My PWM fan controller project, continued....

[Tesla]

I may be missing something, but as I understand it you are trying to maximise airflow through radiator with minimum current draw, so your measure of success to some degree is CFM/amp.
The radiator is a big temp buffer for engine, you do not need to constantly manage temp, just keep it in a safe range for engine to be able to maintain a safe operating temp, intermittant fan operation at max efficiency is probably the simp[lest and best way to go IMO.

PWM controllers are a more efficient way of controlling motor speed than a variable resistor, but that does not mean controlling speed is the most efficient system.
With any fan/motor or pump/motor combination there is an optimum speed of operation, kind of like a torque curve on an engine, at this speed it will move the most fluid/per watt of electricity used.
I think you could just use 1 or 2 off the shelf units, like these:
10 50V 40A DC Motor Speed Control PWM HHO RC Controller 12V 24V 48V 2000W Max | eBay
and play with adjustment while monitoring current flow to determine fan/motor combinations optimum speed.
It would probably be difficult to measure air velocity because of turbulence, but you could set up a simple U-tube manometer with a cup on measuring side that is mounted flush on centre of radiator. Not sure if you have puller or pusher fans, pullers mounted on engine side with full shroud are most effective, in this case your manometer would be used to measure suction on front of radiator.
Once you have determined this setting, then it’s simply a case of determining coolant temp trigger points, simple thermal switches have built in anti-flutter temp on/off range of about 5°C, i.e. switch on at 90, switch off at 85.
Thermal switch controls solenoid to power the PWM units, hence fans and maintain a constant coolant temp range.
I think it will be most efficient to run the PWM/fan units at optimum efficiency for short bursts than running them constantly at varying speeds.

BTW, I've been running without any fans at all for 18 months and would be lucky if I got a 1% improvement in FE, I know there is some benefit, but I have not been able to quantify it as it was too small in my case.

[plasticuser]

It's the intelligence that can be brought to bear that interests me. A dumb PWM system like you linked can only respond to coolant temperature. It operates only in hindsight and cannot anticipate the future. A load-aware PWM controller can feed cooler coolant in higher load conditions. One of the problems is that coolant flow rate depends on engine RPMs, so low load low rev conditions can heat the coolant as much, if not more, than high rev and load conditions in some cases.

My goal is to create a system which can smartly respond to engine conditions, speed, ambient temperature, load, etc to provide efficient control of fans, a grille block, and possibly future options to control an electric coolant pump, DFCO, warm air intake flap control, etc. The idea is to produce generations of increasingly advanced and more capable designs (with open-source code, schematics, gerbers) so people can develop the ideas further, based on a common hardware reference platform that doesn't have the limitations of an Atmel 328 or *guino....

I know that isn't right for everyone, but it's right for many. As is a plain and simple PWM controller. They both have things going for them.

[toc]

Originally Posted by plasticuser

so people can develop the ideas further, based on a common hardware reference platform that doesn't have the limitations of an Atmel 328 or *guino....

What limitations exactly in the 328/guino are you referring to?

I don't believe you have considered the Atmega328 completely.
I know I could put a fan controller together which looked at controlling fans based on load, outside air temp, speed of the car - using the ATMega328.

A previous post from you labelled 12V as hostile and you wanted a clean '3v3'..

12v by itself isn't hostile, so again, I question what it is you are referring to here? What makes 3v3 cleaner than 12v exactly?

[plasticuser]

The 328/quino can't do hundreds of things, run a graphical display well, lots tons of data then upload it to somewhere else or display it graphically.... The Atmel chips are great for embedded custom use, like if this were just going to be a dumb PWM controller, but they don't allow for so many options that a more capable board can do.

As for the voltage issue, a well regulated 3v3 supply is "clean" - the 12v supply in a car is not. It's noisy, prone to wide variation, just a plain old bad environment for delicate electronics, generally.

Why not find an Atmel-based project and posting constructively there, instead of criticizing and asking for justifications in a project that clearly isn't of interest to you?

[toc]

Originally Posted by plasticuser

The 328/quino can't do hundreds of things, run a graphical display well, lots tons of data then upload it to somewhere else or display it graphically.... The Atmel chips are great for embedded custom use, like if this were just going to be a dumb PWM controller, but they don't allow for so many options that a more capable board can do.

As for the voltage issue, a well regulated 3v3 supply is "clean" - the 12v supply in a car is not. It's noisy, prone to wide variation, just a plain old bad environment for delicate electronics, generally.

Why not find an Atmel-based project and posting constructively there, instead of criticizing and asking for justifications in a project that clearly isn't of interest to you?

