Gambit secondary radiator kit

[Gambit]

Hi Y'all,
I’m at it again and wanted to share another open source style kit. I’ll call it “Gambit’s FL5 secondary radiator kit”. DIY and build it yourself and hopefully improve and make it better. Lots of inspiration from Driven Crazy and their Hyperthemia kit on the FK8.

Again the benefit of this DIY kit for the community is that all parts are available on ebay or universally available online. You can mix and match whatever parts you have spare such as silicon or rubber hoses, clamps, pipes, brackets etc to make it work and all at a very cost effective total price.

This is the kit installed in front of the primary radiator. I have not had to cut or drill any additional holes for the mounting points, rather just used existing mounts to fix a straight mending bracket direct to the secondary radiator.

The secondary radiator is an off the shelf ebay item from ASI as linked all below in the itemised table.
I have run the secondary radiator in serial, post the primary radiator. The top factory radiator hose is untouched, however i have cut a section out of the factory return hose to the thermostat and routed in the secondary radiator and its return back into the engine.

Picture of where i cut the factory bottom radiator hose:

I have then gone about installing another water temp 1/8npt sensor fitting with 32mm male hose ends. Which mates nicely to the factory hose end. On the other side i have added a silicon reducer that steps down from 32mm to 25mm and then connected a copper straight coupler fitting to attach a 19mm silicon hose that routes into the 19mm hose ends on the inlet of the secondary radiator. From the secondary radiator outlet i have added the 19mm silicon hose, another 20mm copper coupler to the 25mm to 32mm silicon reducer and then into the engine block.

Thankfully it all works with no leaks and no kinks that impede coolant flow.
Here it is with the crash bar on too:

Here is the itemised table of parts used to construct the kit. Hopefully this helps others in the quest for lower engine coolant temps.
Obviously i have assembled the kit in Australia so have added in a rough conversion cost to USD.

Description	Cost AUD	Cost in USD @ 0.67 rate to AUD	Link
ASI pro performance radiator 533x168x56mm with 19mm hose barbs	110	73.7	ASI Pro performance
Mending plate bracket 75mm x2	10	6.7	https://www.bunnings.com.au/carinya-190-x-40-x-3-5mm-zinc-plated-mending-plate-single_p0046892
20mm copper coupler fitting x 2	7	4.69	https://www.bunnings.com.au/smartex-press-20mm-copper-straight-coupling_p0037866
SAAS 1/8npt sensor fitting kit	21	14.07	https://www.ebay.com.au/itm/3254722...uid=oBH0VqgwQJi&widget_ver=artemis&media=COPY
SAAS 25x32x76mm straight reducer	30	20.1	https://www.ebay.com.au/itm/1451495...uid=oBH0VqgwQJi&widget_ver=artemis&media=COPY
AAS Straight Silicone Intercooler Hose 19 x 19 x 650mm	34	22.78	https://www.ebay.com.au/itm/1252430...uid=oBH0VqgwQJi&widget_ver=artemis&media=COPY
Total	212	142.04

Now for the results.
On a recent track day in combination with the Gambit oil cooler kit, ECT1 and ECT2 (with ECT1 &2 sensors in the factory locations) looked great with ambient temps at 13~ celsius.
Now i hope you have a go at the kit and please share your results or any other similar approaches to cooling.

Sponsored

 

[ProjectDD]

Thanks for sharing. That's a very cost effective solution! My only concern is the neck down to 19mm causing a reduction in flow. That being said, your temps look great so maybe it doesn't have a significant effect on the overall system performance.

 

[Gambit]

Thanks for sharing. That's a very cost effective solution! My only concern is the neck down to 19mm causing a reduction in flow. That being said, your temps look great so maybe it doesn't have a significant effect on the overall system performance.

Thanks! I had that initial concern as well but it kind of just works well. As it passes the primary radiator gap cleanly without issue. Coolant temps albeit in the cooler ambient temps were very stable across the entirety of the track sessions.

 

[svvitch]

Next up oil cooler?

 

[Gambit]

Next up oil cooler?

Oh yes, that’s the next kit I’ll post

 

[Vito.FL5]

At 13C it would be very hard to overheat the car. I would probably set it up in a way that doesn´t restrict the main radiator lines

 

[Gambit]

At 13C it would be very hard to overheat the car. I would probably set it up in a way that doesn´t restrict the main radiator lines

Yes hopefully it will perform just as well in our summer. Albeit Australian summers are much hotter than in the US as I understand.

