DIY Electric Car Forums banner

Tesla cooling system layout

6765 Views 46 Replies 10 Participants Last post by  57Chevy
Hello everyone, I have been searching for information for cooling system plumbing, specifically component order and coolant flow. I have only been able to find bits and pieces of what I need through several sources but nothing to give me a complete answer. I am also trying to figure out if I need to run valving or not (I wasn't planning on it originally). If I do need to what are people using to control it? Here is the equipment I am running:

  • Tesla Model S P85D large drive unit
  • 16 Model S battery modules
  • Model S gen 2 onboard charger
  • Model S DC/DC converter

I also am planning on including a battery heater and have not yet figured out what will be controlling the pumps and fan. Here is what I have drawn up so far.
Rectangle Font Slope Parallel Schematic

Any input from those who have been there and done that would be greatly appreciated! This is indeed my first rodeo! lol
See less See more
1 - 20 of 47 Posts
Have you seen this DD into the Model 3 system?: https://jalopnik.com/the-tesla-model-3s-superbottle-easter-egg-is-a-fascin-1830992728

The biggest difference I see with your diagram is the motor is at the end of the circuit going to the radiator and the batteries are on a separate(or partially separate) circuit in the Mod 3. And, of course, the Tesla has the "Superbottle" to control everything. The motor at the end seems to be a common practice. Could this be to give maximum cooling to the more heat sensitive electronics?
I highly suggest a separate cooling loop for your batteries and then another for the power electronics and motor. The heater is really only needed for the battery and you don't want the heat from the PE and motor in the battery loop in the summer.
  • Like
Reactions: 1
Have you seen this DD into the Model 3 system?: https://jalopnik.com/the-tesla-model-3s-superbottle-easter-egg-is-a-fascin-1830992728

The biggest difference I see with your diagram is the motor is at the end of the circuit going to the radiator and the batteries are on a separate(or partially separate) circuit in the Mod 3. And, of course, the Tesla has the "Superbottle" to control everything. The motor at the end seems to be a common practice. Could this be to give maximum cooling to the more heat sensitive electronics?
I have not seen that, thank you for sharing! I have not really been looking into Model 3 components and cooling layout because I am not utilizing anything Model 3. I am also not utilizing control valves as, as far as I am aware, there is no non factory (open source) controllers to make them work properly. This would change the plumbing configuration but I am not sure how exactly. I am not apposed to alternate pump placement. It looks like the model three only utilizes one pump? I am not sure if using more than one makes placement less critical or not.
I highly suggest a separate cooling loop for your batteries and then another for the power electronics and motor. The heater is really only needed for the battery and you don't want the heat from the PE and motor in the battery loop in the summer.
Something more like this? I guessed at the pump location. I think both pumps pulling from the reservoir will be alright?
Rectangle Slope Line Font Parallel
See less See more
It would be better to put the DC converter and charger upstream of the motor. I'm guessing the motor coolant has the highest temperature and possibly heat output of all of the components. It would probably not good to subject electronic parts to this. And, as I'm sure brian will point out if I don't-God help me, this is understanding the charger is normally not operating if the motor is operating.
It would be better to put the DC converter and charger upstream of the motor. I'm guessing the motor coolant has the highest temperature and possibly heat output of all of the components. It would probably not good to subject electronic parts to this. And, as I'm sure brian will point out if I don't-God help me, this is understanding the charger is normally not operating if the motor is operating.
Excellent point! You guys have been such a huge help. Thank you for your input! How's this look?
Rectangle Slope Parallel Font Schematic
See less See more
Better. But we will actually use two radiators and really have a separate loop for battery (and cabin heat) and the E-motor/Inverter/DCDC/OBC. The battery then can get heated water (for cold weather), just loop water (when the battery is at a nice temperature but the humans still need heat
) and then cooled water from it's radiator if it starts getting hot. I will use an Arduino to control the contactor A to turn the heater on/off, relays B-D for the loop control for the battery, and relays E-G for the pumps. It will also monitor the flow meters and temperatures to make sure it is all working correctly. Like this:
Slope Rectangle Parallel Font Diagram
See less See more
Tesla use several diverter valves to route the coolant based on conditions, i.e. no point heating the battery and then passing the water through the radiator when it is cold outside. Diverter valve keeps battery in its own heating loop. There is also provision for the a/c to exchange heat with the coolant
Yeah, I had a pretty ugly diagram in another thread of my potential coolant loop, which had a similar possibly over-ambitious design goal: Measure ambient temps, battery temps (via BMS CAN bus), and monitor passenger heating demand, and use a series of coolant valves controlled by an esp32 and automotive relays. I've written the code for it and validated it with a table of temperatures for the various scenarios, but like most of my build, still castles in the sky.

