Cadillac CTS-V Forum banner

My Varimax vs Stewart W29 pump comparison!

31K views 84 replies 24 participants last post by  ruskiyab 
#1 · (Edited)
Preface:
We all assume more flow will equal more cooling, so I wanted to try to quantify "How much?" with some basic IAT2 monitoring as I have been running the Varimax for some time. I also have a BMR expansion tank and 3/4" lines throughout (hard lines removed), where I compared the benefits of a reservoir HERE. The system is: reservoir feeds the pump, which pushes into a track attack, and then straight to the ZL1 lid, back to the reservoir with the pressure cap in place. I typically run 30-50% coolant depending on the season (the reservoir tank comparison was 50-50 mix, this comparison will be 30% dexcool before and after as I drained the system this past spring).

Here goes: The Random84 / jb_bgg Stewart W29 intercooler pump installation guide:
Day 1:
I purchased the D3PE stewart pump kit for $189 + shipping, and it’s a nice looking kit: the 2-piece bracket is sturdy, powder coated, and they include hose clamps and (most) of the hardware, including two very nicely machined 1” to ¾” hose reducers — except my kit was missing two machine screws and after I hammered out the factory pump bracket from the subframe to make way for the noticeably larger W29… I realized it didn’t fit!

I have an aftermarket radiator, which likely is the reason why, but the short of it is the pump inlet was butting up against the bottom of the radiator and the outet hose kinked tightly against the radiator fan shroud when pushed into place. So I had to abandon that and hope that DP3E will offer a return / refund for my wasted day. After some ranting in an unrelated thread, jb_bgg was gracious enough to send me some photos of his solution, which I ultimately copied and expanded on here. Many thanks to him for his kindness!

Day 2:
I spent several hours test-fitting the position and location of the pump to allow for a 1” hose feed and return in the passenger-side fender well next to the washer fluid reservoir. I strongly encourage you to use cardboard to establish a template for your upcoming metal work, but all of this is pretty straightforward and can be accomplished with the standard assortment of power tools.

My template:


It took me around 2 hours to make the bracket out of a 12guage plate approximately 8 x 18” in size, $10 at the local big box store. A lighter gauge steel is recommended for ease of use. When determining where to drill your holes for the pump mount, be very judicious in focusing on orientation and location — you will need to be aware of the washer fluid pump line and hose clearance for the 1” inlet/outlet of the Stewart. You will also likely need to either remove the brake air duct behind the fog light or leave enough slack in the line to go under it when the front fascia is installed.

Test fitting (watch out for washer fluid hose and connector clearance):



The Stewart W29 has four mounting holes, using M8 x 1.25 pitch threads that are 15mm deep. You will also need to use spacers given the mounting surface of the pump — in my case I found everything I needed at the local Tractor Supply, I used:
(4) M8 x 1.25 x 25mm hex bolts with 3/8” tall stainless steel bushings (5/16 x 5/8 x 3/8), and locktite is recommended to keep things secure. **Only 15 ft-lbs is required.
(2) M8 x 1.25 x 25mm hex bolts with nylon locking nuts were used for mounting the bracket to the front bumper. You will need a 5/16" cobalt drill bit and plenty of patience to cut these holes. I used the existing rear fender bolt (M6 size) that holds the plastic fender guide in place for the rear of my bracket (see photos).
** Rubber bushings are optional, but account for their thickness when buying bolts — you may need longer 30mm bolts if you use locking washers and/or rubber bushings - IMHO the vibration is negligible and this is not necessary.
Lastly, I used
(2) 1 inch-to-3/4 straight hose connections along with a few feet of 1" hose to mate up to the Stewart.

Day 3:
After figuring out all of my fittings and having extras on hand, I realized I didn't need most of them. I installed the pump, bracket and wiring harness relatively easily: I used the +12 source from the fuse block, and the large negative ground right next to it for power - I also had some factory harness connectors and used the existing OEM pump harness to power the integrated relay via the Stewart. Everything is so close, I did not have to lengthen the provided Stewart pigtail for any of the three connections. Rerouting the hoses was also easy — and if anything is more accessible in the sub-lamp area with bumper removed, without kinking of hoses.

Wiring harness with in-line 25A fuse tied into +12 fuse block and ground:


I used the OEM reference signal (key-on signal) for the factory pump spliced into the Stewart harness with weather connectors (you can see my hose routed under the K member also):


The pump runs very quietly and I get no vibration (the fuel pump is louder). You WILL need to prime the pump, by making sure the inlet hose is completely full of coolant before applying power. There are several ways to do this, but if you find yourself topping off the system with everything buttoned up (like I did), then simply unbolt the pump from your bracket, lower it to the ground and give it a few moments.

