Electric water pump and running without a thermostat
06-10-2009, 08:08 PM
#1
Rennlist Member
Thread Starter
Electric water pump and running without a thermostat
928 Geniuses,
I am curious about hypothetically replacing the stock water pump and thermostat with an electric pump and controller. I understand this would require significant modification to the water bridge, and other parts of the cooling system, but my question is strictly about the removal of the thermostat and mechanical pump.
Assume that the mechanical pump impeller has been removed, and the water bridge has been modified so there are only two ports:
(A) the combined outlet from both heads
(B) the inlet port on the block.
In this scenario:
Thanks for your help in talking through this hypothetical installation.
I am curious about hypothetically replacing the stock water pump and thermostat with an electric pump and controller. I understand this would require significant modification to the water bridge, and other parts of the cooling system, but my question is strictly about the removal of the thermostat and mechanical pump.
Assume that the mechanical pump impeller has been removed, and the water bridge has been modified so there are only two ports:
(A) the combined outlet from both heads
(B) the inlet port on the block.
In this scenario:
- (A) is connected to the upper radiator hose
- An electric waterpump is installed in the lower radiator hose
- (B) is connected to the outlet of the water pump
- There is no mechanical thermostat
- The pump is controlled by an electronic temperature feedback system
- At ambient temperature, 6v is supplied to the pump for 10 seconds, every 30 seconds
- once the water temperature reaches 20 degrees less than the set point, it gives the pump 6v for 10 seconds every 10 seconds (this routine continues until the temperature reaches 5 degrees less than the set point)
- At 5 degrees less than the set point, the controller raises the voltage supplied to the pump until it reaches the set point
- Once the set point is reached, it keeps the temperature constant by modulating the voltage to the pump
Thanks for your help in talking through this hypothetical installation.
06-10-2009, 08:47 PM
#2
BMW went to a elec WP in 2006 on their 3-series cars, thats good enough for me.
Keep the stock thermostat setup. Build some kind of pulley/bearing/shaft assembly
to take the place of the oem pump. Mount the elec pump inline and wire it up to
your choice of adjustable controller.
Keep the stock thermostat setup. Build some kind of pulley/bearing/shaft assembly
to take the place of the oem pump. Mount the elec pump inline and wire it up to
your choice of adjustable controller.
06-10-2009, 08:49 PM
#3
Rennlist Member
You really need to keep the water moving at all times, even if it's just recirculating in the block as the existing system does when cold. Having a pump in the hose won't get you there. IMHO Porsche had good reasons for setting it up the way they did -- it minimizes hot spots, and the circulation warms up the block more evenly than would be the case with no circulation. I would think that retaining this functionality ought to be a priority.
06-10-2009, 09:16 PM
#7
Electron Wrangler
Lifetime Rennlist
Member
Lifetime Rennlist
Member
I was thinking exactly what Dave said.
However you will need to control the pump speed based on RPM and/or Temp Differential so you still need a controller for that. I think you'd want a matched controller & pump - no point doing more potentially destructiive R&D than you need to. I think the thermostat is a cheap and well proven solution so I'd still just use that method for temp control.
Alan
However you will need to control the pump speed based on RPM and/or Temp Differential so you still need a controller for that. I think you'd want a matched controller & pump - no point doing more potentially destructiive R&D than you need to. I think the thermostat is a cheap and well proven solution so I'd still just use that method for temp control.
Alan
Trending Topics
06-10-2009, 09:51 PM
#8
Rennlist Member
One maker of electric fans released an electric water pump here in Oz a few years ago(2000). Lots of people go 'Very interesting!'. A local web based car magazine announced a test/review of the device - this will be interesting...start reading. fitting it to a private import Mustang.....OOPS what this? Rest of the article withdrawn from publication due to threats of legal action by the maker...
There is mention of the original page and a link here
http://autospeed.com/cms/title_Respo...1/article.html
but the page is now blank.
