When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Does anyone know what the minimum and maximum duty cycles the GM e38 ECU pumps out on the PWM fuel pump controller pin in real world use?
Via HPTuners I can vary it anywhere from 0 to 100% when testing, however in real world use what is the general range that the e38 dumps out in terms of duty cycle to the fuel pump controller?
I can see that it is 128 Hz frequency, that frequency does not seem to be able to be changed. Fair enough, no beef there, however it would be nice to know the real world range on the duty cycle.
For example on the PWM fan output the e38 only ever seems to dump a range of 5% to 91%. No more and no less. Is the fuel pump duty range also similar? That is, does it only ever pump out 5% to 91% range?
Reason I am asking is that I am coming close to solving my fuel pump controller with a Dorman 590-001. This fuel pump controller is commonly found on a buttload of Ford products, V8's included, so it is readily available and cheap (50 bucks for the Amazon/eBay clones) and has a nice looking cap and inductor on the output. Early tests show it is great at passing through close to the input voltage at full tilt. Only drawback is that it looks like it accepts just a 5% to 50% pwm input to then run the output from 0 output to 100% output. This is fine, because it looks like I have solved it with a simple Arduino solution to halve my duty cycle, so the 0 to 100% pwm coming out of the e38 now goes through my arduino and (for example) 80% duty cycle coming from the e38 then comes out of the Arduino at 40% duty cycle and the Dorman 590-001 pumps out 80% of its working voltage range and duty cycle. A bit convoluted I know, but it works. As I am now fine tuning my arduino sketch it would be good to know the actual working range the e38 pumps out. Even though I have said that the Dorman controller only takes half the duty cycle, in my arduino sketch it is a scalar more like 1.8 instead of a flat 2.0 that is proving to work well (on the bench at least).
So it sounds like instead of going to an ECM, your temp sensor goes to an MSD box. Most likely, the MSD box has a series resistor going to 5V just like the ECM would and the other sensor wire is likely going to ground. The way to find the value of the MSDs series resistor is to measure the voltage at any of the temperatures that are in your chart.
Lets do an example. Let's say you measure your voltage at the sensor while it's connected to the MSD box and it is 120 degrees. According to your table the sensors resistance is 1338 ohms at 120 degrees.
If you measured 2 volts then Rmsdbox = E / I = (5 - 2) / ( 2 / 1338) = 2007 ohms
If you measured 1.5 volts then Rmsdbox = E / I = (5 - 1.5) / ( 1.5 / 1338) = 3122 ohms
Make some voltage measurements and let's see what the resistance in the MSD is. Once we know what the MSD series resistance is, we can figure out the voltages at 180 degrees and 198 degrees and pop that into the code for starting up the fan and full power.
Hi Carl,
It has been a while since I was able to get back around to this project I was working on. To make things simple, we decided it was easier to take the MSD box out of the picture. We ran an additional ECT sensor in the opposite cylinder head so that the Arduino has its own sensor. It is a 2 wire GM coolant temp sensor. I found on an another forum that the the sensor needs 2,750 ohms of resistance (instead of the 10k resistor for a normal thermsistor) and this seems to work perfectly. After uploading the code to the Arduino, I am now experiencing an issue where the fans are turning on below the 0% duty cycle cutoff. I tried the 1.2v you initially wrote in the code for "voltsForFanStartup" as well as 1v, 0.8v, and 0.6v just to see if it would make a difference and it did not. Do you have any suggestions on what to try next? Put a larger resistor on the PMW output? I currently have 680 ohms on it. I have included the code that I have been working with just in case you see anything that stands out to you.
This code is for a 10K thermister sensor witha neta of 3435, not a GM. The code and wiring for a regular temp sensor is in post number 207 - 209. Also, the thermister uses the 3.3V reference in the arduino, the gm style sensor does not.
Which PWM module are you using?
This code is for a 10K thermister sensor witha neta of 3435, not a GM. The code and wiring for a regular temp sensor is in post number 207 - 209. Also, the thermister uses the 3.3V reference in the arduino, the gm style sensor does not.
Which PWM module are you using?
Also, for the wiring diagram you said in posts 207-209, it is triggered by an ecm? What part of that diagram would I need to modify since this will be a complete standalone fan setup with no ecm input? Also, does the bosch sensor have the same specs as a GM sensor like I am using?
Also, for the wiring diagram you said in posts 207-209, it is triggered by an ecm? What part of that diagram would I need to modify since this will be a complete standalone fan setup with no ecm input? Also, does the bosch sensor have the same specs as a GM sensor like I am using?
Clint
This is how you would wire an Arduino Nano controlling a Mitsubishi/Mazda fan controller using a stock sensor like a GM sensor. Note I haven't tested it.
The code will have to accommodate which Two Wire Temp Sensor and Series resistor used. If you provide the Resistance and temperature table for your sensor, I will write up to code for you and post it up.
I have another project car (not LS) which needed a flatter fan to make clearance for the water to air Intercooler. So I picked up another one of those Volvo PWM brushless fans with the integrated controller for $10.
In order to get the fan going with just and on and off, it needs a PWM generator connected instead of the old fan relay. Instead of grounding the relay coil, the signal from the ECM now grounds the PWM generator. It's that simple. This is how I connected it.