Well it can do a graphical display correctly, it's done GLCDs before...
It can gather plenty of data for it's intended applications... and you can equip it with an SD Card for more storage.
The Atmel chips are indeed capable - PID PWM is about as controlled as you can get something, and they do this well...

Atmel chips want 5V typically - so how does 5V compare to 3v3? How do you propose to avoid the 12V in getting your 3v3 or 5V? I'm simply questioning your usage of clean 3v, compared to 12v and somehow applying this to the atmel chips.

I'm interested in the project, it was the reason why I opened it up in the first place, as I have my own Fan controller that I put together - it did not need to be fantastic, just on when I knew the thermostat was open and the car required cooling - no PWM needed or anything - and it was quick to setup.
I like the idea of making it better - hence why I looked at this thread, but I read through and find what amounts to rubbishing a perfectly capable microcontroller, and trying to adapt a ARM chip running linux to the application that is far more suited to the Atmel chips.

And of course, I read to find 12V labelled dirty, and 3v labelled clean, in an irrelevant context as in BOTH cases you must take 12V and regulate it down.
This then raises the question as to whether you have the ability and understanding to put this together and see it to fruition. Considering your dismissing a perfectly capable controller, and trying to adopt something else that isn't really suited (come on, 1Ghz and Video for what amounts to a MOSFET) - - I don't believe so!

I asked the questions to see if there was some element to your dismissing the ideal Atmel chips and see if what you are planning would make sense - and all I get is a go away, I don't know how to answer your questions with answers that fit.

What goes into selecting a controller?
Simple.. It's target application. You don't throw a new server at a user for web browsing. You give them a simple workstation. Same principle would apply here - give the application the intended chip it deserves, and don't complicate it to the point that it gets unworkable.
It's simple fan control - and you won't get far with the thermostat shut anyway.

After I went through my fan control - It turns out they rarely ever need to run - they run when idle, when I power up the AC, and when slow through traffic (I've got a light in my car telling me they are on so I know!). They rarely ever come on...

The controller that came with the car (faulty, so I made one to replace it) did speed control using MOSFETs - simple control too - temperature gets warm (above thermostat temp), and speed is not > x then it runs the fan proportionally faster. That was from Mitsubishi.

What your going on about is hardly what the application needs - the car is already warm when the thermostat opens, the fans need to modulate speed to reject enough heat to get that thermostat shut, and then stay off.
Clean, simple, no brains control. No need for thousands of logs, no need for any graphical display for it, small chip, a bit of a small circuit (go on, do that with the Raspberry PI!), and a bit of voltage regulation (which will be needed anyway).

Anyway, I'll let you waste your time, It's fairly clear to me that your trying to overcomplicate one of the most simplest of tasks.

[AndrzejM]

Guys... Chill out!
There are many ways of dealing with this issue. It can be solved with simple electronic or even Intel's i7 processor. And it doesn't matter.

I like the idea of PWM controling the fan, and I'm thinking of building the "temperature Management System" (as I call it), based on Atmega328 (which i have anyway), with an LCD screen showing outside temp and coolant temp, with a backlight color changing accordingly to the temp (yellow - too low, green - proper, red - too hot). And not only using PWM channel to control radiator fan, but shutting and opening the grill block.

I'm hoping that soon I'll be able to post my project details.

Cheers!

[plasticuser]

Originally Posted by toc

This then raises the question as to whether you have the ability and understanding to put this together and see it to fruition. Considering your dismissing a perfectly capable controller, and trying to adopt something else that isn't really suited (come on, 1Ghz and Video for what amounts to a MOSFET) - - I don't believe so!

When you resort to ad hominem attacks to make your case, you completely lose my desire to engage with you in any way.

[gumby79]

How about a buld off
All of you sound like you know what your doing and have stalled in your brilliant ideas over witch hardware to build on. Simple answer. Both/ all and let us have some fun with our hobby ,that saves some cash,advanced ourselves and our community together
Lets push back the boundarys of what current technology can do, and make it what it will do.

Ive Been talking with the family computer guru about him helping me work with the Raspberry-pi. He has built a mesh network with this device, this is why I chose to have my pi and eat it to. To be able to prove my mpg/CDa improvements I need a data loger as sophisticated as a new car. Installed in a car that thinks a computer is a brand of tire. You can see I've been thinking about this for a few weeks and found this posting. About people trying to open sorce build a platform that can expand to meet the needs of the Community be it 1or 2 data streems,like your simple gas non- turbocharged dual fan progressive fan controller or like me and could exceed 100 points of data to control the fans and different pumps,shutters propane water heaters, aero Modds ,multi level hybrid. Ground effects Russian dall boat tail all but the water heater are ideal ideas on this website that have stalled have stalled waiting on this project . This first step and giving our creations a BRAIN



1 thing I know for certain , is WITH THIS COMMUNITY I can see 50mpg From a 5600lbs full size truck only 25avg on 5 trips 200mi round trip over the Rockies . for now

How about both of you throw down your best controller. Or at least able to run my 4 parameters on a progressive rate
(-)5000cfm@45mph through 15000cfm sitting still.

