Electric Water Pump How to.

Honestly, I don't think I would eliminate the T stat all together, but maybe put a 165 degree in with a 1/8 bleed hole to get to temp quicker and let the pump modulate from 185-190 degrees, then from 190-200 modulate the fan. Then you maye need some sort of PID control system with all of these interdependent loops and latencies to keep it from oscillating wildly. That's part of the reason I have the fan bang bang, on or off with a 5 degree hysteresis.

Hmmm...all the recommendations I've seen for EWP use is to eliminate the Tstat and use a sensor based EWP controller. Certainly that's the recommendation from Davies Craig in their installation instructions.

Does anyone know if the OEM cars that use an EWP do so without a Tstat? For example, do the BMWs that use the Pierburg EWP do it solely with electronic controller or do they also have some sort of physical Tstat?
 
Find out what model uses the EWP and then check if Rock auto list a T stat for that model, I think they use a 3 port Tstat that routs the water to the intake of the pump when cold, and the radiator when hot. Otherwise you'll take a long time getting to temp with the reduced, but continuous flow of water through the radiator. If the water is stopped, you'll boil water in the head before the hot water gets to the cold manifold... like a SBC with the heater crimped off.

But I can seeone reason to eliminate it... it is a restriction when open.
 

Randy V

Moderator-Admin
Staff member
Admin
Lifetime Supporter
I used a thermostat for a BMW 2002 like this one -

We carefully welded on extension tubes that also adapted the size to the hoses to be used. It can be seen in the center/bottom of this picture.
IMG_2598.jpeg
 

Ian Clark

Supporter
Great find Randy on the BMW 2002 external thermostat housing, an affordable problem solver!

The mechanical route has been the standard for what, more than a century. Works very well, when sorted, in our applications.

There are Pierburg EWPs available in the aftermarket with the plastic BMW remote thermostat housing included. Looks like it also provides a bypass to warm up the engine quickly. There's also an electrical connection for a built in sensor, hmmm.

Tecomotive offers an EWP controller to manage the pump: Tecomotive - tinyCWA A brilliant device.

The simplest way to do an electric conversion would be to follow the manufacturers recommendations, Meziere, Davies Craig, Tecomotive/Peirburg all know what they're doing and have been down this rabbit hole many times...

If you're building a lapping day track car the priorities, parts required and planning are much much much less than building a road/track capable driver with A/C and heat, regardless of mechanical or electric water pump.

Cheers
Ian
 

Attachments

  • Aftermarket Pierburg EWP w Thermostat_Bypass.jpg
    Aftermarket Pierburg EWP w Thermostat_Bypass.jpg
    27 KB · Views: 51
Thanks for the information on "bypass thermostats" (it's helpful) but is somewhat of a different topic than the question posed in post #58 https://www.gt40s.com/threads/electric-water-pump-how-to.57366/post-588792 I think the question of how best to quickly get the coolant system up to target temperature is a good question but not the same question.

Here's hopefully a simplified form of the question. Which is more effective/efficient: a coolant system with temperature control done by pump regulation or flow regulation? The context being that "pump regulation" = electronic temperature sensor integrated with pump controller to maintain target temperature and "flow regulation" = constant pumping together with inline thermostat blocking/enabling coolant flow based on temperature.

If no one has an answer, that's fine, but I thought it worth bringing the original question back to the forefront.
 

Ian Clark

Supporter
The manual for the Tecomotive Tiny CWA is a good read, 28 pages and covers most scenarios, pump flows, with or without thermostat, bypass or not, sensors type and location, radiator turn on/off temps, aftercooling, speed etc.

I can see this device reducing current drain and extending the life of the battery, alternator, EWP and fans.

Also covered are wiring diagram, custom programmability and compatibility with pumps other than Peirburg.

Worth the read no matter what you settle on for equipment.

Cheers
Ian
 

Attachments

  • Tecomotive _tinyCWA_V3.5 manual.pdf
    2.8 MB · Views: 56

Randy V

Moderator-Admin
Staff member
Admin
Lifetime Supporter
Thanks for the information on "bypass thermostats" (it's helpful) but is somewhat of a different topic than the question posed in post #58 https://www.gt40s.com/threads/electric-water-pump-how-to.57366/post-588792 I think the question of how best to quickly get the coolant system up to target temperature is a good question but not the same question.

Here's hopefully a simplified form of the question. Which is more effective/efficient: a coolant system with temperature control done by pump regulation or flow regulation? The context being that "pump regulation" = electronic temperature sensor integrated with pump controller to maintain target temperature and "flow regulation" = constant pumping together with inline thermostat blocking/enabling coolant flow based on temperature.

If no one has an answer, that's fine, but I thought it worth bringing the original question back to the forefront.

Your more focused question might be better asked in it’s own thread rather than trying to direct this thread down that path. I refer you to the subject line and post #1.
 

Howard Jones

Supporter
The overall temp of the coolant (at any given moment) is a function of the volume of coolant flow through the radiator, the amount of heat generated as modified by oil heat reduction and engine room ventilation, the desired stable coolant temp, and ambient air temp/current coolant temp differential as well as airflow through the radiator.

The various methods of regulating coolant temp with EWP's are based on varying coolant flow. Some do it by restricting the flow with a thermostat or simply a round plate with a hole in it replacing the thermostat and running the pump at constant speed. Others do it by what I would call a duty cycle method in which the pump is turned on and off with the longer it is held in the on condition the more the resulting average flow. Davies Craig uses a controller that varies pump speed by monitoring coolant temp with a sensor and electronic system control.

