DIY Gizmo - Honda ATV Programmable Fan Control Unit

[retro 4,048]

Hahaha! It's said that poor boys got poor ways eh! Well it's true, those red thermy lead insulators are lengths of plastic insulation that I stripped off of some solid copper wire. That wire come from the same spool of solid copper that I made my headers with for the power supply. 🙂

 

[Goober 1,555]

I bet just a light coat of Flex Seal would work lol

[Goober 1,555]

Hey @retro how deeply do you immerse the sensors for testing? On my second round—i let the oil level come up just to the top of the threads where they contact the hex flats. When i tested them earlier i let the oil come up to halfway on the threads and the sensors lagged a bit. Thot it wasn’t representative of actual heating conditions.

here’s table and graph showing the thermistor and two sets of data—rebuilt sensors more deeply immersed and an averaged set of data from the same sensors—more shallowly immersed.
Makes you wonder if the thermies were tested bare.

 

Edited April 18, 2021 by Goober

[retro 4,048]

I noticed that the bare thermie for my multimeter lags behind oil temp changes too, unless the oil is constantly being stirred, despite that bare thermie being submerged to the same depth and placed within 1/4" of the end of the oil temp sensor brass that is being measured. So constant stirring of the oil is required.... when I take a photo while calibrating a sensor I gotta snap a pic within 3 seconds after I stop stirring the oil, else the numbers for both the sensor & the measuring thermie begin to drop at different rates. The mass of the brass sensor wicks the oil temp down rapidly the moment stirring stops.

 

I submerge the brass up to the bottom of the flats so that the majority of the mass is in oil. I imagine the brass would cool the oil immediately surrounding it less if the entire sensor were dipped in oil though. Maybe I'll try that the next time I calibrate one of them....

 

You got some good data there... your sensor R/T curves are pretty consistent with published specs.

[Goober 1,555]

I noticed the Honda oil temp sensors also place the thermistor 1/4” from the sensor tip. But instead of shortening botj connecting wires, they shortened one side of the connecting wire as in the pic below, so the longer wire loops down to the bottom of the sensor brass and back up 1/4 inch. Extra wire just adds complexity to the installation 

I can’t remember how I installed mine—my pic suggests i put the thermistor in as close to bottom as I could. Too much Heat shrink on those wires 

[jeepwm69 7,600]

Looks like I need to test a couple of OEM oil temp sensors.  Have some spare engines in the floor of the shop, and I'm concerned that the 07 Foreman I've rebuilt has yet to have the fan come on, so plan is to test the two sensors out of the shop floor engines, and if they are good, swap one into the 07 engine, and then test the one i removed.

[retro 4,048]

You can test the fan motor and the FCU oil temp light functions by grounding the Light Blue wire that plugs into the oil temp sensor. The spec for an oil temp sensor is: 9.5k ohms to 10.5k ohms at 68 degrees F. A room temperature sensor resistance measurement is close enough to 68 degrees to be a trustworthy test.

[jeepwm69 7,600]

23 minutes ago, retro said:

You can test the fan motor and the FCU oil temp light functions by grounding the Light Blue wire that plugs into the oil temp sensor. The spec for an oil temp sensor is: 9.5k ohms to 10.5k ohms at 68 degrees F. A room temperature sensor resistance measurement is close enough to 68 degrees to be a trustworthy test.

 

Fan and light come on when I ground it, but the fan didn't come on at all the other day and it was HOT when we were riding.  Made me worry a bit that the sensor might not be good.

[retro 4,048]

Yup probably a bad oil temp sensor then.

[Goober 1,555]

I’m trying to get back to this project!!!🤥😂

[Goober 1,555]

On 6/28/2021 at 12:18 PM, jeepwm69 said:

 

Fan and light come on when I ground it, but the fan didn't come on at all the other day and it was HOT when we were riding.  Made me worry a bit that the sensor might not be good.

Hey Jeep

you ever test the temp sensor?

i tried to strip down one of those newer sensors; boy that ceramic is tough to remove!

looks like the new thermie can be grounded through the hole in the body tip.

 

[jeepwm69 7,600]

2 hours ago, Goober said:

Hey Jeep

you ever test the temp sensor?

i tried to strip down one of those newer sensors; boy that ceramic is tough to remove!

looks like the new thermie can be grounded through the hole in the body tip.

 

 I have not.  I haven’t come across a bad one yet on a 500. The one that I swapped off of that 07 into my 09 that I was worried about works fine, so that 07 must’ve just been an exceptionally cool running machine lol

[Fishfiles 20,145]

I mounted the GIZMOS display , changed the temp sensor, filled the engine back up with fresh oil , mounted the relays , just have to do the wiring and then I keep getting side tracked ------ 

 

[Goober 1,555]

Finally got back to this project! I needed to replace two sensors right away so i can ride those TRX350A before the trails close. I also have four work TRX350A  that might overheat. So I built four without the GIZMO leads—I already built two with GIZMO leads which i can switch out when i get the controllers completed.