It can also run with any temperature sensor switch that switches to ground. A ground signal turns on the PWM generator to a preset speed setting. This fan will be more powerful that most fans so I have it set for about 60% power. Works fantastic. Set the frequency to 128 and set the duty cycle to any speed you want. Once you set it, it remembers your setting even after the power is shut off.
Flat Volvo brushless PWM controlled fan provided needed clearance for Intercooler.
The PWM generator replaces the relay that used to drive the old fan
This is how you would wire an Arduino Nano controlling a Mitsubishi/Mazda fan controller using a stock sensor like a GM sensor. Note I haven't tested it.
The code will have to accommodate which Two Wire Temp Sensor and Series resistor used. If you provide the Resistance and temperature table for your sensor, I will write up to code for you and post it up.
This should be what you need. Let me know if you need anything else from me. Thanks again
Just "tagging along" with this thread do gather enough info to use a brushless/pwm fan on my 55 Chevy.
The ECM is a Holley Terminator X Max/L33 engine.
I'm considering the Volvo fan. I will likely have to remove most of the shroud as the core support on the car is 16" x 23.5".
I have the tooling to make a custom shroud.
When buying a fan is the used unit a better choice than a Chinese knok off? The car will see long trips, so reliability is a plus.
I see several control schemes to make this work.
Any suggestions for as close as possible to "plug/play" control, for an electronic amateur?
TIA,
Just "tagging along" with this thread do gather enough info to use a brushless/pwm fan on my 55 Chevy.
The ECM is a Holley Terminator X Max/L33 engine.
I'm considering the Volvo fan. I will likely have to remove most of the shroud as the core support on the car is 16" x 23.5".
I have the tooling to make a custom shroud.
When buying a fan is the used unit a better choice than a Chinese knok off? The car will see long trips, so reliability is a plus.
I see several control schemes to make this work.
Any suggestions for as close as possible to "plug/play" control, for an electronic amateur?
TIA,
With the Volvo or BMW or Mercedes brushless fan and a Holley, it's plug and play and reliable. Just direct connect the pwm- to the PWM input of the fan. I believe the Bosch controlled units like the Volvo are very reliable. The last Volvo unit I pulled (in the photo above) was from a flooded car and it still works perfectly.
Thanks for the help!
The Terminator X Max has only 2 fan outputs.
Choices are 12v or gnd. There is no pwm output.
Do I have to use a "box" to add the pwm signal?
If so, would a different fan be more practical?
TIA,
Copied this directly from Holley site
4 Inputs – 12v, Ground, 5v, and Frequency, for things such as additional pressure sensors, or activation triggers for nitrous activation, or a trans brake (Fuel and Oil Pressure Inputs are pre-terminated)
4 Outputs – Ground, PWM - , for accessories such as; fans, boost control, nitrous control, IAC Kit, and More
Thanks for the reply.
DUH!
Once I pulled up the wiring diagram, I saw that all 4 outputs are "PWM", or 12V, including the 2 fans outputs.
[B11 and B12 pins].
When I saw "grnd" I didn't realize that grnd is pwm -.
Hope this is right!
If not, LMK, please!
Last edited by Old Buzzard; 05-12-2022 at 07:41 PM.
This is how you would wire an Arduino Nano controlling a Mitsubishi/Mazda fan controller using a stock sensor like a GM sensor. Note I haven't tested it.
The code will have to accommodate which Two Wire Temp Sensor and Series resistor used. If you provide the Resistance and temperature table for your sensor, I will write up to code for you and post it up.
Hey, were you ever able to write the code for this? Thanks for your help.
Excellent thread. I have the volvo 13 fin fan shown earlier, which I am trying to install in an old truck, it fits the rad perfectly and the fan was just collecting dust. This has been very helpful towards that end. So as I understand, I need simply the arduino nano to feed a PWM signal to the integrated power control unit. I did see that the volvo fan uses a 60 amp fuse in the original vehicle. Also I suspect that the yellow wire is an override for full power, have not yet tested, though.
My question is how can I modify the frequency of the arduino output to 128hz? or to something in the expected 100-300hz range? Did you get a program parameter to run this fan well?
thanks
Excellent thread. I have the volvo 13 fin fan shown earlier, which I am trying to install in an old truck, it fits the rad perfectly and the fan was just collecting dust. This has been very helpful towards that end. So as I understand, I need simply the arduino nano to feed a PWM signal to the integrated power control unit. I did see that the volvo fan uses a 60 amp fuse in the original vehicle. Also I suspect that the yellow wire is an override for full power, have not yet tested, though.
My question is how can I modify the frequency of the arduino output to 128hz? or to something in the expected 100-300hz range? Did you get a program parameter to run this fan well?
thanks
See post 190 and 207 for arduino frequency control. I'd be surprised if there was a full power override wire.
I think I discovered the real shortcut to PWM fans using OEM controllers without an Arduino.
The E-bay controller, which for brushed fans requires a solid state relay.
Anybody care to guess the major component of many OEM controllers? A solid state relay.
Most controllers only need a high amp feed, a ground,and a PWM signal.
Some might run the ground from the fan thru the controller, then to the battery, which is a clean way to do it.
The Mitsu/Mazda6, Ford/GM,and many other controllers do this.
The Dodge/Jeep controller does not, just routes and controls incoming live power.
This all may seem obvious, but it was not made so during the discussion how it directly replaces the Arduino discussed in this thread
12-21-2021, 08:51 PM
