Honda CX 500 Forum banner
1 - 20 of 37 Posts

·
Registered
Joined
·
798 Posts
Discussion Starter · #1 ·
I don't remember where, but I remember someone mentioning a different thermostat to control an electric fan asaell. Have I gone off the deep end or is this true?
 

·
Registered
Joined
·
2,601 Posts
I don't remember where, but I remember someone mentioning a different thermostat to control an electric fan asaell. Have I gone off the deep end or is this true?


We've got three threads going on about that at present, one is named "Is it getting hot in here" or something like that and I can't remember the other ones.



I've got one designed that incorporates several features into it that makes it ideal for controlling an electric fan.



1) It's fully variable speed using PWM control (small size, virtually no heat generated thus a small enclosure)



2) It's user adjustable as far as the turn on point



3) It gives the fan a 1 second "kick start" of full current to get it up to speed before the variable kicks in, most electric fans won't start from a dead stop with just a little voltage applied



4) It incorporates an optional warning light output that signals if the fan stops working or if the bike gets unusually hot



5) It allows you to keep the current temp gauge and sensor in place



6) It monitors the temp of the water LEAVING the radiator before it's fed back into the engine. This allows better control because, if you're cruising, you may be getting hot going into the radiator but the existing wind flowing through the radiator may be enough by itself to cool it.



I don't have the bike in hand (friend is helping to finish the restoration) but if I'm lucky I hope to pay it a visit on Sunday to come up with an ideal sensor placement that effectively makes the installation easy and hopefully invisible.
 

·
Registered
Joined
·
798 Posts
Discussion Starter · #3 ·
Sounds cool. I'd noticed a couple, but none with defininitive answers.

The trailtech vapor has a cool tube like piece for the temp sensor that could go on the rubber hose from the radiator to the chrome pipe if you made it bent.
 

·
Registered
Joined
·
724 Posts
I don't remember where, but I remember someone mentioning a different thermostat to control an electric fan asaell. Have I gone off the deep end or is this true?


you are probably thinking of the post i put up. i got a coolant fan switch for an 81 honda civic. it screwed into the temperture probe at the back of the thermostat housing. hot wired fan through the switch but haven't gotten it hot enough to see if its works yet.
 

·
Registered
Joined
·
2,601 Posts
Sounds cool. I'd noticed a couple, but none with defininitive answers.

The trailtech vapor has a cool tube like piece for the temp sensor that could go on the rubber hose from the radiator to the chrome pipe if you made it bent.




What's a trailtech vapor?



Yea, it was fairly easy designing all the electronics, the big decisions had to do with how and where to monitor the temperature and still leave the gauge circuitry working. Since I've found the stock temp sensor to be far from accurate I gave up on sampling that voltage as a reference, besides that I hadn't thought about the fact that ideally you should be controlling the temperature of the water being fed back into the engine anyway.



Yea, it's overkill but I found a series of fan controller ICs that already had a ton of functionality built in. Thought about making both the kick in temp point and the warning temp point adjustable but decided to leave it just as the kick in temp being adjustable to allow for people that live in colder climates. Haven't quite decided where to set the overheat trip point. The stock thermostat is designed to finally reach fully open somewhere between 93*C - 97*C (199*F - 205*F) so I'll probably set it at 99*C (210*F) which is just before the boiling point of water and most certainly hotter than I'd want my engine running. This gives you an operating range of 180 - 210.



I will entertain other people's opinions though, where would you set it?
 

·
Registered
Joined
·
798 Posts
Discussion Starter · #7 ·
By overheat you mean when the light kicks on? Id say 99 is good. Still give you a little extra headroom if your running coolant so you know there is a problem before its too serious, but also wont be sending of false alarms.



Im not even close to an expert though. Im not used to working on vehicles with radiators.
 

·
Registered
Joined
·
2,601 Posts
By overheat you mean when the light kicks on? Id say 99 is good. Still give you a little extra headroom if your running coolant so you know there is a problem before its too serious, but also wont be sending of false alarms.



