electric water pumps and turbo coolant flow
 

I've seen a lot of people install EWPs all in the same place (in the lower radiator hose).

I put mine in the same place, but I'm concerned that putting the EWP here moves the turbo coolant return from the suction side of the water pump to the pressure side of the water pump...

does this reverse the flow of coolant through the turbo? or stop the flow all together?

I've read of people having to put smaller booster pumps in heater lines to get them to properly move water after an EWP install... I don't see how the turbo is any different which has me worried....

anyone else look into this?

You are correct in that the coolant flow path will alter by moving the water pump the lower radiator hose.

In order to flow properly, any coolant circuit should be routed with an entry downstream of the water pump and the return upstream of the pump.

You will see that in the factory setup on all 13B and 20B turbo models, the turbo coolant return line is plumbed back into the lower outlet of the water pump housing via a barb fitting that points up directly towards the pump.

Ditto with the heater. In older Mazdas and FC models, the heater return line is on the lower radiator outlet. In the FD the heater return is also in the lower outlet of the water pump housing.

Now assuming you remove the thermostat and block up the bypass, then move the water pump to the lower radiator hose.

In my view it won't reverse the flow through the turbo - both entry and return are downstream of the pump. With the direction of the return fitting, what is likely to occur is you will get a very small amount of flow through the turbo circuit, based both on convection and negative pressure as the coolant is pulled through the engine and then down through the radiator. But the flow will not nearly be as much as if you routed the turbo coolant circuit to return to immediately upstream of the EWP. That's where the return should go. Same as the heater return. With that setup, no booster pump (EBP) is necessary.

Without an EBP, heater performance will suffer during warm-up times due to lack of flow, but this is somewhat offset by overall faster warming up times.

To hook this up, it's all quite simple. Davies Craig make a fitting that you can stick on the inlet of the EWP. OR you can tap a heater return into the lower radiator outlet. Stick in a T-piece and you can also re-route the heater pump. This will restore the factory coolant flow path.

The Davies Craig fitting is depicted below.


http://daviescraig.com.au/media/183/...5531020.s8.jpg

See: EWP Heater Return Kit (8380) -

Now, I've been using EWPs since they were first released in Australia in the early 2000s. After trying different setups, the best one I've found for a regular street car is completely different to what anyone else has done.

Basically, I have kept the bypass thermostat system. The problem with that is that, like the heater and turbo return, the bypass return is integrated within the pump housing itself. There is a channel from the bypass port back into the upstream side of the factory water pump. This means the bypass won't function properly and the coolant path will flow straight up the lower water pump housing outlet and directly out the bypass, minimising the flow path through the engine.

My solution was:
1. weld up the bypass return channel on the water pump side.
2. drill a hole from the outside of the water pump housing to tap into the bypass return channel
3. tap a thread into the newly drilled hole
4. screw in a barbed fitting
5. run a new return hose from the barb (thermostat bypass) and route it the heater return line (which on my series 3 RX-7 runs to the lower radiator hose)

Done!

This way I can run a factory thermostat with all the benefits. With the combination of Davies Craig controller and thermostat, temperatures are rock solid, warm up times are way faster than factory, heater performance is the same as factory and I get the benefit of the 3 minute run on with the EWP from the Davies Craig controller.

I did a write-up on this in my AusRotary build log years ago: AusRotary.com ? Login

Unfortunately, I hosted the images on ImageShack and they removed them all. I'll dig them back up again at some point.

Now, this is not as compact or lightweight as the simple inlet/outlet flange setup I see most people use. But without a thermostat, I found that temperatures were just too cool especially at nights on the freeway where temps would often drop back down from 85'C to 70'C in a few minutes - this is even with minimal flow from the EWP. You need that coolant circulation through the block to maintain and evenly distribute heat IMO.

that's what I figured thanks.

I'll just plumb the turbo into the feed pipe for the EWP...

I'm having a hell of a time picturing what you did with the thermostat. pictures would be awesome. It would be really nice to be able to maintain the thermostat functionality mainly for the overcooling possibility.

Here is my current setup (not quite finished obviously) and yes, it's an FC... I posted this in the FD forum because a lot more FD guys use EWPs.

https://lh3.googleusercontent.com/1G...=w1679-h944-no

https://lh3.googleusercontent.com/cs...=w1679-h944-no

https://lh3.googleusercontent.com/nj...=w1679-h944-no

Looks good to me.