Draw your own conclusions. The units have now multiplied and may be very different, but autospeed still will say absolutely nothing.
jp 83 Euro S AT 52k
There is mention of the original page and a link here
http://autospeed.com/cms/title_Respo...1/article.html
but the page is now blank.
Draw your own conclusions. The units have now multiplied and may be very different, but autospeed still will say absolutely nothing.
jp 83 Euro S AT 52k
06-10-2009, 10:37 PM
#9
Nordschleife Master
Hans thanks for posting this up.
We have been chatting back and forth about this information offline and it may be a good idea to run it by this group.
Some points for those out there.
When using an electric controller like the Davies Craig (DC) unit. A thermostat is not required at all. If one were to just run the waterpump at full voltage without a controller then yes one needs a thermostat.
So to make it clear this is a one or the other situation. NOT both.
Joe (90-S4),
It is impossible to retain the stock thermostat and run the pump where we want in a hoseline. If the stock thermostat is used then the stock pump location MUST be used. This then gets very complicated and imo impossible to keep under the stock TB covers for the ability to run a SCer.
As well Hans wants to eliminate the stock water bridge for his project. So again, it is not what is desired/will work.
Dave does bring up a very good point about always wanting to keep the water recirculating during warm up. I would prefer it to, but it is not 100% required and the controller will work to control the engines running temperature with issue.
Alan,
The control adjusts the pumps RPM/output via voltage. The hotter the coolant the more voltage it supplies. It adjusts the voltage going to the pump up and down to maintain your temperature to an adjustable point. So if you want your engine to run at 90deg C by varying the voltage of the pump it can pump just enough water through the engine to maintain that temperature (reason for no T-stat needed). So there is no other contollers that is needed for this.
The pump does not need to be matched to the controller the way the controller functions. But you do need to ensure that the electric motor on the WP is designed to be able to be run with variable voltage. Most brushless motors are able to do this without problems. Electric motors with brushes can burn the brushes with too low of a voltage being put through them from my research.
As to the cost of this conversion it would not be that much if the stock bridge was retained. It would be the cost of the DC temp controller which is a couple hundred US. Then it would be the cost of the electric water pump, whatever pump you decided to go with. From there you would remove the T-stat and use a plug found at say home depot to plug it. You would also then have to pull off the WP and take the impellor off the backside and replace. So this could probably be done realistically for around $500
But this is not fixing the bearing/seal problem of the WP. And this is where I will end that.
We have been chatting back and forth about this information offline and it may be a good idea to run it by this group.
Some points for those out there.
When using an electric controller like the Davies Craig (DC) unit. A thermostat is not required at all. If one were to just run the waterpump at full voltage without a controller then yes one needs a thermostat.
So to make it clear this is a one or the other situation. NOT both.
Joe (90-S4),
It is impossible to retain the stock thermostat and run the pump where we want in a hoseline. If the stock thermostat is used then the stock pump location MUST be used. This then gets very complicated and imo impossible to keep under the stock TB covers for the ability to run a SCer.
As well Hans wants to eliminate the stock water bridge for his project. So again, it is not what is desired/will work.
Dave does bring up a very good point about always wanting to keep the water recirculating during warm up. I would prefer it to, but it is not 100% required and the controller will work to control the engines running temperature with issue.
Alan,
The control adjusts the pumps RPM/output via voltage. The hotter the coolant the more voltage it supplies. It adjusts the voltage going to the pump up and down to maintain your temperature to an adjustable point. So if you want your engine to run at 90deg C by varying the voltage of the pump it can pump just enough water through the engine to maintain that temperature (reason for no T-stat needed). So there is no other contollers that is needed for this.
The pump does not need to be matched to the controller the way the controller functions. But you do need to ensure that the electric motor on the WP is designed to be able to be run with variable voltage. Most brushless motors are able to do this without problems. Electric motors with brushes can burn the brushes with too low of a voltage being put through them from my research.
