECM even if those algorithms reside in our version of the E67 ECM. We only want to "turn on" the coolant and the A/C pressure algorithms that we believe/hope to be resident in the LSA crate motor E67 controller but "turned off".

2. We need to make sure we can reset or tune the A/C pressure settings to emulate the way the Vintage A/C system controls cooling fans.

3. We need to verify that we install the GM A/C pressure sensor in the correct location within the refrigeration circuit.

4. We need to think the whole strategy through again to see what we might be missing.

Of course, we may still need to fall back to a more conventional cooling fan strategy if this all proves undoable for some reason. It's always good to have a Plan B.

Lou, all this stuff is amazingly topical to me as well. I've got a good buddy (Jeff Tate, one of my fellow mods over at PT.com) who works the parts counter at a Chevy dealership near Atlanta. He can help source all the parts needed here: PWM module, fan, connectors, A/C sensor and connector, plus the pins needed to add the wires to the ECM connectors.

I follow the need to disable transmission oil and maybe engine oil temps from the algorithm, plus change the parameters from 360 psi to 260 psi on the pressure side of the refrigerant circuit. I'm now beginning to wonder about fan dimensions, and how it will fit in a 69 Camaro body. I'm not sure sure the difference between the V/A compressor and the GM matters that much... as long as the fans generate enough cooling to keep the high side pressure below 300ish psi.

Of a greater concern is that the ECM has no way to turn off the compressor. That's a function of the trinary switch -- which only passes on the compressor on request as long as the high side pressure is under 260 psi.

So another question would be: can the E67 be made to take in the compressor request from the dash switch, and only signal the compressor clutch to engage if the pressures are safe? And to withdraw that request if the pressures exceed a programmable threshold (say 325 psi)?

E67 can and does control compressor - in factory installations. Issue is it gets AC request from HVAC module over CAN bus. And I don't see anyone that is going to solve that puzzle anytime soon.

Best solution is to allow Vintage Air (or whatever) unit control compressor, and ECM control fans. Again, you can set fan DC based on AC pressure (which is why you need to install sensor in AC line and wire to ECM).

Thanks again for all the insight and answers. I love this stuff!

If the V/A (or ISIS) controller manages the compressor, then there must be a sensor to know when to turn off the compressor when liquid side pressure gets too high. So we're back to a trinary switch (maybe just a binary switch?) to do that -- negating one of the key advantages of having the ECM manage the fan. I suppose it's still better for the ECM to manage the fan (does it soft-start the fan to reduce inrush?), but this hybrid system will have two pressure sensors that aren't integrated.

I'm going to talk to ISIS about their V/A controller and whether or not it can be modified to support the CAN-bus protocol to be able to talk to the ECM. I'm sure Jay's question will be: what is the CAN-bus sequence to tell the ECM to enable the A/C clutch?

Edit: I realize that just the CAN-bus electrical connection isn't enough. GM has a proprietary protocol that would have to be reverse-engineered to get this to work. I also asked ISIS about their ability to read CTS and A/C sensor data and manage the fans. I'll report back.

Tried to read all of this, not all of it sunk in & now my head hurts. In a nutshell, can I use a vette fan & PWM module & hook it up to my GMPP LS3 controller kit with extra harness & retune the ECM to work with the PWM fan instead of just on/off?

Simple answer - yes

Dave,

When my GM PWM fan modules arrive I am going to play with them on the test bench to learn more about how they work and respond to various input signals.

Thanks to you, I know I need to provide them with a 128 Hz PWM signal which will be no problem.

I can measure the impedence of the module when it arrives but I don't know how to determine what voltage range of PWM signal it will be looking for

Do you happen to know what voltage the module is looking for in the input signal it receives from the E67 ECM?

Thank you so much for your knowledge. You are the Man!

ECM signal is switching to ground - not voltage.

Most output functions of ECM are that way: injectors, coils, etc. Ground signal is switched with power supplied directly to device.

One notable exception: fuel pump. That's a 12V signal.

DC Controls has a stand-alone PWM that can handle 85A continious. I had to change to an FK55 for the MarkVIII fan since the initial turn-on current was melting the ends of the connector body.

http://www.dccontrol.com/fk85.pdf

I'm learning a ton here. Thanks for the great discussion and info sharing.

Now imagine sitting in a bar listening to a conversation between Mikels and Stielow. :willy: :bang: :willy:

That would be the coolest bar I would have ever been in......

We Engineers think this stuff is intoxicating...we don't need no stinkin bars...

Well OK, in Rolla we did need our bars because there were no women.

I guess we are a pretty sad lot

Hell, I think this was the highlight of my week!

If I missed it it, I apologize. But what compressor are you running? The factory compressor is controlled by the ecm not just at the clutch but it varies the pitch on the vanes right? I didn't think that variable compressor is compatible with the VA unit?

If you are running a sanden compressor then why not leave the ecm out of the equation for the ac? You could run a 5 volt ref to one side of the trinary and from there to the ecm to tell it that ac is on. It would only make when the specified pressure is reached and you could set it to run at, say 40% on the pwm when the AC calls for it.

It really doesn't take that much fan to cool off the condenser. When we charge systems here, the fan only cycles for 8-10 seconds at a time and doesn't cycle again for at least another 90 seconds. You can watch the high side pressure drop quickly, you would swear it had a leak. So in my mind it would be simpler to do it that way and not have to worry about the oem pressure sensor.