Im not even close to an expert though. Im not used to working on vehicles with radiators.


Even though I realize that the antifreeze/coolant raises the boiling point to around 240 it just seems to me that anything reaching the 210* range would be considered as too hot as, according to Honda specs, the thermostat is supposed to be fully open once you reach the 199* - 205* range. Of course that temperature is based upon where the thermostat is located which would be before the radiator.



Hitting 210* coming out of the radiator would tend to indicate a failed fan or obstructed radiator core, of course the circuit also sets the fault light if it senses the fan isn't rotating at a time when the circuit is commanding it to do so.



It wouldn't be too hard to add another potentiometer to make that temperature setting adjustable as well but this would present two disadvantages:



1) Adjusting one setting would affect the other; you'd be going back and forth to get them both right and we want this simple

2) Well made, durable pots add a few more dollars to the cost of the unit and I'd like this to be as affordable as possible.



Ease of installation, virtually invisible mounting, 100% weatherproof and the capability of being able to work with virtually ANY 12V electric fan are priorities as are being pretty much plug and play. Even the turn on speed could be a preset but I still think it's necessary should you want to use it on a different bike, mount the sensor in a different place or live in a cold climate where it's desirable to maintain the engine at a slightly higher operating temperature.



No answer from my PC board guy yet since it slightly breaks their rules but I'm hoping to design a run of one big board that can be cut apart to make several circuits. So far it woould cut up into PC blanks used to produce:



1) The fan controller

2) The 7V regulator replacement as either version

3) An exact reproduction CDI

4) That dot/bar display circuit that can be used for so many things.



and a few other circuits that aren't CX/GL related but of interest to others. For instance I designed an FM radio preamplifier setup that has an incredible amount of gain at an extremely low noise. The preamp box mounts on the antenna mast near the antenna and in the house is the small power supply box that has remote adjustments for gain, input antenna matching and center tuning frequency. Using a three element cubical quad antenna of my own design I'm able to pull in a station from Tulsa that's even difficult to receive if you're 30 miles closer to it.



The dot/bar display can be used for almost anything you want to monitor and if I can find a small enough rotary switch it could be multiple input. If you mounted the right sensors it could be switchable between battery voltage, oil pressure &/or temperature, possibly an ammmeter and if I ever get time to work on my fuel gauge idea it could show that as well. The really small analog rotary switches used to be plentiful but have all but disappeared in favor of digital selection. Anything I build that is intended to go on the bike also has to be pretty darn vibration and weatherproof so I haven't decided how far to go or in what direction on that one. I have found some that will fit the bill but they aren't cheap.



The whole idea of this is to be able to sell the circuit boards by themselves or as full kits including the parts for self-builders, or simply as fully built and tested devices.



Anyone have any ideas of some other simple handy circuits we might want to include? My design time is going to be a bit limited and laying out a PC board, even with the full suite of specialized software, isn't as easy as it sounds.
 

·
Registered
Joined
·
2,601 Posts
The part I was referring to is this

http://trailtech.net/7500-3050.html

but ideally a bend in it to make it easier to mount to the bottom of the rad.


Pretty darn expensive for something you could make out of a brass pipe fitting you just drilled a small hole in. I wish I had the bike here and should soon enough if I get my final parts in and then get all the painting done. Wherever it ends up being mounted it's going to be tight, that's why I'm hoping there's an easy enough place to glue the sensor down with some thermally conductive epoxy covered with some sort of thin layer of insulative material to minimize the effects of outside air interfering with the temperature to be read. Kind of defeats the purpose if I've got it glued to the radiator but the air around it is 20 below as some of our riders insist on riding in.



What with all the heated grips I'm surprised I haven't heard of anyone adapting the heated seat part of a modern car into the mix but they chose not to give us much of a stator to start with, there's very little extra current to work with especially on the CDI bikes.