As to the cost of this conversion it would not be that much if the stock bridge was retained. It would be the cost of the DC temp controller which is a couple hundred US. Then it would be the cost of the electric water pump, whatever pump you decided to go with. From there you would remove the T-stat and use a plug found at say home depot to plug it. You would also then have to pull off the WP and take the impellor off the backside and replace. So this could probably be done realistically for around $500
But this is not fixing the bearing/seal problem of the WP. And this is where I will end that.
06-10-2009, 10:56 PM
#10
Rennlist Member
Thread Starter
Colin,
Thanks for your reply. I think that removing the thermostat altogether from the cooling system is the easiest way to add a remote pump. Obviously, some type of temperature control is required, and the best option seems to be a variable pump controller. I guess the real concern is will the 30 seconds in-between pump pulses during the initial warm-up period be safe. I understand the logic behind the programming in the controller, as it allows the water in the jackets to heat-up to temp faster than circulating like the stock configuration. Since we would be using a similar temp as the stock (either the 80 or 85 degree setting on the controller), the pump will begin cycling almost constantly (10 seconds on, 10 seconds off @ 6v ~ half speed) at 60 degrees or so, I cant see how this could cause any harm, but wanted to pass it by the smarter guys in the room.
Anyone care to comment specifically on the warm up protocol of the controller?
Thanks for your reply. I think that removing the thermostat altogether from the cooling system is the easiest way to add a remote pump. Obviously, some type of temperature control is required, and the best option seems to be a variable pump controller. I guess the real concern is will the 30 seconds in-between pump pulses during the initial warm-up period be safe. I understand the logic behind the programming in the controller, as it allows the water in the jackets to heat-up to temp faster than circulating like the stock configuration. Since we would be using a similar temp as the stock (either the 80 or 85 degree setting on the controller), the pump will begin cycling almost constantly (10 seconds on, 10 seconds off @ 6v ~ half speed) at 60 degrees or so, I cant see how this could cause any harm, but wanted to pass it by the smarter guys in the room.
Anyone care to comment specifically on the warm up protocol of the controller?
06-10-2009, 11:07 PM
#11
Somebody said something about making a thick plate that had the shape of the WP profile
and installing a shaft with a very robust sealed (never sees water) bearing, with the pulley
on the end. Once done, why couldn't you utilize the existing inlets/outlets already on the oem
water bridge. In fact if you changed positions and pumped to the top of the motor and pulled/exited
from the block you'd have a reverse flow system. This sounds straight forward, what am I missing ?
I think this is all very do-able, hope we can keep this thread going..
Thanks,
and installing a shaft with a very robust sealed (never sees water) bearing, with the pulley
on the end. Once done, why couldn't you utilize the existing inlets/outlets already on the oem
water bridge. In fact if you changed positions and pumped to the top of the motor and pulled/exited
from the block you'd have a reverse flow system. This sounds straight forward, what am I missing ?
I think this is all very do-able, hope we can keep this thread going..
Thanks,
06-10-2009, 11:48 PM
#12
Chronic Tool Dropper
Lifetime Rennlist
Member
Lifetime Rennlist
Member
I'll throw out some thoughts here. My creds-- I own a 928, and also do a little work on heat transfer issues for boilers in power plants.
The existing system depends on continuous flow during the warmup period to avoid hot spots and promote even heating of the block, as Dave reminds us. The thermostat is closed to the radiator at this time, and the circulation is directed through the bypass port behind the termostat. Any system for the 928 will need to include this cold circulation capability if you want to use the aluminum block and heads. Keep in mind that many (most Ameican) cars bypass some of their flow around the radiator full time, including SBC for which the electric pump is an available aftermarket option. There's a short connection from the water jacket on the front of the intake back to the pump suction. The 928 system takes this s step beyond SBC (and the other similar systems from Ford and Mopar) by blocking the recirc flow when the engine is warm. That kinda rules out an in-the-hose solution unless the bypass port is permanently closed off, and some sort of recirculation capability is included on the radiator side of the pump system. You will functionally move the thermostat from tthe block to another bypass manifold. Might be cool to just build a radiator with this thermostatic bypass capability, by the way.