I have a series 5 setup too. Just plumb the turbo and heater return (if you still have heater) into the feed pipe for the EWP as you suggested. Easy.

For my thermostat setup, I'll try to explain what I did using your pictures. You have plugged the bypass channel at the top. You need to do that when deleting the thermostat. So it's setup correctly for a thermostat-less operation. But with my setup, I've left it open at the top where the thermostat sits and blocked the bypass channel from the BOTTOM inside the recess where the factory water pump impeller sits. Now, the idea then is to get the coolant to run out of the water pump housing into to the EWP whenever the bypass is open (ie thermostat closed). To achieve that, I've tapped a fitting into the bypass channel and run a hose from that to a heater return inlet pre-EWP. I've quickly tried to do some annotations in the attached photos.

I think I got it.

How did you block off the bypass channel? Weld in a plate? I don't remember what it looks like...

it's a narrow tunnel basically. I had it welded up with aluminium filler. It's a contoured shape and difficult to access, so I couldn't really stick in a welch plug or something similar to what you have done.

Some more info on this.

The other option I've long considered is to use a remotely thermostat housing mounted in-line in the top radiator hose.

Moroso and a few other companies make universal remote thermostat housings eg Moroso Remote Thermostat Housings | JEGS The problem is that these lack a bypass and won't work properly on a rotary engine layout (that I can see). I think some OEMs use remote thermostat housings with bypass incorporated. I researched this years ago - maybe a MR2 Spyder and a Jeep?

If you could put a suitable aftermarket market product, then you could replace the entire water pump housing with the inlet/outlet flange like many race cars do and still have a thermostat controlled system.

If anyone is aware of such a product, I'd be keen to have more details.

Not exactly what you describe, but this company machines 3 holes into their thermostats to provide closed-condition flow.
http://www.stewartcomponents.com/index.php?route=product/category&path=61

That's a common practice in racecars to stabilize coolant temperatures and minimize trapped air. I've been doing that forever...

OP your oil fill thing is missing a nut or a bolt

Thanks guys. I hadn't seen those Stewart thermostats.

I have heard of drilling holes into the thermostat however. In fact, Davies Craig recommends that for using a conventional non-bypass thermostat when installing an EWP you drill one or two 5mm holes to allow proper circulation. It could work well in combination with the controller, but I still think the idea of recirculating warm coolant is a better option to reduce warm-up times, if possible. An elegant off the shelf solution may not exist.

OK. I managed to find the OEM solution again!

I was close. Not a Jeep and MR2 but a Land Rover and a Lotus Elise both of which use Rover K Series engines.

Land Rover Freelander 1.8L K series used a remote thermostat housing with a bypass: This is the Land Rover Discovery version (silver): Land Rover Discovery 180F Thermostat PEL500110
Similarly, the same Rover part was used in the Lotus Elise: https://www.eliseparts.com/products/...ef-thermostat/

https://www.eliseparts.com/img/up/e5/1_96643fd.jpg
https://www.eliseparts.com/img/up/ed/216-0c8c1.jpg

The second grey one is meant to be the improved design with better flow balance to prevent thermal shock when the thermostat opens. It comes with thermostat installed with an 82'C opening temp - factory spec for a rotary.

ABS plastic housing is not ideal though. We know what happens with the Mazda AST. But it not really an expensive service item and Rover parts are available everywhere.

I think this would work.

So, you could:

1. replace the entire water pump housing with an inlet/outlet flange and side mount the alternator. Or you could keep the alternator top mounted with the product Turblown is developing.
2. install the K series remote housing with 82'C stat in line in the hose from engine to top radiator inlet
3. run a bypass hose from the bypass port to a heater return fitting in the EWP inlet.
4. into that bypass hose, you also plumb the turbo return hose and the heater return.

When the thermostat is closed, coolant will run from engine directly to EWP, bypassing the radiator.
As the thermostat opens, the bypass also closes, gradually introducing cooled water passing through the radiator.
Once the thermostat is fully open, EWP is pumping according to engine requirements hitting full speed when or shortly before the radiator fans turn on.
Under all conditions, the turbo and heater return circulate, as this coolant path does not pass through the thermostat housing - it goes directly to the EWP.

Who wants to try it first?

Upon further digging, the factory Freelander / Discovery has a thermostat opening temp of 92'C - too high for a 13B.