 

[keller]

Hi Y'all,
I’m at it again and wanted to share another open source style kit. I’ll call it “Gambit’s FL5 secondary radiator kit”. DIY and build it yourself and hopefully improve and make it better. Lots of inspiration from Driven Crazy and their Hyperthemia kit on the FK8.

Again the benefit of this DIY kit for the community is that all parts are available on ebay or universally available online. You can mix and match whatever parts you have spare such as silicon or rubber hoses, clamps, pipes, brackets etc to make it work and all at a very cost effective total price.

This is the kit installed in front of the primary radiator. I have not had to cut or drill any additional holes for the mounting points, rather just used existing mounts to fix a straight mending bracket direct to the secondary radiator.

The secondary radiator is an off the shelf ebay item from ASI as linked all below in the itemised table.
I have run the secondary radiator in serial, post the primary radiator. The top factory radiator hose is untouched, however i have cut a section out of the factory return hose to the thermostat and routed in the secondary radiator and its return back into the engine.

Picture of where i cut the factory bottom radiator hose:

I have then gone about installing another water temp 1/8npt sensor fitting with 32mm male hose ends. Which mates nicely to the factory hose end. On the other side i have added a silicon reducer that steps down from 32mm to 25mm and then connected a copper straight coupler fitting to attach a 19mm silicon hose that routes into the 19mm hose ends on the inlet of the secondary radiator. From the secondary radiator outlet i have added the 19mm silicon hose, another 20mm copper coupler to the 25mm to 32mm silicon reducer and then into the engine block.

Thankfully it all works with no leaks and no kinks that impede coolant flow.
Here it is with the crash bar on too:

Here is the itemised table of parts used to construct the kit. Hopefully this helps others in the quest for lower engine coolant temps.
Obviously i have assembled the kit in Australia so have added in a rough conversion cost to USD.

Description	Cost AUD	Cost in USD @ 0.67 rate to AUD	Link
ASI pro performance radiator 533x168x56mm with 19mm hose barbs	110	73.7	ASI Pro performance
Mending plate bracket 75mm x2	10	6.7	https://www.bunnings.com.au/carinya-190-x-40-x-3-5mm-zinc-plated-mending-plate-single_p0046892
20mm copper coupler fitting x 2	7	4.69	https://www.bunnings.com.au/smartex-press-20mm-copper-straight-coupling_p0037866
SAAS 1/8npt sensor fitting kit	21	14.07	https://www.ebay.com.au/itm/3254722...uid=oBH0VqgwQJi&widget_ver=artemis&media=COPY
SAAS 25x32x76mm straight reducer	30	20.1	https://www.ebay.com.au/itm/1451495...uid=oBH0VqgwQJi&widget_ver=artemis&media=COPY
AAS Straight Silicone Intercooler Hose 19 x 19 x 650mm	34	22.78	https://www.ebay.com.au/itm/1252430...uid=oBH0VqgwQJi&widget_ver=artemis&media=COPY
Total	212	142.04

Now for the results.
On a recent track day in combination with the Gambit oil cooler kit, ECT1 and ECT2 (with ECT1 &2 sensors in the factory locations) looked great with ambient temps at 13~ celsius.
Now i hope you have a go at the kit and please share your results or any other similar approaches to cooling.

Hey Tony. I used to run that drivencrazy's kit on my fk8. I had problems exactly because it was connected in series with the main radiator. Such large difference between ECT1 and ECT2 (16 C) means flow is being restricted. A well designed system has a small temperature difference, but lower overall values for both (i.e. large flow). Something is off because your ECT2 being 69 C means your thermostat should be completely closed. The factory thermostat opens in the 78-90 C range. Are you using an aftermarket thermostat? Was that ECT2 sensor calibrated?

 

[Gambit]

Hey Tony. I used to run that drivencrazy's kit on my fk8. I had problems exactly because it was connected in series with the main radiator. Such large difference between ECT1 and ECT2 (16 C) means flow is being restricted. A well designed system has a small temperature difference, but lower overall values for both (i.e. large flow). Something is off because your ECT2 being 69 C means your thermostat should be completely closed. The factory thermostat opens in the 78-90 C range. Are you using an aftermarket thermostat? Was that ECT2 sensor calibrated?