The hope though was that all scenarios would harvest heat from all heat generating components when needed rather than waste any through the radiator, while needing only a single coolant heater for all scenarios (battery heat, passenger heat). Since then I may be leaning towards a couple of loops and redundant bits in favor of valve switching simplicity.

Previous newb thread: Coolant loop design.
I see, you're selectively cutting off flow through the radiator but opening it through the heater loop for battery heat/cool. Or alternately bypassing via (c). Pretty clean. Any concerns with B,C,D not responding and/or any way to monitor that? Or phrased another way, is there a 1 input, 3 output flow selecting valve out there so that at worst there is flow the wrong way but no way to be overall closed?
Coolant loop valve fail is not fatal. You just limit car power or charge rate in that unlikely scenario. The key is having all the sensors in place to do that.

That said, 4 N.O. valves on a common input manifold can do your 1:3 select (1 to any, actually). The 4th valve is for your redundancy, if you must - not necessary, IMO. You can also use N.C. valves to save power, but a fail in the solenoid circuit can mildly screw you - again, you can enter a limp mode.

Flow meters are expensive and a total waste of money, IMO, when you already have open loop speed controlled pumps and temperature sensors. Even when you both make >$200/yr, flow meters are expensive and unnecessary.

By "heat pipe", I'm guessing they mean an electric fluid heater, a heated pipe, and not a heat pipe.
Agree that the second diagram is an improvement Mowza. However, there is still room for improvement.
You do not want to cook your batteries via the waste heat of the motor and inverter.
They operate at different allowed temperatures. For a motor coolant of 70 degrees centigrade is not a problem, for a battery it is.
See also my blogpost on Ideal battery temperature?
So you need to add a way do disconnect the two loops. When needed you can then use the motor/inverter waste heat to warm up the batteries. This is also what Tesla does (but they do more), see my blogpost Using Tesla thermal management system parts.

In my project I even had three separate cooling circuits.
http://instagr.am/p/Bw3l5uXHmCU/
  • Like
Reactions: 1
Flow meters are expensive and a total waste of money, IMO, when you already have open loop speed controlled pumps and temperature sensors. Even when you both make >$200/yr, flow meters are expensive and unnecessary.
Basic flow sensors for Arduino are like $12-15...
I see, you're selectively cutting off flow through the radiator but opening it through the heater loop for battery heat/cool. Or alternately bypassing via (c). Pretty clean. Any concerns with B,C,D not responding and/or any way to monitor that?
That's what the flow sensors are for. If I am expecting flow and don't have any, I can flag a message. As a backup to that, the BMS and VCU is also monitoring all the temps of the inverter, motor, batteries, and DCDC/OBC so if anything is too hot (or cold) they will also protect the components.
Better. But we will actually use two radiators and really have a separate loop for battery (and cabin heat) and the E-motor/Inverter/DCDC/OBC. The battery then can get heated water (for cold weather), just loop water (when the battery is at a nice temperature but the humans still need heat
) and then cooled water from it's radiator if it starts getting hot. I will use an Arduino to control the contactor A to turn the heater on/off, relays B-D for the loop control for the battery, and relays E-G for the pumps. It will also monitor the flow meters and temperatures to make sure it is all working correctly.