I need to bench test the sensors and record the values —i guess today would be a good day considering i plan to fry some catfish LOLs. Ambient temp 70F, 82F; at rest sensors:

Sensor ID, KOhms@70F, 82F

9E26, 9.67, 7.65
12E21, 9.52, 7.47

4E23, 9.43, 7.29
7E19, 9.46, 7.39

9E03, 9.56, 7.57

9E17, 9.69, 7.61

Very happy with these 

 

 

Edited August 22, 2021 by Goober

[Goober 1,555]

They all passed by getting below 1KOhm by the time the oil was 200F and below 231Ohm by 300F. 
then I found I hadn’t mixed enough hardener in the epoxy on the first sensor i did. It puffed up like a muffin. I had problems with the applicator. Well at least I could peel away all the unset epoxy!

[Fishfiles 20,145]

@Goober , what do consider the normal operating temp range and  the high temp alarm on temp ?  And what are You setting your Gizmos up at ?

[Goober 1,555]

Hey Fish! I haven’t completed my Gizmos yet. Gotta finish the power supplies. Since it’s taking me so long, thought I would simply install rebuilt sensors.

 

the newer model service manuals may not list the alarm temps along with sensor thermistor values, but the 86-89 TRX350 does. 

 

The fan is designed to come on at 100C (212F); thermistor resistance value 1.05 to 0.95 KOhm. This is supposed to kick on the cooling fan but the oil light stays out. 
 

At the overheat temp of ~330F, the thermistor is supposed to read less than 230 Ohms—essentially no resistance. The fan should already be running but now the oil light should come on. As retro said, that’s too late.
 

@retro initially set up the Creamsicle fan to come on at 100C (212F) and go off at 85C (185F). the OEM Fan Control Unit normally responds to the sensor resistance value in the circuit, but can’t display operating temp. Beauty of the gizmo it displays all that and allows custom settings 

 

Edited August 23, 2021 by Goober

[_Wilson_™ 6,564]

the newer model service manuals may not list the alarm temps along with sensor thermistor values, but the 86-89 TRX350 does. 

 
The fan is designed to come on at 100C (212F); thermistor resistance value 1.05 to 0.95 KOhm. This is supposed to kick on the cooling fan but the oil light stays out. 
 
At the overheat temp of ~330F, the thermistor is supposed to read less than 230 Ohms—essentially no resistance. The fan should already be running but now the oil light should come on.

 

 

I'm not disputing the temp range on the trx350, mines an 86.... only had the fan come once, but the high oil temp light also came on at the same time.... and both went off at the same time,  and i believe 330f is correct.  never had a wire hack... so I'm baffled on why mine acts like this, but,  then again I'm  not in the habit of over heating the plant. 

[Goober 1,555]

Ohms check your sensor leads. Should be 9.5 to 10 KOhms. my sensors were in the 13 KOhm range. 

 

When resistance is higher, your engine temp will be hotter than you want it to be when the fan kicks on. 

I don’t recall getting an oil light with the fan on—I thought it was odd because i only heard a new sound and had to hold my hand over the louvers to feel the fan running. But the light comes on when i ground it out. Now that i have a decent sensor i can try it again.

[_Wilson_™ 6,564]

thanks goober . oil temp would kind of flicker off and on, but the fan was on steady, but when the fan shut off, so did the light, but i always kept that bike immaculate.... (no hacks or rodent damage) would it be the rare 86 DC current model be a reason the oil temp light and fan came on at the same time ? because i know it wasn't a weak battery issue ... but like you, i didn't hear the fan so much as i felt the extra heat wave all the sudden, then i kept a check on both the oil light and fan, till they went off, but as a standard the oil light would come on for a short time when the ignition was cycled, but the over heat didnt damage the engine at all.... told my friend Keith ... i wasn't going to run it any longer till it cooled off. so i headed to the shop, and turned a shop fan on it, and went on about my days work.

[Goober 1,555]

Be more successful in testing your oil sensor and more confident in judging whether it’s good or bad.
 

the Honda manual could improve the method recommended for testing oil temperature sensors. I’ve had poor luck using an open flame to heat the pan of oil as described to perform the test. 

 

I’ve had much better performance using an electric stovetop and peanut oil. I used a two quart stock pot—good height to width ratio for heating oil. I immerse the sensor in the pan of oil just to the base. Set up my fluke meter and thermometer. Once i got the pan and sensor temp stabilized (at or below 75F) then i kick on the heat and start logging values. 
 

My tablecraft electric stovetop gets the oil from 65-70F up to 338F in 15 minutes. So every 20 seconds, i was logging a 5F change in temp and the corresponding resistance value. Not only do I know when the sensor reaches cooling system set points, i see how responsive the sensors are to the rise in temperature.
 

This method helps to keep consistency in my readings where before i was doing it outside with the wind, not very well organized, heating the oil too fast. Oh and the house smells like peanuts not a garage. 
 

 

[Goober 1,555]

These 86-87 TRX350A and 85-86 TRX250 all test good, just gotta put connectors on them.

 

[Goober 1,555]

Has it really been over a year since we talked oil temp sensors?  I have not made much progress on this job—six of the original seven are still on my bench. Sent one away to another owner but have heard nothing back about performance. 

[retro 4,048]

Yeah it's incredible! I feel pretty bad about it too. I bought and built a new Prusa 3D printer last year for making FCU housings, buttons and other gizmos -- I haven't even powered that Prusa up yet, it's been sitting inside the temp-controlled cabinet that I got for it for more than a year. The control boards, cabling, fans and voltage regulators for the 3D printer project are still in the shipping boxes too. It's been incredible man.... got no hope to get back on this stuff until mid-summer at the earliest. Hopefully you'll find more time than I can.