While it could be done here's another idea that might work but would probably be extremely cost prohgibitive. Devise some sort of "pancake alternator" that fit on the camshaft stub that's left after the manual fan is removed. If you used it just to run the headlight there's a good 5 amps you could free up from the main circuit.
 

·
Registered
Joined
·
8,423 Posts
What with all the heated grips I'm surprised I haven't heard of anyone adapting the heated seat part of a modern car into the mix but they chose not to give us much of a stator to start with, there's very little extra current to work with especially on the CDI bikes.


You should know better




As we speak I'm recovering a spare seat and putting some of my heat wire



http://www.pdsrecording.site90.com/cxgl500/heatedgrips.htm



on it so I can have a warm bum in Winter




I've taken my heated grips off as I've made some heated inner gloves now
 

·
Registered
Joined
·
2,601 Posts
I should have figured. :)



So long as I'm making a tiny PC board for this fan controller I'll probably stuff a tiny buffer amp in there that will allow an output pin which can be fed to a temperature gauge made out of one of those dot/bar ICs should a person decide to add one later.



Also remember now's the time if you've got an idea for a useful circuit I might find room to stuff into the final template before I send it off to the circuit board etchers.



Going to be a long weekend of design + a planned visit to see my bike for the first time in ages.
 

·
Registered
Joined
·
8,423 Posts
So long as I'm making a tiny PC board for this fan controller I'll probably stuff a tiny buffer amp in there that will allow an output pin which can be fed to a temperature gauge made out of one of those dot/bar ICs should a person decide to add one later.


That would be excellent.Then if a temp gauge goes there's no need to fiddle around replacing it even if you can find one
 

·
Registered
Joined
·
269 Posts
I've got one designed that incorporates several features into it that makes it ideal for controlling an electric fan.


I'd buy that for a dollar, do you think you will build and sell them?
 

·
Registered
Joined
·
269 Posts
I would set the fan to turn on about 10deg colder than the thermostat is at full open.

That lets your radiator do all the work it can before the fan comes one to help.

Sense it is a PWM controller, you can have more than one speed.

So there could be a "low" speed, say 60% duty cycle, that comes on at 10deg down, and a "hi" speed at 100% duty if it gets to 5 deg down. your radiator should give about a 20 deg drop from inlet to outlet.



I would set the warning light to 215deg. that puts your engine at about 235deg if your radiator is not plugged, and less if you are having a problem. (blocked air flow, plugged cores, no fan)



I would request an output for another light that is on when the fan is on, but that is easy to wire in down stream.

and a grounding output that could be run through a switch to ground for manual "on"



Use this in conjunction with a aftermarket 7" fan and it may never get to "hi" mode!

FYI, my stock 650 fan pulls 4 amps. most 7" fans I have found pull about 6 amps.



I hope to upgrade to a better 7" fan soon and would love to use your controller with it.
 

·
Registered
Joined
·
2,601 Posts
I would set the fan to turn on about 10deg colder than the thermostat is at full open.



I would set the warning light to 215deg. that puts your engine at about 235deg if your radiator is not plugged, and less if you are having a problem. (blocked air flow, plugged cores, no fan)



I would request an output for another light that is on when the fan is on, but that is easy to wire in down stream.

and a grounding output that could be run through a switch to ground for manual "on"


Chances are both setpoints will be adjustable, there's no telling how one might need to set it up if they live way up North.



The output that will be meant to hook to an optional dot/bar display temperature gauge could just as easily be run through a zener diode to an LED to any GND if you absolutely want something to serve as a "fan is running" indicator.



Hard day at work, I'm going to relax a bit then get back on it.



As far as setpoints remember this is going to be monitoring the temperature of the coolant as it LEAVES the radiator heading back into the engine. The main reason for this is no matter how hot the water is going into the radiator, if you're getting plenty of cooling from the wind such that the returning water is already being cooled enough there's no reason for the fan to kick in. It helps to compensate for ambient temperatures but can also sense the difference in radiator efficiency.
 

