So there are belt driven engine water pumps, and there are electric water pumps but both use a thermostat to limit flow to the radiator.
Is there a variable speed, electric, water pump that we could use on our engines?
It would be off when cold, slow when some cooling was required, and full on when the engine was "hot" (or just binary).
Because that would be neat (my google fu, has turned up nothing)
The only EWP that I can think of off the top of my head that does what you're asking for is a Davies Craig unit.
Although, I'm sure their controller could be made to work with a Meziere or similar pump.
Ah! Perfect! Thanks 
My google-fu must be getting weak in my old age.
Late model BMWs have exactly what you're talking about, though the come with an electronically controlled thermostat and complex ECU based control. Oh, and they cost about $300 for a replacement unit when it won't go into high flow mode at a track day.
A vent/purge cycle is built into the ECU though which is kind of novel after doing a coolant change-- fill until coolant comes out bleeder, close it up, hook up battery changer, turn ignition on and then hold the gas pedal down for 10 seconds. Sounds like a dishwasher starting up under the hood!
If playing with old school electric motors to turn a mechanical pump you could connect the motor to a heater blower motor switch and resistor. That would be manual, but I suppose you could use a series of sensors run to relays feeding the heater fan resistor to control the speeds.
We've been using the Davis Craig on one of our supercharged Miatas recently. It works like a charm and has increased the cooling capability of the vehicle considerably. No thermostat, it's all done with pump duty cycle.
Rob_Mopar wrote: If playing with old school electric motors to turn a mechanical pump you could connect the motor to a heater blower motor switch and resistor. That would be manual, but I suppose you could use a series of sensors run to relays feeding the heater fan resistor to control the speeds.
i shall build this!
-J0N
^Very clever. On a car with enough room in the engine bay, the electric motor could have a pulley attached and could drive the original pump on its original pulley. You could use an Arduino if you want electronic speed control like the Davies Craig EWP.
How about a variable V belt sheave like you find on some infinite speed control drill presses? Use a screw drive old Japanese style manual tensioner driven by an electric motor for speed control.
I have a Davies Craig with CPU for my RX7. Haven't installed it yet, but it came heavily recommended. The CPU can also control the rad fan.
Most standalone engine management systems should be able to handle this as well.
Kenny_McCormic wrote: How about a variable V belt sheave like you find on some infinite speed control drill presses? Use a screw drive old Japanese style manual tensioner driven by an electric motor for speed control.
Sounds like you've building a CVT for the water pump!
But a much simpler way to achieve the same thing would be a clutched water pump pulley, the new ND Miata has one. I'd get one if I could find one for a 4AGE.
In reply to GameboyRMH:
I thought about hacking on a AC compressor clutch, but IDK if it would last long enough with pulsed operation like that. I suppose it would be slipping a lot less considering the weight of a water pump impeller vs the rotating assembly of a AC compressor.
It doesn't need to be pulsed - just engage when the engine reaches full operating temp, and then rely on the thermostat. The engine will warm up much faster and save some energy when cold.
The one on the Miata is actually engaged when it's not powered - it has a friction clutch and giving it power disengages it.
Most of the high tech heating systems for your house use variable speed pumps and fans. Must be something you could gleam from a wiring diagram. Does the electric water pump vary its speed with the voltage delivered to it or is it just a go- no go thing. If it varies speed with voltage, then you could install a potentiometer that varies voltage output dependent on the signal coming from a temp sensor. If it doesn't vary, then you would need to turn the voltage on and off which isn't good for electric motors.
Voltage control is so last century, pretty much all modern electric motors use PWM control, which is "go/no go" happening really really fast.
An Arduino mated to Rob's setup should create a pretty automated variable system. The code should be something like:
Any temp under 100 deg f would have no output so the pump would be off.
Dag gum, that's so easy I may just try it! I can solid state relays to switch the higher amperage lines via the digital outputs.
I really like the Davies Craig CPU, I wonder if megasquirt will pick up on this idea (wink wink)
In reply to GameboyRMH:
I think the design goal here is to use a minimal amount of power to run the pump so you can use that power to move the car instead. Spinning it slower (as its geared to spin fast enough to keep things kosher while climbing a hill in death valley at high noon) when it's not needed to pump as much coolant would be effective for that.
In reply to Hungary Bill:
Is the wink wink sarcasm? 
Because you can do that on Megasquirt. Even full pwm.
GameboyRMH wrote: It doesn't need to be pulsed - just engage when the engine reaches full operating temp, and then rely on the thermostat. The engine will warm up much faster and save some energy when cold. The one on the Miata is actually engaged when it's not powered - it has a friction clutch and giving it power disengages it.
The Davis Craig one still pulses when the engine is cold. Keeps you from developing hot spots. But it's fairly low duty cycle, something like 5 seconds every 30.
Swank Force One wrote: In reply to Hungary Bill: Is the wink wink sarcasm?
Nope. I honestly had no idea it was a possibility!
(embarrassed to say I have a MS3 sitting assembled in my basement)
Cool part about the Davies Craig is that it looks like a turbo and you can easily confuse idiots.
Old thread, but: How large/small are these things?
I have access to one of these bad boys for testing/mockup purposes: http://www.ebay.com/itm/NEW-OEM-WATER-PUMP-2013-2014-FORD-C-MAX-LINCOLN-MKZ-2-0L-DOHC-GAS-ELECTRIC-/221825977668
I want to use it on the RX-7 so I can cool the engine off in grid without having to have the engine running. But GOSH the thing looks so tiny!
It's a neat setup, controller is built in, give it power and ground on the big fat contacts and PWM on the other two. And unlike an aftermarket pump, this one is junkyardable. car-part search shows a place in Lansing with new ones for $110.
Not having your cooling coupled to engine speed means you can cool things more efficiently. Specifically with high-revving engines the pumps can actually cavitate at high RPMs.
In reply to DaveEstey:
I used to overheat my RX-7 at track days if I kept revs over 6000rpm. Later, when rallycrossing, I noted mass overheating at high revs. Went to a little 3" crank pulley and overheating stopped.
Pumps have efficiency ranges, past their effective speed they just make vacuum bubbles rather than moving coolant.
You'll need to log in to post.
