Designed my own electric fan controller
#1
Designed my own electric fan controller
Just installed it tonight and it works great. The cheapie I bought failed in less than a year -- probably from water and mud, lol.
Anyway, it's on about a 2" by 2" board. The design goals were:
1. must be built from junkbox parts (spend NO money but use parts salvaged from old junk stereos and whatever I have laying around)
2. must use the existing coolant temp sensor that feeds the gauge and PCM
3. must be able to set the spread between turn on and turn off (set temperatures wide apart to minimize cycling)
4. must mount INSIDE the vehicle to be out of the weather
I was in a hurry to get this done, so I haven't taken pics, and for now it is just tapped in and wrapped with tape and wedged into the general mess behind my radio in the center of the dash -- but it works great.
The fan now comes on at a "true" coolant temp of 210 degrees F (as measured by my scan tool), and cuts off at 195. I have an input to pick up the AC compressor enable as well, but I didn't wire it yet. The wide setting between on and off works nicely. It takes a while to come on, then runs awhile, then rests awhile and doesn't short-cycle as much as the other one did.
On the dash is a 3 position small rocker. Center is off, position 1 (press to the left) is fan on manual control, position 2 (press to right) is automatic control.
The relay to control the fan is up near the fan, and this circuit just turns on that relay.
I tapped the coolant temperature sensor wire in the harness bundle near the emergency brake.
Not too many parts in it either. An MC1458 dual op-amp (only using one section), an opto isolator (4N35), a darlington PNP transistor (TIP125), a LM7809 voltage regulator, and a silicon diode (like a 1N4001) are the entire active parts. Two 10 turn trimpots, and a handful of resistors and capacitors make up the rest.
If anyone is into electronics and wants to reproduce it, I'll draw it up nicer, scan it, and post it.
At some point I'll take the breadboarded one I made out and photograph it, though it's not real impressive, lol. It just works well.
Anyway, it's on about a 2" by 2" board. The design goals were:
1. must be built from junkbox parts (spend NO money but use parts salvaged from old junk stereos and whatever I have laying around)
2. must use the existing coolant temp sensor that feeds the gauge and PCM
3. must be able to set the spread between turn on and turn off (set temperatures wide apart to minimize cycling)
4. must mount INSIDE the vehicle to be out of the weather
I was in a hurry to get this done, so I haven't taken pics, and for now it is just tapped in and wrapped with tape and wedged into the general mess behind my radio in the center of the dash -- but it works great.
The fan now comes on at a "true" coolant temp of 210 degrees F (as measured by my scan tool), and cuts off at 195. I have an input to pick up the AC compressor enable as well, but I didn't wire it yet. The wide setting between on and off works nicely. It takes a while to come on, then runs awhile, then rests awhile and doesn't short-cycle as much as the other one did.
On the dash is a 3 position small rocker. Center is off, position 1 (press to the left) is fan on manual control, position 2 (press to right) is automatic control.
The relay to control the fan is up near the fan, and this circuit just turns on that relay.
I tapped the coolant temperature sensor wire in the harness bundle near the emergency brake.
Not too many parts in it either. An MC1458 dual op-amp (only using one section), an opto isolator (4N35), a darlington PNP transistor (TIP125), a LM7809 voltage regulator, and a silicon diode (like a 1N4001) are the entire active parts. Two 10 turn trimpots, and a handful of resistors and capacitors make up the rest.
If anyone is into electronics and wants to reproduce it, I'll draw it up nicer, scan it, and post it.
At some point I'll take the breadboarded one I made out and photograph it, though it's not real impressive, lol. It just works well.
#2
sounds just like the one i built a little while ago to use for my cooling fans for my amps! im an idiot though and left the thing in arizona back in june, along with building instructions, and a parts list, lol, so it'll be real nice to see what you've come up with! id like to build a few more!