An electric water pump uses quite a bit of extra electric power. I haven't looked at current consumption numbers for the aftermarket pumps, but lets say for discussion that the pump motor generates one horsepower at full load. That's almost 60 amps just to run the pump. A 'smart' pump controller would have that at full power only at high loads and high RPM's, right? On a hot day, slogging in traffic with the AC and headlights on and coolant temp going up as it seems to, will the pump controller be smart enough to limit water flow increases and wait for airflow from the fans to catch up? If both go to full load at the same time, there will be almost 100 amps required. I'm not sure the existing 928 alternator will make enough power at lower RPM's to meet that demand, so you'll be dipping into battery reserves even mre than you do now with just the fan load. So plumbing issues aside, you'll need to plan on a much bigger alternator.
At some point you'll still need to put an idler pulley in the neighborhood where the current pump shaft is, just so you can use a known belt. Perhaps there's a shorter belt available and a way to guide the belt to a slightly different idler point, or get rid of that idler completely and still figure a way to keep the belt fully engaged on the two cam drives.
Those are my instant thoughts. Maybe more later after more discussion. I guess I'd be leaning towards finding the most robust commonly available bearing cartridge, and then get it on to new pump housing that has a boss big enough hold that bearing and seal setup. The bearing and shaft will likely be longer and stick out further in the front, so a deeper bell drive pulley would be needed to use the existing belt drive. That way the forward end of the bearing boss would be deeper inside the pulley, effectively. So what if the nose of the pulley sticks out through the existing plastic cover some...
The existing system depends on continuous flow during the warmup period to avoid hot spots and promote even heating of the block, as Dave reminds us. The thermostat is closed to the radiator at this time, and the circulation is directed through the bypass port behind the termostat. Any system for the 928 will need to include this cold circulation capability if you want to use the aluminum block and heads. Keep in mind that many (most Ameican) cars bypass some of their flow around the radiator full time, including SBC for which the electric pump is an available aftermarket option. There's a short connection from the water jacket on the front of the intake back to the pump suction. The 928 system takes this s step beyond SBC (and the other similar systems from Ford and Mopar) by blocking the recirc flow when the engine is warm. That kinda rules out an in-the-hose solution unless the bypass port is permanently closed off, and some sort of recirculation capability is included on the radiator side of the pump system. You will functionally move the thermostat from tthe block to another bypass manifold. Might be cool to just build a radiator with this thermostatic bypass capability, by the way.
An electric water pump uses quite a bit of extra electric power. I haven't looked at current consumption numbers for the aftermarket pumps, but lets say for discussion that the pump motor generates one horsepower at full load. That's almost 60 amps just to run the pump. A 'smart' pump controller would have that at full power only at high loads and high RPM's, right? On a hot day, slogging in traffic with the AC and headlights on and coolant temp going up as it seems to, will the pump controller be smart enough to limit water flow increases and wait for airflow from the fans to catch up? If both go to full load at the same time, there will be almost 100 amps required. I'm not sure the existing 928 alternator will make enough power at lower RPM's to meet that demand, so you'll be dipping into battery reserves even mre than you do now with just the fan load. So plumbing issues aside, you'll need to plan on a much bigger alternator.
At some point you'll still need to put an idler pulley in the neighborhood where the current pump shaft is, just so you can use a known belt. Perhaps there's a shorter belt available and a way to guide the belt to a slightly different idler point, or get rid of that idler completely and still figure a way to keep the belt fully engaged on the two cam drives.
