DIY electric water pump - on the cheap

[serialk11r]

Load and coolant temperature is the way to do it I think. Load is what determines average cylinder temperature for the most part, and higher coolant temperature means you want to circulate it faster to improve cooling.

[nemo]

Originally Posted by Daox

Now, we have the duty cycle that the temperature effects. For example, lets say this is set to 50%. This means the temperature has 50% of the 80% max that it controls (so its really 0-40% of the total duty cycle).

Let me rephrase my question. Why wouldn't you want the temperature sensor to be able to go to 100% of the 80% if a very high temperature reading existed.



If you use an electric fan it may use less power to turn the fan on at various times as opposed to speeding up the pump.

[Daox]

Originally Posted by nemo

Let me rephrase my question. Why wouldn't you want the temperature sensor to be able to go to 100% of the 80% if a very high temperature reading existed.

The example quoted was just an example. In actuality, it likely will have 0-100% of the 80% max. The scaling will be tweaked in the program to maximize efficiency. The reason being, when the engine is up to temperature and at idle, the heat load will be much lower than WOT at 5000 rpm. Thus the temperature alone does not need to be putting out 100% simply when the engine is up to temperature if you're only at idle. The 100% will be reserved for 200F+ temperatures.

[Maco]

Hi I am not saying this is the right way to do it, but if I was going to do it this is how I would go about it because its easy to do.
I would use the 12V from the cars battery but I would replace the battery with a proper traction battery so if it runs down it wont kill the battery.
I would wire the motor through the existing fans thermostat so it gets turned on and off as the engine gets hot, you can get 70 deg C thermostats if you need to turn it on more often before you up the voltage to 24V.
On the mechanical side I would remove the water pump and make a solid press fit coupling so that the motor drives the pump directly inline then you can mount it where ever you like in a sub frame and it will be quiet and eliminate chains and pulleys.
On the output side of the thermostat I would rig up an LED light in parallel with the motor, and put the LED in the cockpit so that I have a visual check that it comes on regularly. The logic being that the fan thermo couple will regulate the temperature for you by turning the pump motor on and off automatically so the pumps RPM will become irrelevant provided there is sufficient cooling at the 12V pump RPM if it just stays on then you may have to go to 24V. this is the simplest way I can think of, its self regulating and at least you know the pump is pumping.
If the fan thermo couple cant handle the pump motors amps just wire it into a contactor so it activates the coil in the contactor rather than having the motor amps go through the thermo couple, the contactor coils pull negligible amps and if you use a contactor the main power can be 24V if necessary but you can specify a 12V coil in the contactor so the activation circuit is still just the standard 12V

If you over heat your engine it will be very expensive to fix - new head gaskets etc - so from a "save money" prospective it may turn around and bite you, especially on those hot summer days.
Provided you monitor the water temp all the time you will be fine - You could rig up a system that would sound an alarm if the temp goes over 100 degrees by fitting a second fan thermo couple and connect it to your cars hooter - lol then at least if the electric water pump is not doing its job then you will know for sure before you break any thing, where you mount that 100 degree sensor is very important. I have driven cars that are over heating but show a low temperature on the temperature gauge this is because there is no water in the radiator any more and the manufacturers fitted the sensor too high up in the radiator. So that sensor must be as low as possible - maybe in the engines sump oil but I dont know how hot the oil gets so I cant recommend a temperature there, but its probably the safest place to monitor overall engine temperature. So you can do it, just make sure you add a monitoring circuit as well then it will be pretty safe.

[blownoiler]

That old saying that goes " a dollar not spent is a dollar saved" comes to mind here, rigging up a reliable system that will pay for itself in fuel savings is difficult, even more difficult is having a system that is 100% reliable as an engine overheat can cost a lot of dollars. Until you work out an almost failsafe fuel saving system via an electric pump, you can save some parasitic loss by running the waterpump v belt on the loose side. Let it slip a bit, and if you want to retain full throttle/high heat load capability then you could rig up an electrically activated belt tensioner for those full throttle moments. You could run with this idea and polish the v belt pulleys to a high shine to ensure that the belt will slip easily. (don't blame me if it overheats ).
Power transmission via v belt is around 97 ish percent efficient if my memory serves me correctly, I'm wondering what the losses would be turning the pump with an electric motor, I suppose that will depend on where the power to run it is coming from.
I have to wonder what happens to coolant flow pressure when the thermostat has not yet opened and the thermostat is bypassing coolant via a small return pipe/orifice to the engine, surely coolant flow is hindered in comparison to the main radiator flow path, maybe there is a gain in efficiency lurking in the bypass side of the cooling system!