Interesting points so I am running the factory thermostat. ECT2 sensor has remained unchanged from factory. I previously had the system run in parallel to the factory radiator implemented in the same fashion as the Mishimoto fk8 secondary radiator. ECT1 and ECT2 numbers were very close however ECT delta between running in parallel and running in serial. Serial has a much lower ECT1 number on average from my data logging.

Running in parallel, was causing some issues as I had to turn the heater on, and then it took a lot longer for the thermostat to open for some strange reason, and it would only stay low if the heater was left on, which defeats the purpose in summer conditions on track.

I am in the process of installing a ECT3 sensor using a Defi FD sensor array. This ECT3 sensor will be post secondary radiator on the return into the engine. That’s looking forward to data logging that sensor.

another track buddy commented that my ECT numbers are too low as well, and that it may compromise Efficient burn off of compounds in oil in order to ensure that the engine is running in its optimum operating condition. However, ambient temperatures were low at 13°C at Philip Island with a cold headwind that day. The true test will be on a much warmer summer day.

 

[keller]

Serial has a much lower ECT1 number on average from my data logging.

A lower ECT1 is all that matters. But yeah you should compare in similar weather and similar load.

Running in parallel, was causing some issues as I had to turn the heater on, and then it took a lot longer for the thermostat to open for some strange reason, and it would only stay low if the heater was left on, which defeats the purpose in summer conditions on track.

This is odd. Our cars run water through the heater core all the time, so it shouldn't need to have the heater on to work properly. Indeed the thermostat should take longer to open up because your engine was being continuously cooled down by the secondary radiator. After all, the heater circuit is not regulated by the thermostat.

I am in the process of installing a ECT3 sensor using a Defi FD sensor array. This ECT3 sensor will be post secondary radiator on the return into the engine. That’s looking forward to data logging that sensor.

Oh, I thought your ECT2 was being measured at the return hose to the thermostat as shown in that picture. So in fact your ECT2 reading is before the secondary radiator and the water returning to the thermostat housing should be even cooler than 69 C.

The fact you added a radiator _after_ the ECT2 sensor and it still changed the ECT2 measurement is a hint that your larger ECT1 to ECT2 difference is just the result of a restriction. It's counterintuitive but a larger ECT1 to ECT2 difference actually means less heat is being rejected. As an example, when you are cruising on a road and your thermostat gradually closes, ECT2 will go down to ~25 C (ambient temps). This huge ECT1 (~80 C) to ECT2 (~25 C) difference is the result of a huge restriction (the thermostat) and your radiator is barely rejecting any heat.

 

[Gambit]

Yes the secondary radiator run in parallel took a lot longer to get ECT 1 temps to a stable level.

Indeed, what you see in the pic above is my ECT3 sensor which is post secondary radiator and before the engine, yet to be logged.

Have a look at this chart, which depicts a recent track day i had at Phillip Island with the data overlaid from another FL5 that ran at Phillip Island. Unfortunately we were not there at the same time, but what i have done is aligned the data the first frame post VMAX on the straight to somewhat estimate what both systems are doing in totality in order to recover from the last pull.

I never got a chance to test the secondary radiator parallel configuration on the track, however i do have a datalog on the street in similar ambient temps of 13~ Celsius. Joel from driven crazy (developer of the hypothermia kit) dyno tuned my car in person and looked over my parallel secondary radiator configuration and seeing first hand data suggested changing to a serial configuration. What can be seen between the two setups is how large a difference the delta is between ECT1 and ECT2.

However lets go another step and overlay a street log of the serial configuration on top, albeit log taken during the heat of the day in much warmer ambient temps, over similar roads but with more traffic.

The difference between a parallel and serial implementation of a secondary radiator setup is up for debate on which is more efficient. Ideally you would want to test both configurations against each other on the same track on the same day to truly determine which was more efficient.

However and overall when comparing at the roughest approximation the difference between 'no' secondary radiator and 'a' secondary radiator on track, without a doubt a secondary radiator setup delivers much lower and consistently lower ECT 1 temps, even when giving it a beating on track.

Sponsored