Priming the pump by getting air out of the feed line:


Initial impressions:
Running IAT2s are easily ~5* cooler at speed, typically in the single digits (8-12* over ambient in 75-85* weather), whereas the varimax was always 12-20* over ambient with my setup and 30-50% dexcool. Although I plan on more testing, IAT2 during pulls from 45-130+ mph show modest but consistent reduction of around 4-5* compared to the Varimax, which aligns with other off-the-cuff impressions from a few days of driving.

Honestly, I was expecting a more significant IAT2 reduction during pulls, and have several 45-130+ mph runs for the purpose of monitoring this. Interestingly, the IAT2 decrease remains consistent (increases at the same rate as with the Varimax, just 4-5* cooler) - however when I increased capacity with the reservoir the IAT2 delta increased with time (the larger coolant volume was taking longer to heat up).

In all fairness, I suspect I am not utilizing the Stewart to its fullest potential: I suspect a potential restriction with ¾” lines elsewhere in the system; or perhaps a restriction with the style of reservoir tank I am using (BMR), or just cumulative resistance from multiple <3/4" transitions in the lines that are inevitable.



Conclusion:
In my opinion, for twice the price and a more complicated installation, I do not see the Stewart W29 as a great buy relative to a Varimax for a typical “under hood” installation utilizing a reservoir tank — Although from a big picture standpoint I gained greater IAT2 benefits from the Stewart, it was actually equal to or less of a change than from the reservoir tank alone: The reservoir / extra capacity netted me 4-5* lower IAT2 at the end of a single 5-6 second run only, whereas in this test the Stewart showed lower IAT2 across the board, but would be 3-4*F less under the same conditions (ie a single gear) that I used to test the reservoir tank. Only when I expanded my test over a longer 45-130mph pull did the delta really become appreciable, given the potential error in my ad hoc testing - and that's what I've shared here.

I think the Varimax will continue to be the standard for a cost-effective Plug-n-Play upgrade. The “hopped up” Stewart version is likely more effective, but also significantly more money with unknown durability. At the end of the day, by the time you buy a Stewart W29 and a reservoir tank of your choice, you’d be better off with a trunk tank specifically for the four-fold increase in system capacity — all you need are a few extra feet of hose IMHO. It seems that “capacity” is the primary driver of cooling effectiveness in our cars: the more volume you have, the better off you are assuming that the heat generated is the same, since the Supercharger is the primary heat source while in boost (and heat dissipation is mostly limited by your condenser efficiency).

I also feel that, with most of the available reservoir tank designs, the 12AN fittings and various filtering devices intended for using ice are limiting the pump feed along with the necessary use of ¾” hose and fittings through most of our intercooler system, and likely reducing the effectiveness of higher volume pumps like the Stewart.

I do plan on converting to a trunk tank setup in the near future to see if this theory holds up. 
 
See less See more
7
#2 · (Edited)
UPDATED I actually posted in the wrong thread...

Here's my trunk tank data:

First test: using the same "standard speed" Stewart pump from my underhood tank setup last year, I compared IAT2 to the new trunk tank setup - which has roughly 4-5x the water volume as the underhood setup. Pump speed was the same - HOWEVER remember that total RESISTANCE to flow has increased dramatically. This means the fluid velocity was likely noticeably slower with the trunk tank setup. After all, it takes ~40 feet of 3/4" lines to run from the trunk to the front of the car - and yet, no major difference in the IAT2 curve (accounting for gear changes) or in maximum IAT2 observed after a run:


Second test: I compared the standard speed Stewart to the MAX speed repogrammed stewart in only the trunk tank setup. I purposefully used a test where the MAX speed pump had the disadvantage of a +6* F ambient temperature - yet the IAT2 curve was flatter and the peak IAT2 temperature was lower! Clearly, the MAX speed pump was able to overcome the additional resistance noted above. As other members have mentioned, the primary way heat is expelled from the intercooler system is the temperature delta of the water in the heat exchanger relative to ambient air temperature. Faster flow = more heat dissipation:


Third test: Next I compared the MAX speed trunk tank to some data logs I have from the original "standard speed" underhood setup. The trunk tank wins... BUT that's likely due specifically to the additional fluid velocity (beyond that of an underhood tank!); after all, I previously showed that markedly increasing total fluid volume by itself didn't do much:


Last test: I added the underhood tank / STD pump data with the trunk tank / STD pump run, and then the worst-case scenario MAX speed pump run in one final graph, to show that by my best ability to measure, the primary benefit to all of this shit was a stronger pump:


You can see that the IAT2 curve is notably more blunted, and does not peak as high, as the other tests. I think it's fair to say that although there are other variables, the MAX speed reprogrammed stewart pump is the single largest contributor.