[ctgottapee]

While a fun project, it won't increase efficiency much for the $, compared to say opening up the current water pump and removing a couple blades or trimming them to improve the efficiency, and replacing the pulley wheel size.

The engine is always spinning making a nice spot to add on a pump that also almost always needs to spin.
The car engine and system was designed to fall in an RPM range of efficiency with the water pump attached. Removing it, as you have already removed the alternator, will likely push the rpm range out of the intended efficient zone for your car.

Find a way through pulley wheel change and/or internal blade trim to get it's draw as limited as possible. Add a cheap inline helper electric pump as a backup emergency for high temps situations. Davies makes a cheap one of them sized for the small heater loop hoses, which you could also use to help circulate for pre-warming.


You could also try to simulate the Davies pump with an A/C clutch on your current water pump and belt system by using their programmer module that would turn on the clutch rather than turn on an electric pump.

[ctgottapee]

My V8 Cadillac, as well as most big V8 vehicles comes with transmission and oil cooling as standard, and you could possibly use such a system for pre-warming or drive time warming.

The radiator has two side panels, one for the transmission, one for the engine oil, with in and out routings. The trans/engine oil is then cooled/warmed as it pumps through. You'd have to find a way to tie into your lines/system. All three fluids are kept within a certain range as the radiator is kept within range by the cooling system.

It's meant more for cooling, but in the winter time it also serves to warm both trans/engine oil up too, as the radiator is warmed by antifreeze.

[Daox]

I agree, this isn't likely going to be a huge gain. I certainly don't recommend it as anywhere near a first mod. This is planned for a fairly modified car. I wouldn't even consider it unless you at least have an alternator delete done already. Also, its an idea that I hadn't heard of talked about here. Instead of spending a few hundred dollars on a pump, this costs under $100 with new parts, and can be salvaged for much less.

Its not just about the efficiency gain either. There are other benefits like having heat when the engine isn't running and being able to cycle the pump when the block heater is plugged in. It will pair well with my idea of using engine coolant to speed up transmission warm up times, especially with my short commute. It'll also free up a bit of horsepower at higher rpms which never hurts. So, it pays real well with other mods. If you don't have any of those other mods, it probably won't help as much.

[serialk11r]

Originally Posted by Daox

I wouldn't even consider it unless you at least have an alternator delete done already. Also, its an idea that I hadn't heard of talked about here. Instead of spending a few hundred dollars on a pump, this costs under $100 with new parts, and can be salvaged for much less.

I'd actually say it's a great mod to start out with if it's easy to reroute a shorter belt. In dollars per horsepower gained, this has to be one of the cheapest ones!

[Daox]

Its been a while since talking about this topic, and I'd like to get back to it. This time though, I'm thinking about putting this setup on the Metro. It already has a deep cycle battery in it and an alternator disable mod done to it. The only thing left being powered on the crankshaft's accessory belt is the water pump... so lets fix that.

So, first things first, how fast does the Metro spin its water pump? Well, the crank pulley is 5" and the water pump has a 5-3/8" diameter. So, it has a 1:0.93 ratio. I threw the numbers into excel to get this chart. At idle, our lowest water pump rpm is 700 (assuming 750 rpm idle speed).





For now, I'll probably consider 700 to be the low end of what is needed once the car is up to temp. This sounds somewhat logical to me, your two cents is welcome.

The high end will ideally be determined by the speed at which the water pump starts to cavitate. Getting rpms up there will be interesting though. The electric motor only spins 1250 rpm at 12V. I'd imagine I'd want a max water pump rpm of around 2000. That means I'll need a 8" diameter pulley which is pretty large period... I may have to look into making a new water pump pulley too and reducing its size.