Those are my instant thoughts. Maybe more later after more discussion. I guess I'd be leaning towards finding the most robust commonly available bearing cartridge, and then get it on to new pump housing that has a boss big enough hold that bearing and seal setup. The bearing and shaft will likely be longer and stick out further in the front, so a deeper bell drive pulley would be needed to use the existing belt drive. That way the forward end of the bearing boss would be deeper inside the pulley, effectively. So what if the nose of the pulley sticks out through the existing plastic cover some...
06-10-2009, 11:51 PM
#13
Nordschleife Master
Joe,
It was I who had talked about the plate and external bearing assembly. I am still considering it and I will let you know how I get on with it if you like.
If one wanted to not use the controller and wanted to use an external thermostat the 1978 BMW 325i had a remote thermostat in its own housing that can be purchased for around $30 give or take. The ports are big enough for the engines needs, it is 2 part cast aluminum piece that is crimped together, so when it fails you simply remove and replace, but this adds alot to the plumbing.
Hans,
Having the water stop circulation during the warm up cycle for 10-30 seconds during warm up will have no ill effects on the engine. Now if it was to do this at normal operating temperature, then yes that would be a problem and create heatspots. But when below operating temps it is not a problem.
It was I who had talked about the plate and external bearing assembly. I am still considering it and I will let you know how I get on with it if you like.
If one wanted to not use the controller and wanted to use an external thermostat the 1978 BMW 325i had a remote thermostat in its own housing that can be purchased for around $30 give or take. The ports are big enough for the engines needs, it is 2 part cast aluminum piece that is crimped together, so when it fails you simply remove and replace, but this adds alot to the plumbing.
Hans,
Having the water stop circulation during the warm up cycle for 10-30 seconds during warm up will have no ill effects on the engine. Now if it was to do this at normal operating temperature, then yes that would be a problem and create heatspots. But when below operating temps it is not a problem.
06-11-2009, 12:10 AM
#15
Nordschleife Master
Bob,
You posted while I was typing my reply, but touched on some points that I wanted to go over.
How could it be detremental for the aluminum block to have the water not pumping and just warming up quicker during the warm up cycling. Keep in mind this is only really at temps below 60C.? I do not see any way that it could cause damage, but I may be missing something so please expand.
Sticking out of the cover with an electric motor is not a problem for some. But there is limited space there from the factory and zero space there for those running SCers which are the most likely people to want a system like this.
As to the power issue. It doesnt require as much as you might think.
The CSR water pumps (chev, ford, mopar, generic etc.) all run with a MAX operating amperage of 13amps. They flow between 35- 70GPM, that is 132.5 - 265LPM. Depending on how they are set up. The big Davies Craig unit flows 115LPM for a comparison. For your setup go with the CSR 923, when I got mine I got the CSR 925, as they didnt have the 923 at that time. So I just welded up an aluminum manifold to bridge the two. Probably not the best but it will be fine for what I want.
And it is possible to simply remove the impellor off the current water pump (easist and cheapest) and use it to continue running as is.
You posted while I was typing my reply, but touched on some points that I wanted to go over.
How could it be detremental for the aluminum block to have the water not pumping and just warming up quicker during the warm up cycling. Keep in mind this is only really at temps below 60C.? I do not see any way that it could cause damage, but I may be missing something so please expand.
Sticking out of the cover with an electric motor is not a problem for some. But there is limited space there from the factory and zero space there for those running SCers which are the most likely people to want a system like this.
As to the power issue. It doesnt require as much as you might think.
The CSR water pumps (chev, ford, mopar, generic etc.) all run with a MAX operating amperage of 13amps. They flow between 35- 70GPM, that is 132.5 - 265LPM. Depending on how they are set up. The big Davies Craig unit flows 115LPM for a comparison. For your setup go with the CSR 923, when I got mine I got the CSR 925, as they didnt have the 923 at that time. So I just welded up an aluminum manifold to bridge the two. Probably not the best but it will be fine for what I want.
And it is possible to simply remove the impellor off the current water pump (easist and cheapest) and use it to continue running as is.