#3
#4
Maybe $20 to $25 at most, I should think. I don't know if they have the opto isolator, but the exact one is not required. I just used it as a "level" shifter (from the 9 volt supply of the circuit, to the 12 volt supply of the transistor) and to keep potential relay coil switching transients out of the IC if the contacts on the switch don't break cleanly
I just drew up a sketch, I'll scan it and post it. It's not so neat but it's better than my assembly sketch/notes, lol. I'll do a better one later if required.
I'll post it in a little bit.
I just drew up a sketch, I'll scan it and post it. It's not so neat but it's better than my assembly sketch/notes, lol. I'll do a better one later if required.
I'll post it in a little bit.
#5
That's sounds great, John. Mine runs off a plain old fixed radiator switch with a 20F deadband - how boring.
Here's a feature I'd like to see: a speed controlled cut off for the fan's AC function. You could have the fan automatically turn off above whatever set speed you choose, 35 MPH for example. I hate the idea of running the fan at highway speeds for AC when it isn't needed at all. Sure, you could have a manual switch to turn it off but it would be that much cooler if it were automatic.
FYI, your truck uses two separate sensors for the gauge and the PCM. The more critical PCM sensor (ECT Sensor) has a dedicated signal return but the gauge sensor (ECT Sender) uses chassis ground. I guess you must be using the ECT sensor for the PCM if you're able to read it with your scan tool.
Here's a feature I'd like to see: a speed controlled cut off for the fan's AC function. You could have the fan automatically turn off above whatever set speed you choose, 35 MPH for example. I hate the idea of running the fan at highway speeds for AC when it isn't needed at all. Sure, you could have a manual switch to turn it off but it would be that much cooler if it were automatic.
Originally Posted by n3elz
2. must use the existing coolant temp sensor that feeds the gauge and PCM
#6
That's a good idea, Bob, and a few of us have mused on that. It should be fairly simple using the VSS, a comparator to square the signal, and an LM567 (I think that's the number -- it's from memory) "tone detector". I believe there's an application note for a circuit where it tells you you're above a certain frequency. I'll look that up and think about it -- I like the idea.
#7
Okay, here's the circuit for the basic controller.
You may need to ask questions -- I haven't put many notes on it. Those of you with experience this is all you need.
Good starting point is to adjust the "threshold" pot to put 5.25 volts on pin 2 of the op-amp, and adjust "hysteresis" to half scale. This will put you in the right ball park.
DECREASING the threshold voltage, RAISES the temperature where the fan cuts on.
DECREASING the resistance of the hysteresis pot, WIDENS the temperature gap between on and off. values below 20K or so are probably useless (too wide). Fully ed up you'll only get 3 to 5 degrees or so. 50K should give you around 5 to 10 degrees. It's easiest to tune with a scan tool handy to watch the actual temperature.
EDIT: IMPORTANT NOTE -- The transistor is drawn as an NPN and it is not -- it's a PNP transistor.
You may need to ask questions -- I haven't put many notes on it. Those of you with experience this is all you need.
Good starting point is to adjust the "threshold" pot to put 5.25 volts on pin 2 of the op-amp, and adjust "hysteresis" to half scale. This will put you in the right ball park.
DECREASING the threshold voltage, RAISES the temperature where the fan cuts on.
DECREASING the resistance of the hysteresis pot, WIDENS the temperature gap between on and off. values below 20K or so are probably useless (too wide). Fully ed up you'll only get 3 to 5 degrees or so. 50K should give you around 5 to 10 degrees. It's easiest to tune with a scan tool handy to watch the actual temperature.
EDIT: IMPORTANT NOTE -- The transistor is drawn as an NPN and it is not -- it's a PNP transistor.
Last edited by n3elz; 11-24-2006 at 03:27 PM.
#8
Hmmm...maybe not, Bob, lol. I didn't even look very close at the diagram -- I used the one that goes to the gauge so I didn't reach my design goal, lol.