You can see that the EWP method uses a system that either stabilizes the pump impeller speed and relies on varying restrictions to control the flow rate or stabilizes restriction and varies pump speed or could use varying duty cycle although I am not aware of an automotive system that does it this way. IMHO I think either the DC or duty cycle system is better suited to a dual-use or street car because it can react to the wide range of heat generated/ airflow/ambient air temp that is seen in that kind of car. A track-only car can better use a constant speed pump and a restriction. This is because the heat load does not vary as much and the average heat load remains at the top of its range most of the time as well as the airflow also remains high.

The traditional method utilizing a mechanical pump running a derivative of engine speed tries to do both at the same time. Varying restriction with the thermostat and at the same time continuously changing pump speed. The problem is the pump speed isn't really reacting to changes to the heat generated or vertical seed (airflow( and the thermostat has a very limited ability to gauge heat load quickly.

Joel. There is really only flow regulation. Just several different ways to do it. Both airflow and coolant flow.
 
Last edited:

Randy V

Moderator-Admin
Staff member
Admin
Lifetime Supporter
Howard - not so sure I would do the duty-cycle method you just spoke of. Coolant in the heads, particularly in the vicinity of the exhaust port is going to get hotter than the rest of the engine in a very big hurry. One of the key reasons I chose a constant speed pump and a bypassing thermostat was to stabilize the temperature throughout the engine.
Small block Fords, in particular are challenged in the head-gasket department and don’t tolerate hot heads or detonation..
A side benefit was the heater would get full circulation all the time.
 
I think we're conflating "duty cycle" with variable speed. I think they're the same. You vary the duty cycle (PWM) to the signal pin to use the variable speed function of the pump's internal brushless perm mag A/C motor controller.

You're not actually starting and stopping the pump.
 

Howard Jones

Supporter
Brian. I think that if you review post # 11 you will see that I agree with you. In order to not confuse others I will not use the term duty cycle again. This is a term that is often used when discussing load/time in electrical systems. It is one that I have been familiar with over my 50-odd years in the industry. We all tend to use the vocabulary that we are accustomed to, however, since my intention to add information to the discussion of EWPs isn't served by confusing terminology, hereafter when discussing flow control with EWPs it can be assumed that I refer to varying pump speed.

It should be noted that DC's system does vary pump speed and that the speed range of the pump does include zero rpms on occasion especially when the actual coolant temp of the system falls significantly below the target temp programmed into the controller.

I like the PWM control and BMW pump. If they had been available when I selected the DC system I might have gone that way. So far the DC system has proved reliable in my car for a few years. It will be interesting to see how the newer combination proves out over time.
 
Last edited:
The overall temp of the coolant (at any given moment) is a function of the volume of coolant flow through the radiator, the amount of heat generated as modified by oil heat reduction and engine room ventilation, the desired stable coolant temp, and ambient air temp/current coolant temp differential as well as airflow through the radiator.

The various methods of regulating coolant temp with EWP's are based on varying coolant flow. Some do it by restricting the flow with a thermostat or simply a round plate with a hole in it replacing the thermostat and running the pump at constant speed. Others do it by what I would call a duty cycle method in which the pump is turned on and off with the longer it is held in the on condition the more the resulting average flow. Davies Craig uses a controller that varies pump speed by monitoring coolant temp with a sensor and electronic system control.

You can see that the EWP method uses a system that either stabilizes the pump impeller speed and relies on varying restrictions to control the flow rate or stabilizes restriction and varies pump speed or could use varying duty cycle although I am not aware of an automotive system that does it this way. IMHO I think either the DC or duty cycle system is better suited to a dual-use or street car because it can react to the wide range of heat generated/ airflow/ambient air temp that is seen in that kind of car. A track-only car can better use a constant speed pump and a restriction. This is because the heat load does not vary as much and the average heat load remains at the top of its range most of the time as well as the airflow also remains high.

The traditional method utilizing a mechanical pump running a derivative of engine speed tries to do both at the same time. Varying restriction with the thermostat and at the same time continuously changing pump speed. The problem is the pump speed isn't really reacting to changes to the heat generated or vertical seed (airflow( and the thermostat has a very limited ability to gauge heat load quickly.

Joel. There is really only flow regulation. Just several different ways to do it. Both airflow and coolant flow.

Thanks Howard. I think I have a good understanding of what's going on. I'm guessing a person who could back this up with actual quantitative analysis is likely busy in a day job in auto industry and not hanging out playing with their automotive toys.

I dug up the Davies Craig Installation Manual for the EWP and Controller. I found this interesting graph for how their controller operates:

20231020_085606_0002.JPG


It varies the voltage and does pulse the pump as well. It looks like the pulsing starts at key on and then increases until nearing target temperature where the voltage is ramped up to increase flow. This makes sense as there is only 2 wires going to the EWP from this external controller unlike the pumps that take in a PWM signal to an internal controller circuit.

It's understandable that a street car with AC would need a much more sophisticated coolant control than a track car. There's just a lot more variables in the equation. My thought while building the Miura has been to have minimal manual coolant flow restrictions as the system already has many feet of tube and several 90 degree bends just due to packaging considerations. So I'm going to initially try it without a manual thermostat and will adjust if necessary.
 
My mistake. I didn't realize that controller did actually pulse the pump when well under target temps. I see a good reason to do so as reliably controlling speeds open loop below 30% or so is iffy at best. I'm assuming the EWPs with an internal controller are better with this.
 
Hi, Newbie Rob H here. Has anyone used the Davies Craig #8670 adapter for the MWP delete? Also, has anyone used the PWM Mastercell option to run the EWP 150?
 
Back
Top