Rover released a lower-rated t-stat for warmer climates that opens at the more useful operating temp of 82'C. Part no. is PEL500110.

Here's a link to a sample part: LAND ROVER DISCOVERY 2 1999-2004 THERMOSTAT & HOUSING 82C / 180F PEL500110 | eBay

Actually, this is designed to fit into the lower radiator hose. It would still work in the top, but you could also set it up on between the radiator and EWP like this:

Attachment 605272

The "from head" label is coolant from the outlet from the engine. You would need a T-piece in this hose, which would then also run to the top radiator outlet.

Here's how the factory system looks:

http://www.rimmerbros.co.uk/ItemImag...GRID001582.gif

Going off this diagram:

(3) goes to engine outlet
(8) goes to EWP inlet
(12) goes to radiator lower hose
(23) goes radiator upper hose

Even though it on the radiator outlet with cool coolant, the pump will circulate hot coolant around the thermostat to ensure proper opening temperature.

Some useful explanation is given here:

Am I just being a bit thick? 1.8T cooling system - Page 5 - The 75 and ZT Owners Club Forums

Finally, if you want something sexy and aluminium, these guys appear to make just about any inlet/outlet/bypass size configuration: Custom Remote Thermostat Housing

OK last one I promise and apologies for the multiple posts. Partly this is for more own reference later and partly to share the information to save others on the research.

BMW 2002 E10 and Fiat 124 also use a remote inline thermostat with bypass, but both are aluminium not plastic and therefore appear superior than the Rover ones. The Fiat is the best looking but more expensive. Not sure on thermostat options.

The BMW comes with 71, 75 and 80'C thermostat options and is dirt cheap. Found one brand new for US$16 without much research: http://www.autohausaz.com/pn/11531253249

Score!

http://www.autohausaz.com/images/11531253249.jpg

Part number for the 80'C t-stat is 1153146805680. For Australian players, Tridon makes an aftermarket version with 82'C thermostat (same as Mazda), Part no. TT260-180P - there are couple of other variations made by Tridon with different inlet/outlet sizes and shapes. Worth looking at Tridon catalogue.

What a great thread for anyone with a EWP! I will be giving the Fiat 124 remote thermostat a try. Should fit what I have going on perfectly.

is this really an issue or are we overthinking and complicating things needlessly?

does it mean that Turblown's or SBG ewp kits are incomplete and may cause turbo bearing to overheat?

I'm adding one of the ewp kits to my build and now I have this to worry about? i thought it was set it and forget it...

Turbos don't need to be water cooled at all.... most early ones were just cooled by the oil going through them. But water cooling is much better and it suggested by turbo manufacturers to extend the life of the turbo bearings...

As far as I can tell, with both Turblown's and SBG's kits they keep the stock turbo outlet location which means there is minimal coolant flow through the turbo with the revised pump location.

turbo life will be affected, how much? no one knows. but rotaries don't really run cool... there is a reason our stock turbos were water cooled.

For me it has nothing to do with water to the turbo as mine is oil cooled only. Car runs way to cold. Especially on the freeway.

i had something custom made. If there is enough interest I can see about a group buy. I will post a pic tonight.

Exactly this. Without a proper bypass there is also risk of thermo shock and possible engine damage from the EWP going from pulse mode to full operation there is a great temperature differential between the radiator and the block.

A proper bypass also ensures proper heater operation and turbo cooling as well.

Just make sure the Fiat temp rating is ok. The BMW and Rover have suitable alternatives but i couldn't find a rating for the Fiat. A lot of piston cars have 190-195F thermostats which is too hot for a rotary.

edit: according to this it's 180F - perfect!

http://www.partsgeek.com/mmparts/the.../fiat/124.html

The configuration is like the Rover. You fit in the bottom radiator hose before the EWP then create a T in the top radiator hose to feed the bypass. Here's how the factory setup looked:

Thanks for the video. Lets me see how this is going to work much clearer. Getting rid of the water pump housing is going to be nice.

Would this be a good flow path?

I deleted my heater core..

From my Rear iron to the feed of my turbo (old heater core line)
The return of my turbo back into the hot side of my radiator

So the bypass thermostat helps with warm up and too cold situations. But what about the turbo and heater core coolant flow, how is that solved?

Both of them just have to be moved from after the water pump to before the water pump.