Yes, using ground is a problem in that the ground loops introduce error (and you can see it when the fan turns on). You have to adjust the hysteresis up a bit to compensate -- but it still works very well and didn't require mounting another sensor
So, I failed to reach that one goal, dang it. You had to spoil my fun! -- however, I'm happy with how it works so there you go. I'll have to go back and look at that diagram.
I ASSUMED they used the same sensor WITHOUT LOOKING -- poor engineering practice but it was only for me and I'm in a hurry to have it in for Centralia so I don't have to monitor my cooling system or leave the fan on all the time. In fact, some of those goals were all about "hurry up and get it done", lol.
Yes, using ground is a problem in that the ground loops introduce error (and you can see it when the fan turns on). You have to adjust the hysteresis up a bit to compensate -- but it still works very well and didn't require mounting another sensor
So, I failed to reach that one goal, dang it. You had to spoil my fun! -- however, I'm happy with how it works so there you go. I'll have to go back and look at that diagram.
I ASSUMED they used the same sensor WITHOUT LOOKING -- poor engineering practice but it was only for me and I'm in a hurry to have it in for Centralia so I don't have to monitor my cooling system or leave the fan on all the time. In fact, some of those goals were all about "hurry up and get it done", lol.
Originally Posted by rwenzing
That's sounds great, John. Mine runs off a plain old fixed radiator switch with a 20F deadband - how boring.
Here's a feature I'd like to see: a speed controlled cut off for the fan's AC function. You could have the fan automatically turn off above whatever set speed you choose, 35 MPH for example. I hate the idea of running the fan at highway speeds for AC when it isn't needed at all. Sure, you could have a manual switch to turn it off but it would be that much cooler if it were automatic.
FYI, your truck uses two separate sensors for the gauge and the PCM. The more critical PCM sensor (ECT Sensor) has a dedicated signal return but the gauge sensor (ECT Sender) uses chassis ground. I guess you must be using the ECT sensor for the PCM if you're able to read it with your scan tool.
Here's a feature I'd like to see: a speed controlled cut off for the fan's AC function. You could have the fan automatically turn off above whatever set speed you choose, 35 MPH for example. I hate the idea of running the fan at highway speeds for AC when it isn't needed at all. Sure, you could have a manual switch to turn it off but it would be that much cooler if it were automatic.
FYI, your truck uses two separate sensors for the gauge and the PCM. The more critical PCM sensor (ECT Sensor) has a dedicated signal return but the gauge sensor (ECT Sender) uses chassis ground. I guess you must be using the ECT sensor for the PCM if you're able to read it with your scan tool.
#10
#13
#15
Damn, I have a ways to go in my electronics course if I am supposed to read all those schematics by december lol :)
I can understand most of it, except why you need so many resitors to achieve all this throughout the circuit...oh well, ill learn it in a few weks prolly :)
I can understand most of it, except why you need so many resitors to achieve all this throughout the circuit...oh well, ill learn it in a few weks prolly :)
Last edited by hougy; 10-16-2005 at 11:11 PM.
#16
#17
I was just referring to my schematic when I noticed something that most probably wouldn't care about anyway -- but if you're building this thing it's important. The transistor is drawn as an NPN transistor, and it's a PNP.
It continues to work marvelously -- unless I bypass it to "off" and leave my truck run, which I've done TWICE now. Once at a meet (really boiled some coolant) and once when meeting Mike @ SJ Emergency Light to grab his photo's of a meet.
Also, I added a 0.3 ohm 50 watt resistor (off the efan of a junked Pontiac Fiero, believe it or not) to put my fan into a "normal" ranger. It was always seriously oversized and over powered for the application, and folks made fun of it.
Well, it's pretty quiet now, and drawing about 50 to 60% of the amps it did draw -- pulling about 11 or 12 now -- over 20 before, and it needed a 30 amp fuse to not blow it when it started up, lol.
I idled it for long periods with the A/C on, and it remained at temp. You all may not be able to make fun of my fan anymore!!!