In my case, it's a Sanden. I was looking for a solution that doesn't cycle the fan at max speed for 10-15s every 90s... but would run a lower fan speed for longer times. I don't know of any aftermarket fan controller that supports that, so I thought having the ECM do it would be a nice integrated feature.

That is our current thinking too. We are also using the Sanden compressor.
We plan to let the Vintage Air system control the compressor as it is designed and use the ECM control to the fans since it looks like that can work.

So you are trying to eliminate the on/off cycle. If the fan soft starts and runs at half speed for the AC I think it would work. This is really only critical during warm up, I think, once up to temp wouldn't normal fan cycling move enough air to keep the trinary switch out of the loop?

That's exactly how crate engine controller will work with the pressure sensor in AC system and PWM fan control. Will modulate fan output only enough to control pressure. This is how Jackass, Red Devil, Mayhem and Camaro XV are set up.

I hate listening to a car and all you hear is the fan screaming.

Ok Charlie - next time we'll use small words and speak slowly (or buy you enough margarita's you won't care). I get this same abuse from my wife and kids - now you too?

I don't know what your talking about - there were tons of women in Rolla, just not many of them.

Exactly! We've got an analog input (ECT goes from 180 to 181, or refrigerant goes from 175 to 176), but a binary response: the fan goes from 0 to full speed.

How about this question: do Jackass, Red Devil, Mayhem, and XV use a trinary (or binary) switch? Let the ECM manage the fan, but the fail-safe requirement when the refrigerant gets dangerously high is still allocated to the binary/trinary switch?

I'm from the age: If it has spark and fuel it should run.

I thought we were about the same age....

This discussion "sparked" my interest, "fueled" my knowledge and I'm so full of "....." I gotta run.

Yes - compressor controlled by Vintage Air (including bi/trinary switch), with PWM fan controlled by ECM.

Typically fan request for AC pressure never more than 30-40% (and only that high on very high demand AC environment). At that level, can barely hear fan running.

At full tilt (90% DC for most PWM fan controllers), an 850W fan tends to make a little noise - but only see that level during track use due to ECT, IAT or oil temp. I'm ok with fan making some noise then - especially if controlling temps.

Excellent. I think I have a plan then. Great stuff, great discussion.

thanks!

Very interesting read!

Would this type of setup work on my C5 LS1, MarkVII, VA in my Nova? Or is my old school PCM not capable?

Thanks guys,

Dhamen

No real easy answer to your question and I'm no expert by any stretch but here is what I "Think" I have learned that may offer some insights:

So far, I have found evidence of 3 GM ECM types or series that have been used in applications that used some sort of PWM control of cooling fans and/or fan clutches as follows:

E38 ECMs have been used in C6 Corvettes and use PWM control of the cooling fan using the fan control module.

E40 ECMs have been used in some Trailblazer SS applications and use PWM control of the fan clutch.

E67 ECMs have been used in C6 ZR1s and use PWM control of cooling fan using the fan control module.

There may be more. These are just the ones I have found in my research so far.

Here are pics of these three ECM types:


E38
http://i1140.photobucket.com/albums/...ps9ed02ff9.jpg


E40
http://i1140.photobucket.com/albums/...ps2d322fe2.png


E67
http://i1140.photobucket.com/albums/...ps1d165872.jpg

You will notice the easiest way to tell them apart is by the arrangement of connectors and fins.

I have been told the ECM itself is analgous to your desktop computer. By that, I mean within reason, each computer of the same type or series is theoretically "capable" of doing the same job as long as the right software programs are loaded into them. In other words, two computers that are made the same are only capable of running the same operating system or programs if the same software is loaded into them.

Following that logic, if one computer (or ECM) of a specific model number has the internal hardware to do a job in one application, it should be able to do it in another application provided the right software is resident inside, that software is enabled, and it is wired to the correct inputs and outputs.

So, using deductive logic, (which you should treat as suspect until you or someone who has actually done it validates or disputes it), would suggest that if your ECM is one of these types, it should be "capable" of PWM control of cooling fans because it has been used to do so in other applications.

The next challenge would be to find out if your ECM has any software inside that can be enabled to provide PWM control or if it would be possible to load that software into your ECM and then turn it on.

If all that looks possible, you would then need to figure out if you have wires in your harness for the appropriate input and output pins, whether the fan control module will work in your application, if you have the right input sensors, etc.

Like I've said earlier, first time retro-fits are like the old whack-a-mole game. Solving a problem in one area can cause another problem to pop up somewhere else in the system. It can often be an interative process with inherent tradeoffs.

Unless you are a gluten for punishment or just like the challenge of figuring things like this out yourself, getting the advice of or hiring a professional is highly recommended. That is exactly why we plan to follow Dave's trail as much as possible.

I am definitely not a professional or experienced resource. I'm just a fellow newbie trying to validate what I think I am learning as I go.

Good luck!

Thanks for the reply. Maybe when I get further in my project I'll play with it and see what I can get done.

BTW your engineering degree is showing. :D

Here's the report back from ISIS:

So does anybody know: does the GM E67 use J1939 protocol over CANbus?