-----------------------------------------------------------------------------------------------------------------

Upcoming tests, weather permitting:
Hose routing: I will be switching to the "alternate" hose routing outlined in my first post - so that the hottest water from the intercooler brick goes to the intercooler, maximizing the "delta" of coolant temperature to that of ambient air. That should hopefully reduce water temps in the tank and increase heat shedding.

Additives: After this (or maybe at the same time), I will be adding ~3 bottles of water wetter to the system, and seeing if that makes any appreciable difference. After all, my trunk tank data so far is all using straight water, but the "underhood tank" data had some dexcool and 1 bottle of water wetter in it, and it's possible that accounts for a few degrees in IAT2 measurements.



RECOMMENDATIONS:

At this point, for simplicity and cost, my recommendation to anyone without a coolant reservoir is this: an underhood tank of your choosing (Norcal, BMR, D3PE, etc), a reprogrammed Stewart pump, and use 1" hose as much as possible. A larger aftermarket heat exchanger is likely beneficial... but I never tested one by itself. !) Lastly, I would say the ZL1 lid is still helpful, but I would put it last on the list due to cost.

The trunk tank will be useful for drag racers specifically for allowing the maximum volume of ICE - but for a daily driver I do not recommend it. It's slightly more costly, but the length of hose needed to go from the trunk to the engine then back again is significant, and it's very challenging to use 1" hose. So you're adding what I believe to be significant resistance in the form of 3/4" hose over ~40 feet that is just not needed for a daily driver.


Lastly, understand that there IS variability in my measurements. Your mileage may vary - and there is quite a bit of variability in my testing, much less from one person's car to another. I've done my best to mitigate variables but at the end of the day it is what it is. I recommend taking other members' results with a big grain of salt as we're all using different combos, coolant ratios, hose sizes, etc; so saying one guy sees "___ over ambient" is likely not an apples-to-apples comparison with my combo and my testing.

And I am limited by my resources, my time and my patience, so this is not the be-all for cooling system efficiency, BUT I think it's a pretty good groundwork. IAT2 is NOT a direct comparison to power, nor is it a direct equivalent to literal "air temperature" as the air is heating flowing water that is being read by my IAT2 sensor - but it's as useful a approximation as I can come up with for these purposes.

Enjoy!
 
#3 ·
Have you verified that your pump is programmed to run at full speed?




_________________________________
Contact me for
•Modified lids • Lid customization services (brick reinforcement, milling, -12 AN manifold conversions) for both ZL1 and CTS-V lids • Billet aluminum V badges • Go Fast Parts • More products coming soon!
 
#8 · (Edited)
This is the "standard" Stewart pump. I did not spring for the "programmed" higher speed / higher amp pump, however that may help the results slightly.

Let's face it, the Stewart should be spinning significantly more than the Varimax based on Lingenfelter testing (and others), yet the results? modest at best. So getting the amped up version *may* help slightly, clearly having almost double the pumping volume is not translating in a marked real-world reduction in IAT2.

Is the Steward better? Absolutely, no doubt.

As good as I expected? No.

Presently, increasing your coolant capacity appears to be the priority, with pump speed being less significant. Both improve cooling efficiency, but if I had to place a priority on one over the other, it would be more coolant volume without a doubt.

In other words, spend your money on a larger tank instead of a faster pump.
 
#5 ·
It's made by EMP, if it is indeed the one that Stewart carries, I don't believe it's programmed to run at full speed.




_________________________________
Contact me for
•Modified lids • Lid customization services (brick reinforcement, milling, -12 AN manifold conversions) for both ZL1 and CTS-V lids • Billet aluminum V badges • Go Fast Parts • More products coming soon!
 
#6 ·
Nice!

One thought, if you re-routed the flow it might improve the performance of the system a slight bit. I would suggest the pump send the water to the lid first, then the hot water would route to the radiator and then back to the tank. This will send the hottest water to the radiator and ultimately transfer heat at a greater rate to the outside air.

Thanks for sharing your hard work, we appreciate it!!!
 
#10 ·
Who wants to trade me then? :D


(but, of course there is also a "massive" difference between the standard Stewart and the Varimax pump... )
 
#13 ·
0.9 gpm difference in a ZL1 intercooler circuit according to Lingenfelter testing.




_________________________________
Contact me for
•Modified lids • Lid customization services (brick reinforcement, milling, -12 AN manifold conversions) for both ZL1 and CTS-V lids • Billet aluminum V badges • Go Fast Parts • More products coming soon!
 
#16 · (Edited)
Someone asked me which pump I am using, and since there is some doubt as to my results I wanted to post a photo here. I bought it new from Karch back in February - but again I'll reiterate that I suspect my reservoir tank and fittings are creating a noticeable restriction, but since most people will have that problem with the typical setups, it's still a valid test IMHO.