·
Registered
Joined
·
8,423 Posts
Use this in conjunction with a aftermarket 7" fan and it may never get to "hi" mode!

FYI, my stock 650 fan pulls 4 amps. most 7" fans I have found pull about 6 amps.



I hope to upgrade to a better 7" fan soon and would love to use your controller with it.


My Ducati 748/916/996 fans only draw 2 amps and are very efficient.



HTH
 

·
Registered
Joined
·
2,601 Posts
Use this in conjunction with a aftermarket 7" fan and it may never get to "hi" mode!

FYI, my stock 650 fan pulls 4 amps. most 7" fans I have found pull about 6 amps.


It's continuously variable, no steps to it. It goes from 0% to 100% duty cycle in a linear factor related to the temperature sensed. The MOSFET used for the fan driver is good for 100A but you'd need a small heat sink to handle that amount of current, as it stands it only has a typical on resistance of 0.0021 ohms so even if you were running a 10 amp fan at 100% you'd only be losing 0.021V from the 12V driving it.
 

·
Registered
Joined
·
2,601 Posts
Of course now enters yet another question. Exactly how much does the radiator cool the water?



If I'm running along at 180*F (82*C) we know that's when the thermostat just starts to open and you probably wouldn't need any fan unless you were just sitting still.



If I'm hitting 210*F (99*C) this is when the thermostat has reached the fully open position and the engine is demanding maximum cooling so you'd want the fan fully on.



The above numbers came directly from Honda's service manual. "Starts to open"? "Fully open"? Sounds like one of those Standt SuperStats that actually do slowly open and close. Even Wikipedia descibes them all as working this way: http://en.wikipedia.org/wiki/Thermostat#Automotive



My guess on the actual Honda thermostat is it's just like a regular car one and I know darn well they don't just slowly open and close. If you start out cold on a highway drive you can actually see when the thermostat opens. Most I've noticed hold closed until you get a bit above where your temp gauge would normally sit then wha - you can see the temp gauuge drop a fair amount when the stat opens and it takes a little more driving until it creeps back up to the normal area.

------------



OK, enough for theory. Let's say we're riding along, temp gauge sitting at normal range, thermostat is probably fully open sending 195*F (90.5*C) water into the radiator but how much is it getting cooled off by the radiator before it's returned to the engine to be returned to the 195* it sends back out again?



Although the sensor could be mounted anywhere, logic states that monitoring the water temp coming back out of the radiator would provide much closer regulation on the fan speed. It also just happens to be the easiest place to mount the sensor. There's a nice open blank place next to the lower hose - you wouldn't even have to remove the shroud to get at it.



When I realized this I knew the "on" temperature was going to have to be lower than 180* and while I still intend to make it adjustable I'm trying to figure out just how much adjust ment range to put in it?



This brings on yet another problem. If I make it to where you could adjust the sense temp to 160*F then the high end point would lower by an equal amount putting it at 190*F unless of course I included a second adjustment to allow for a "slope" in the gain. If I do that the two would somewhat interact with each other; if you changed one you'd have to tweak the other.



It would make it more difficult to set up but we've got to get this right the first time.



Ideas?
 

·
Registered
Joined
·
347 Posts
Marshall, yes, ALL wax-plug thermostats (even used on automobiles) do 'slowly' open and close, depending on the opening/closing temperature of the thermostat.



What you are forgetting, is that until that thermostat opens for the very first time, the coolant in the radiator is VERY cold, while the coolant in the engine block is VERY warm - once the thermostat absorbs enough thermal energy to open, that initial VERY cold coolant flowing back into the block causes your coolant temperature gauge to drop immensely.