Yep. The Davies Craig inlet adapter with 19mm barb fitting is perfect for the return.

In the Lotus/Rover/Fiat/BMW style pre-EWP thermostat setup i described, the heater and turbo return would fit between the thermostat and the EWP.

I plan to create a diagram to show how it would all hook up.

Moving them to the pump inlet bypasses the radiator so you'll be recirculating hot coolant from the heater and turbo...

Correct and that's exactly what you want and exactly how it is plumbed in factory automotive cooling systems.

In order to provide proper flow through the heater and turbo circuits, the feed has to be downstream of the water pump (whether mechanical or EWP) and the return has to be upstream. If you make the return upstream of the radiator, then:

(a) in a thermostat-less EWP system, it will work ok but those circuits will be passing through the radiator all the time. This will add to the problem of overcooling and diminish heater performance. The heater core will itself actually take heat out of the system and therefore will be introducing coolant into the radiator at cooler than system operating temp. I foresee less of a problem with the turbo coolant circuit in less scenario but the flow through that circuit won't be as good as if the return went straight before the EWP.

(b) in a top thermostat EWP system (like my current factory position bypass thermostat setup), the thermostat will cease to properly regulate system temperatures, because a proportion of coolant flow will always pass through the radiator irrespective of whether the thermostat is open or closed. This will also significantly increase warm-up times, overcooling and reduce heater performance.

(c) in a bottom thermostat EWP system (like the Lotus/Rover/Fiat/BMW style), the heater and turbo coolant circuit won't work at all, because when the thermostat is closed, the return is no longer upstream of the EWP and consequently that coolant path will close.

By bypassing the heater and turbo coolant circuits, you are essentially equalising the temperature of those circuits with engine coolant temperature. The heater and turbo coolant circuits won't overheat provided that overall system temperature doesn't. Recirculating the coolant through the heater and turbo won't actually decrease overall cooling efficiency of the system. This is because under either scenario all factors affecting temperature reduction across the radiator remain equal (water pump flow, inlet/outlet diameter, inlet/outlet hose diameter, airflow, core design, core surface area and thickness, conductivity of material etc). It's not like you're pumping purely hot coolant either - the bypassed coolant, mixes in with the cooled coolant from the bottom radiator hose.

Are you using the Davies Craig EWP LCD controller?

Yes.

Solution for EWP
 

Here is my custom made setup. I am going to open a group but interest thread on these to get an improved version produced to shrink the footprint, and make some other design changes I feel are necessary. They will have the option of fittings or no fittings, and will be something like a -20 AN o ring for flexibility. It uses a GM thermostat, and completely bypasses when cold. This would be between the pump and the radiator.

https://cimg7.ibsrv.net/gimg/www.rx7...d2ad1036c6.jpg
https://cimg8.ibsrv.net/gimg/www.rx7...43bf34f16c.jpg
https://cimg9.ibsrv.net/gimg/www.rx7...f61d18e984.jpg

I can't quite figure out how that works but it looks amazing! Looking forward to seeing more details.

looks good. any pictures of it installed?

The basic function is: water pump forces water towards the thermostat. If thermostat is closed, the water goes across the block internally and bypasses the radiator. If the thermostat is open, the bypass is plugged and the water is forced through the radiator.

Anywhere in between is normal thermostat modulation. This allows for continuous flow during warmup, avoiding any hot spots, whilst retaining the ability to warmup the engine and related fluids quickly.

My goal with this deaign is to allow for the use of an EWP with continuous flow during warmup, a fast warmup like factory, and then basically direct flow once up to temp.

The internal passages are all 1" diameter. The big question is what kind of a market is there for it?

I am currently working with a friend of mine who is a badass machinist to make the new and improved version.

looks like an oil thermostat, same working principle right?

pretty much. But with a replaceable thermostat. Chose GM style with bypass blockoff plate so people could choose what temp they wanted along with being able to service without entire unit replacement.

I'm really interested in going EWP route. Do you guys have pictures of your setup installed with the BMW/FIAT thermostat?

Also, did anyone try going the route of using a PWM output with SSD relay to control the pump? If you did, any feedback?

Hopefully I'm not thread jacking.

I will be trying the PWM output with a SSD, should have useable info by summer.