I'll probably add a bypass relay to run it at full speed when needed.
It continues to work marvelously -- unless I bypass it to "off" and leave my truck run, which I've done TWICE now. Once at a meet (really boiled some coolant) and once when meeting Mike @ SJ Emergency Light to grab his photo's of a meet.
Also, I added a 0.3 ohm 50 watt resistor (off the efan of a junked Pontiac Fiero, believe it or not) to put my fan into a "normal" ranger. It was always seriously oversized and over powered for the application, and folks made fun of it.
Well, it's pretty quiet now, and drawing about 50 to 60% of the amps it did draw -- pulling about 11 or 12 now -- over 20 before, and it needed a 30 amp fuse to not blow it when it started up, lol.
I idled it for long periods with the A/C on, and it remained at temp. You all may not be able to make fun of my fan anymore!!!
I'll probably add a bypass relay to run it at full speed when needed.
#18
#19
Well, THAT schematic has not changed so much, really. However, I could show the resistor, bypass relay (not yet installed anyway) and switch for it.
One thing I'm thinking about doing is never allowing the fan out of AUTO mode, but instead raising the temperature where it kicks in. Under normal circumstances, that would cause the fan to kick off for water crossings or what not, but if I forgot about it, it would still come on again before the vehicle overheated. My absent-mindedness twice over the switch makes this seem like an attractive modification.
Edit: One other thing: that second threshold could be used to put the fan in "HIGH" (activating the bypass relay) so if for any reason the slow speed was inadequate, the fan speed would be increased to try to bring down the temperature.
One thing I'm thinking about doing is never allowing the fan out of AUTO mode, but instead raising the temperature where it kicks in. Under normal circumstances, that would cause the fan to kick off for water crossings or what not, but if I forgot about it, it would still come on again before the vehicle overheated. My absent-mindedness twice over the switch makes this seem like an attractive modification.
Edit: One other thing: that second threshold could be used to put the fan in "HIGH" (activating the bypass relay) so if for any reason the slow speed was inadequate, the fan speed would be increased to try to bring down the temperature.
#20
#22
#23
#24
Ok well....I don't understand a word of that lol
Using the Jusnes e-fan and controller....I'm going to assume one position to cut off the power or ground.
Another to bypass the temp sensor....and the last one to leave it like it is now, which is controlled by the temp sensor and AC control (purple wire behind the HVAC)
Using the Jusnes e-fan and controller....I'm going to assume one position to cut off the power or ground.
Another to bypass the temp sensor....and the last one to leave it like it is now, which is controlled by the temp sensor and AC control (purple wire behind the HVAC)
Last edited by buzzair; 11-24-2006 at 03:44 PM.
#25
Here's how it could work:
If the Jusnes controller uses a relay that is grounded all the time on one side of the coil, and it applies power to the other side of the coil to activate it, then you would wire the switch with the center wire to the fan relay coil +, an end wire to +12 (switched with the ignition key), and the other end wire to the Jusnes controller output to the relay.
However, I don't know if Bob's controller has the relay "integrated" -- many do. So the above method wouldn't work.
You could use a second relay, controlled by the Jusnes controller, and use the switch as I've described.
Some controllers have an outside "override" position for the HVAC that can be used, but you'd need a diode. It would be helpful to see the Jusnes controller's documentation and then I could draw something up for you.
If the Jusnes controller uses a relay that is grounded all the time on one side of the coil, and it applies power to the other side of the coil to activate it, then you would wire the switch with the center wire to the fan relay coil +, an end wire to +12 (switched with the ignition key), and the other end wire to the Jusnes controller output to the relay.
However, I don't know if Bob's controller has the relay "integrated" -- many do. So the above method wouldn't work.
You could use a second relay, controlled by the Jusnes controller, and use the switch as I've described.
Some controllers have an outside "override" position for the HVAC that can be used, but you'd need a diode. It would be helpful to see the Jusnes controller's documentation and then I could draw something up for you.