Anyone know what the "reprogramming" entails? If it's as simple as taking off a cover and moving a jumper around....
 
#17 · (Edited)
Someone asked me which pump I am using, and since there is some doubt as to my results I wanted to post a photo here. I bought it new from Karch back in February - but again I'll reiterate that I suspect my reservoir tank and fittings are creating a noticeable restriction, but since most people will have that problem with the typical setups, it's still a valid test IMHO.



Anyone know what the "reprogramming" entails? If it's as simple as taking off a cover and moving a jumper around....
It's not that simple, it has a controller attached as you can tell, you connect to it via CAN or serial and use the EMP tool to reprogram the max speed. Now, yours may already run at that of its not an actual Stewart pump. One way to check would be to measure the current draw. If you're seeing close to 20A or more, it's safe to say it's running full speed.

As for less restriction, that'll help any pump.


_________________________________
Contact me for
•Modified lids • Lid customization services (brick reinforcement, milling, -12 AN manifold conversions) for both ZL1 and CTS-V lids • Billet aluminum V badges • Go Fast Parts • More products coming soon!
 
#22 ·
i dont think the reprogram will help much either. from what i recall it is limited to pull a max of 25 amps now and would roughly up that too 30-35 ~ amps.

The 3/4" lines are what the issue is , you need an entire 1" system including h/x , reservoir and core. More restriction/head pressure + heat with making pressure.

eliminate bottle necks, and you can expect a pretty great gain in iat and hp .
 
#23 ·
The Stewart pulls about 9A while the reprogrammed about 19A according to Lingenfelter. You can't really do 30-35 through the 12 AWG wire going into that connector, full speed would be 25A.




_________________________________
Contact me for
•Modified lids • Lid customization services (brick reinforcement, milling, -12 AN manifold conversions) for both ZL1 and CTS-V lids • Billet aluminum V badges • Go Fast Parts • More products coming soon!
 
#26 ·
Yes. The 12AN are essentially 3/4" ID. However, as some of the more technically inclined guys will point out, EACH restriction is cumulative. So having four 3/4" restrictions plus the heat exchanger fins plus the brick/intercooler, plus any bends, etc all add up to more resistance than a single 3/4" fitting.

So most of us who use the ZL1 lid and a typical track attack or other heat exchanger are all going to have higher resistance than a true 1" all-the-way-through setup. That's exactly why the EMP Stewart pump was thought to be desirable: more power to overcome those inherent restrictions relative to the OEM pump or even the Varimax.

FWIW I will be having a custom trunk tank made and running full 1" lines to the heat exchanger and ZL1 lid - but I don't expect the weather to hold out long enough for a useful comparison... nor do I expect it to make a huge improvement in the flow of the pump itself. I expect the gains to be primarily from having four-fold (or more) fluid capacity: more water takes longer to heat up (and to cool down).
 
#28 ·
Starting from scratch, I think the best bet is a trunk tank with 1" outlets to a Stewart, 1" lines and then teeing down to the heat exchanger and lid (with 12AN fitting conversion), then back to the tank.

That's my best guess at this point.

Sent from my SM-N920V using Tapatalk
 
#30 ·
When I do a fluid flush and refresh I just use a little suction on the line from the HX to the lid to prime the pump. Re: flow between regular and max RPM there is a pretty significant difference, and Lingenfelter can reprogram. Otherwise you have to buy the interface cable from Stewart, download software, etc. etc. Much less hassle to pay the price.
 
#34 ·
even if everything is 3/4" including the lines too the zl1 lid , the i/d of the F manifold and brick fins is not. The only real way to significantly improve cooling is to up everything too 1" hose and heat exchangers with equivalent flow.
 
#36 ·
~9 gallon trunk tank on order! <Giggles>

Sent from my SM-N920V using Tapatalk
 
#37 · (Edited)
I did some math yesterday or whatever...on pressure drop of various size hose.

What I came up with was at 5gpm a 5/8 hose would have just over 2psi pressure drop in 25ft...a 3/4 hose was just under 1psi pressure drop in 25ft.

As flow increases...the pressure drop curve is pretty steep in any case. Closer to 10gpm things are ugly and 3/4 or 1" pretty well becomes mandatory.

Also removing the screen in the norcal makes a tremendous difference. If you run ice...add a piece of mesh to the entire bottom of the tank and then add your ice.
 
  • Like
Reactions: random84
#38 · (Edited)
General rule, 2x the flow requires 4x the deltaP or pressure drop.

So you want to double the flow thru your existing system? it requires a pump that puts up four times the deltaP at a flow rate that is twice what your current pump is putting out.

Sorry to burst the bubble of all the non-engineers out there but this is a reality that even the mighty internet forums/FB/etc cannot violate lol
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top