I will try to create a drawing tonight that takes into account rerouting turbos and heater return to before EWP, and with remote thermostat from a fiat and see if its correct, KYPREO

Another pump to consider is the pierburg CWA400 EWP. This pump comes off of the bmw 335i. It flows a ton of water and has a built in solid state relay. Also you can get a controller from this company Tecomotive - tinyCWA that will control it properly as well as run the pump after shut down and control an electric fan after shut down as well.

https://www.auto-medienportal.net/bi...d/detail/24283

the CWA400 pumps out 150 L/m which is the same as the davies craig pump. it's cool that there is another choice though.

to follow up on this, I finished up my "fixed" EWP turbo drain plumbing last night

just welded a -6AN bung to the coolant pipe that feeds the EWP and extended the turbo drain line. (which was already -6AN)

https://lh3.googleusercontent.com/9E...=w1679-h944-no

Not sure if right thread, but has anyone verified the temp calibration on the Craig Davies EWP and the stock water temp sensor?
The location of my EWP temp sensor is on the outlet neck (where the cap goes and where you install the thermostat). Same as this one here:
https://cdn3.volusion.com/wpyhq.jsay...jpg?1459350630
So essentially ~4" from the stock temp sensor on the back of the water pump. You can see both sensors.
The thermostat is bypassed.

The EWP sees higher temps, about 20 degrees more or so.
I am using the stock water temp sensor and calibration on the adaptronic.

Earlier today testing, below were the temps
EWP water temp 200*F
Stock water temp 180*F
Oil Pedestal Temp ~160*F
Ambient Temp ~60*F

The problem is the EWP alarm goes off at about 200*F, before my fans even come on.

Bringing back an old thread. I was wondering if you found out any information regarding the difference in temperature from Stock to EWP. I just finished installing my EWP on my car, and have the water temp sensor at the fill neck as well. I decided to remove the thermostat, due to concerns mentioned above. The car isn't running yet so I'm not sure if there will be an over-cooling issue or not yet. I have the EWP connected to solid state relay controlled by a AEM infinity. One concern that I have is that I hooked up a variable DC power supply onto the EWP and ran it to help bleed the system. I don't know if it's just me but just visually inspecting the water from the fill neck, it appears that the water isn't moving that much. Perhaps someone can give some insight on what I'm seeing.
https://cimg6.ibsrv.net/gimg/www.rx7...f7d3bb1df3.jpg

I've just installed the Sakebomb EWP kit and am over cooling, 15 minutes on a highway at 70 mph and the car drops to 160F. So far I have left the LCD controller set to 85C (default at power up), I will try this increased to 90C but as this now also controls the temp the fans kick in there is a limit to how high I want to go.

In regards the turbo cooling flow, would a simple mod be to T into the drain from the AST at the bottom of the radiator? This would bring the turbo return up stream of the EWP.

Yep, that sounds fine. Tapping the turbo return anywhere upstream of the EWP that draws from the bottom of the radiator will work well. I am using sharing it with the heater return line myself.

This alone probably will not solve the over-cooling problem. To fix that you need to incorporate a thermostat. This will then allow you to set the controller to 85C with the thermofans kicking in at a more sensible temperature. Personally, my preference is to use a separate controller for the fans eg aftermarket ECU so that the switch on/off temp can be configured with precision separately from the EWP target temp. The controller is great and the algorithm works well to get the engine up to operating temp quickly (when used in conjunction with the thermostat) but unfortunately the thermofan control is tied to the EWP temp which isn't ideal in my opinion.

Yes I agree with your thoughts on pump and fan control.

I reached out to Davies Craig on it, I'm surprised the controller is so restricted, I doubt it would cost any more to have user programmability for different pump duty cycles and fan switch points. Davies Craig don't have any intention of offering this and their recommendation was to re-install the thermostat with a couple of 5mm holes drilled in it.

That is an OK stop-gap solution but requires modification to the thermostat to remove the bypass section and manually block up the bypass channel. Otherwise coolant flow will pass straight across the front of the engine and not through it. While going to that effort, it is better to incorporate a proper bypass thermostat, as this (when used in combination with the controller) gives you faster-than-factory warmup times and rock-solid temperatures.

I'm not too fussed about the fan control because pretty much every aftermarket ECU offers that. Otherwise, you can get programmable thermatic switches which will do the job. Davies Craig actually makes a cheap standalone one. I used this for years without issue. It switches off the fan 5'C below the set-point, which you can set wherever you want it.