That is another excellent point. It makes perfect sense to divide the heat load and have each system not dependent on the other. I really like the use of flow meters from a diagnostic/monitoring standpoint. One would definitely want to know that the cooling system is functioning properly and receive a warning when it is not.
While I was originally planning on utilizing a standard heater core for cabin heat I came across a PTC heater core from an EV smart car which is very similar in size to the heater cores used in Vintage Air systems. I was going to see if I could swap the core into the Vintage Air unit and use electric heat. I am glad you shared your design in the event I will not be able to do so as it gives me a backup plan. I am not currently versed with Arduino but will be using it for other vehicle systems/features so I will be at some point during this build.
I was also looking to replace the stock in-dash heater core with a PTC heater core...but this doesn't help the battery when cold. So, this is why I wanted to utilize a common heated water system for both the battery and the cabin and have the chance to swap loops once the battery is warm yet continue to heat the cabin...
Agree that the second diagram is an improvement Mowza. However, there is still room for improvement.
You do not want to cook your batteries via the waste heat of the motor and inverter.
They operate at different allowed temperatures. For a motor coolant of 70 degrees centigrade is not a problem, for a battery it is.
See also my blogpost on Ideal battery temperature?
So you need to add a way do disconnect the two loops. When needed you can then use the motor/inverter waste heat to warm up the batteries. This is also what Tesla does (but they do more), see my blogpost Using Tesla thermal management system parts.

In my project I even had three separate cooling circuits.
http://instagr.am/p/Bw3l5uXHmCU/
Thank you for sharing all the info! I did not see the page from your first link but I have seen the page from the second link a few times now. Overall I was hoping for a relatively simple and less complex cooling/heating arrangement but it seems that is not the most feasible or ideal approach for a number of reasons. I really appreciate all of the input/feedback so far from everyone. I feel much more informed an able to take on the task at hand.
I was also looking to replace the stock in-dash heater core with a PTC heater core...but this doesn't help the battery when cold. So, this is why I wanted to utilize a common heated water system for both the battery and the cabin and have the chance to swap loops once the battery is warm yet continue to heat the cabin...
To make sure I have my head wrapped around this correctly, the reason you chose a standard heater core was because a cold battery running both a battery heater and PTC heater core would not be ideal due to its already reduced output correct? The thought being that the operator would just have to wait before turning on the heat like they would in a gas engine vehicle which would leave the battery running only the battery heater. That makes sense using only one electric component vs two reducing overall demand.
I do have two questions. Question one, would the battery heater have enough heating capacity to heat both the batteries and the heater core at the same time when demand is the highest (just starting out on a cold day)? Typical coolant temps entering the heater core of a gasoline engine vehicle are around 180-200 degrees F at full op temp. What I don't know is what the actual water inlet temp requirement would be to achieve an ideal vent temp on a cold day. I know there are laws regarding the heat transfer rate to the air passing through the heater core and the resulting air temp that leaves it but I don't know what they are either. I imagine they change based on the temp of the air being pushed into the core. In my case it the HVAC blower is always pulling from inside the cabin where that temp would change as the cabin heated up. This would presumably reduce the required core temp to achieve that same vent outlet temp.
Question two, when the demand for heat is required for both the batteries and the cabin, can your system regulate the flow of that warm coolant to both components? The concern being the water temp requirement for the heater core might need to be higher to obtain an ideal vent output temp. The flow of the higher temp coolant would need to be choked down to the batteries to prevent them from being overheated.
See less See more
So...first...this is just a concept that we will build and check functionality.

The main idea is that I would like to combine the battery heater and cabin heater using one heating element so I don't need two contactors rated at ~400VDC and 30ish amps (i.e. Gigavac). The idea is to just get the battery up to a nominal temp in cold weather for charging and initial driving. i.e. 10 degrees C using the heated water and then switch over to the bypass loop (no heating or cooling) for the battery (or even the cooling loop later if the temps get too high...but I doubt this...not in our application and in the cold weather). After that I only need the warm water for the heater core for the meat puppets in the cabin.

Our concept is using a 4kw heating element. We will try it and see how well this works. Might be a struggle if the weather is like it is now (-9degF yesterday) but again...we are trying and learning. Maybe with it on and running we don't get enough heat out and I have to switch over to a PTC heater for the cabin...we will see.
See less See more
1 - 20 of 47 Posts
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