Now what I need to control my 4 parameters of cooling with 3of the largest cfm fans available (3k-5k cfm ea.).
I am not biased Im just wasting 25hp of fan to keep the thermostat closed a @165°f, the stock cluch fan kicks on @ 180f Off @ 175f the same as the opening of the thermostat
Previous Owner advised me not To go back to the stock 98°C because"it will kill your mpg"
By making the 25hp Fan run all the time.


.
viscous--vs-electric-variable-engagement-vs-all

Originally Posted by gumby79

I have a 91.5 ic 1st gen Cummins says my fan pulls 25hp engaged a ckass8 up to 75hp.

Option1
Update to 3rd generation electronic fan cluch.
Could a person use a variable speed fan controller to achieve the same variable % of lockup? Eg the hotter the radiator the higher the voltage the fan sees/the higher %of lockup? Or is the signal a 0-5vdc standerd automotive computer language ? If so this can be worked around.
Option 2
Stock replacement? And adjust the bi metal spring to come on at say190f Instead of (my original equipment )180f on 175off coolant temps.
Option 3
full custom progressive 9,000-15000 cfm electric fan, would be comparable stock 10,000cfm macanicel fan?


1 coolant

step1 180f 1000cfm step 2 3000cfm 185f step 3 190f 2nd fan @1000Cfm 1st @full step4 195f 2 fans @full 3000cfm ea step 5 200f fan #3 @ full 3000cfm for a total of 9000cfm

2 iat(Intake Air Temperature )

Step 1 130f 1 fan full step2 200f 2 fans step 250f All 3 @full
As per cummins any iat Above 90f will reduce power at a rate of 1%per 10°f above 90f https://drive.google.com/file/d/0B0l...p=docslist_api

3 ac/defrost compressor

Add a head pressure switch from a class8 for 1 fan only.
The stock Dodge AC safty SWITCH WILL hit MAX turn the FAN ON for a moment 1-5 sec, simultaneously turning off the compressor,
Pressure drops turning the FAN OFF (as a safe level has been achieved )turning the compressor back ON for a few seconds 1-5

3b Ac / defrost compressor
Wire the compressor to a fan . Compressor ON fan full ON compressor OFF fan full OFF . No good with options 1,4and5 (opinions wanted )

4 trans temp
150f a separate fan on the air to atf cooler with a reverseing contactor relay to reverse the fan pulling hot air from under the hood @80f to keep it warm. @-10 and lower so that as per Dodge tsb#07-01-89 TSB 07-01-89 I dont have to drop to 3rd to keep the atf from freezing and burning up my trany killing FE yes this is a problem Butte, MT is listed as the 18th coldest city in the continental United States
__________________

have expansion capableity of;
brushed(BDC) and brushless(BLDC) motors
4Intake temp
4 Intake psi/vac
VSS
TPS
ALT LOAD
FULE CONSUMPTION
2 WATER
1 ENGINE OILTEMP
1 TRANS TEMP
1 diff Temp
G-force
Altitude
Gps
Air speed and direction

All temps Must read to -60f if something gets out of Wack AIT can exceed 400°f and be well below absolute max
Now for my contribution :
Clean X vdc any from supply down to 4.8vdc @10A CC BEC or 20A CC BEC PRO a simple resister for final adjustment below 4.8vdc. BEC ( Battery Eliminating Circuit )link

Servos to open / close thinges like active grill use pulses width modulation on signal for position or can be hacked at the position pot. As well as liner actuators for heavier applications eg Active extended boat tail diy-active-rear-aero- link


BLDC / BDC Motors and controller RC Esc (motor controllers) includ a BEC
Forward only from avation (something that self-deploys and pi brings in @xxmph)
Forward /brake/reverse ,(force open hold force closed) Forward / reverse car, boat
Motors to replace or upgrade BDC With BLDC or vica verca eg fans ,ac Blower motors ,wipers ,power windows, ect,ect
avation motor link
land motors up to5kw 85%+efficiency
Even water cooled this is only 1 manufacturer, their are too many to shake a stick at
lets Have some fun with our hobby that if dune right, will fattens our wallets.
Gumby stay flexible.