As a matter of fact, that behavior is exactly why no OEMs now use linear gauges - they are (almost) all a variation of a log scale. In other words, that first 1/4 of the gauge might be 0-120F, then the next 1/4 is 130-180F, then the next being 190-220, and the last quarter is 230-250+



In other words, in 'degrees of range'



Quarter:

1st - 120 degrees range

2nd - 50 degrees range

3rd - 30 degrees range

4th - 20 degrees range



OEMs do this, otherwise if you took a constant TRUE measuring of the coolant temp, it would constantly fluctuate from something like +/- 10% of the thermostat rating, and Average Joe would spazz at the needle moving so much.



They 'buffer' the signal in various ways (GM uses fully enclosed coolant temp sensors in their OBD1 systems to smooth out the CTS signal, for instance) - Toyota was one of the first to use a properly responsive CTS for the ECU, and 'buffer' at the gauge (in automobiles, anyway).



Ok, enough engineering for now, I gotta go fab my muffler mounts LOL



My point is - just keep both signals adjustable.
 

·
Registered
Joined
·
2,601 Posts
OK, here's an update on where I'm at in the design. I've got enough of the key components in to start running some tests but actual bike tests won't start for a while yet. Hoping to test it out on a 650 with the stock fan as well.



From what Shep has found (under varied operational conditions on a cool day) these bikes tend to run between 75*C - 85*C (167*F - 185*F) at the cylinder heads and maintain a fairly consistent 60*C - 65*C (~145*F) water outlet temperature from the radiator - in other words they seem to be happy so long as they're getting around that temperature back from the radiator regardless of what's being fed into it. The mission is therefore to control the fan sped to maintain that condition.



1) Since I'm taking the temp from the bottom of the radiator by the outlet there's tons of mounting room, it's also simple to get to and to snake and hide the sensor wires up from there. Seing that thermally conductive epoxy appears to be somewhat difficult to find in smalll quantities I figured I'd just go with a larger sensor in a standard TO-220 case that could be mounted with a strap across it and regular epoxy, or however you want to do it. Having the larger sensor with a metal tab on the back side attached directly to the sensor die of the IC also means that it's going to pay most of its attention to the temperature of what it's mounted to instead of any air flowing around it thus no need for the heat conductive epoxy + some insulation over it method.



A common TO-220 case looks like this, matter of fact this is the actual sensor I chose:

http://www.futureelectronics.com/en...rs/temperature/Pages/3630068-LM35DT-NOPB.aspx



2) There will be an internal jumper that changes the sensing range from low to high so it can be used at the top of the radiator if desired. THough these aren't exact numbers, tentatively the low range will cover from 50*C - 75*C (122*F - 167*F) and the high range from 75*C - 100*C (167*F - 212*F) The jumper changes the range while a 25 turn trimmer potentiometer is used to fine tune the setting as 1 turn = 1*C



3) There will be an adjustable "gain" ratio for the output which controls how quickly the fan speed is increased in relation to temperature rise. This also serves to determine the point at which the warning light comes on; if the circuit is trying to drive the fan faster than 100% then some sort of fault is suspected in the cooling system.



4) There will obviously be output wires for the fan motor, and this thing can handle about any fan (or combination) you could ever throw at it with virtually 0 voltage drop thus full speed fan operation will never be compromised.



5) There will be an output wire for abovementioned optional warning light which will also come on if the fan itself stops turning when it's supposed to be. This will detect a burned out fan or one that has become jammed in a "locked rotor" condition.



6) Temperature output wire. There will be two output choices selectable by a jumper - one will be for *C and the other for *F. This can be used to drive one of those little digital meters you can get on eBay or, alternately, one of those dot/bar display circuits. It will include an offset adjustment so you can choose to have it display the temperature actually returning from the radiator or you can add voltage such that it will more closely display the operating temperature of the engine. In either case the output will be accurate for where the sensor is mounted (as in if you put it at the top) but only a close estimate if you're adding a bias adjustment to it.



If I choose to make a dot-bar display I'll probably add in a switch that will allow it to double as a voltmeter.
 
1 - 20 of 37 Posts
Top