Supplemental cooling for the charger.

[jray3]

After the charger failure, I'm keeping a closer eye on the new one and have been reminded that after several hours of charging, the top cover is extremely hot to the touch (actual temperatures coming soon), much warmer than the outbound coolant line. My pump is running and there is vigorous circulation, but of course, that heat exchanger is in the bottom of the box where it can't do the most good. Since the little filter box on top of the charger is in firm contact with the sound insulation on the underside of the motor room hatch, it has little chance to cool off.
For how much heat we're dealing with, at 96% assumed charger efficiency, 4% of 3300W= 132W or 640 Watt-hours per full recharge.
Throw in the DC-DC operation and I'd bet we're dealing with as much as 1 kWh of waste heat. To remove just 0.5 kWh of heat we'd need to soak up 1706 Btu.

SO, I'm considering the following unprioritized list of options.
#1- Remove some insulation to create an air gap above the charger. (and then likely add Dynamat to offset any increased noise)
#2- Add a ducted fan fed by the charger AC input to blow cool air across the top of the charger or pull hot air away.
#3- Add a liquid-cooled cold plate atop the charger and plumb it into the coolant loop.
(here's a candidate https://www.shopaavid.com/Product/416401U00000G Once could alternatively make a copper coil or use a tiny radiator designed for computer overclocking. I also a tiny heater core that serves to preheat winter combustion air in a Honda Odyssey.
#4- Add a bladder full of phase-change wax to serve as a heat sink atop the charger. Paraffin with a phase change at 84 degrees F will absorb 87 Btu/lb in the melting process. So, a whopping 20lbs of wax would be needed to reduce the heat by half, and not help at all on a warm summer day... wow. no go there
#5- Cut a hole in the motor room hatch and put a butterfly vent in so that the hot air rises into the cabin. Free heat in winter, but less effective in summer when needed most. https://www.etrailer.com/Enclosed-T...MItfjtpeqX3AIVGMRkCh3JYQoxEAQYASABEgLEK_D_BwE
#6- prop open the motor room hatch during recharging and open the windows a crack...

Comments please. Is any of this worthwhile, or is a charger replacement every 50-90k just to be expected?

 

[Aerowhatt]

SO, I'm considering the following unprioritized list of options.
#1- Remove some insulation to create an air gap above the charger. (and then likely add Dynamat to offset any increased noise)
#2- Add a ducted fan fed by the charger AC input to blow cool air across the top of the charger or pull hot air away.
#3- Add a liquid-cooled cold plate atop the charger and plumb it into the coolant loop.
(here's a candidate https://www.shopaavid.com/Product/416401U00000G Once could alternatively make a copper coil or use a tiny radiator designed for computer overclocking. I also a tiny heater core that serves to preheat winter combustion air in a Honda Odyssey.
#4- Add a bladder full of phase-change wax to serve as a heat sink atop the charger. Paraffin with a phase change at 84 degrees F will absorb 87 Btu/lb in the melting process. So, a whopping 20lbs of wax would be needed to reduce the heat by half, and not help at all on a warm summer day... wow. no go there
#5- Cut a hole in the motor room hatch and put a butterfly vent in so that the hot air rises into the cabin. Free heat in winter, but less effective in summer when needed most. https://www.etrailer.com/Enclosed-T...MItfjtpeqX3AIVGMRkCh3JYQoxEAQYASABEgLEK_D_BwE
#6- prop open the motor room hatch during recharging and open the windows a crack...

Comments please. Is any of this worthwhile, or is a charger replacement every 50-90k just to be expected?

I have a long history with air cooling components on EVs. The Charger is a perfect candidate because it sits high in the bay where heat rises and becomes trapped. While the car is moving there is plenty of airflow, but when parked there would be very little. A finned heat sink on the cover plate. With a Nidec blower situated such that it gives good flow across the heatsink and creates enough turbulence it the compartment to not be recirculating hot air over and over would do the trick nicely IMO.

Aerowhatt

 

[mikedufty]

According to wikipedia over 30,000 i-MiEVs have been sold. I think if there was a fundamental design fault with the charger that required more cooling, you would be seeing more reports of failures than we are.

I suspect you would be very unlucky to have two fail.

 

[Don]

Comments please. Is any of this worthwhile, or is a charger replacement every 50-90k just to be expected?

I guess if you're always in need of a quick turnaround, your 3300 watt figure just can't be avoided - But, if not, recharging L1 at 12 amps (1440 watts) will cut your waste heat by more than half and it's a completely free option, costing you nothing and no mods needed. Also, our L2 EVSE is set limited to 12 amps, so we have never once used the full 3300 watts the charger is capable of. Quite probably, 2,880 watts is much less of a strain than the full 3300 watts is. Also, I never plug in to recharge until the car has been sitting, cooling off for at least a couple hours, usually longer

On our '17 Chevy Volt, I have the EVSE set to just 8 amps at 240 volts - Once I park it to charge, I never need to drive it again before the next day, so why push things??

We have always used L1 more often than L2 since we bought the first car more than 6 years ago. We never plug in L2 overnight because there's just no need - The car will be full in the morning no matter which you use and when we don't need a bunch of amps in a hurry, L1 is the obvious choice for us. A very large percentage of the time, we don't recharge to full anyway and 4 or 5 hours of L1 is easier to manage (for us anyway) than 2 or 3 hours of L2. - If we plug in L2 and forget about it, the car usually gets a full recharge when we really only wanted 75 or 80%.
When one car is already using the L1 EVSE, we sometimes use L2 for the other car, but usually just for an hour or two and not to fully recharge it

Don

 

[jray3]

Good point Don, I’ll try to run a test to confirm the lower max temp on lower amps. The top of the charger is thermally insulated so effectively that I wouldn’t be very surprised if it reaches a similar temperature after 10-12 hrs and just spends more time cooking!

Our usage patterns usually dictate either L2 recharging, or to use another car (how else does one accumulate 110,000 i-miles in 6 1/2 yrs?). Though there are many days when L1 could suffice, but doing that mental math on every plug-in is not acceptable to my wife.

 

[Aerowhatt]

So I had hooked up a remote sensor to the charger top plate last night and had a lot of driving this morning. On the road the unit working as only the DC to DC with all that air flowing around from motion only runs 3 or 4 degrees F above the ambient temperature.

Charging on L1 at 12 amps it peaked at 128F which is 48F above ambient. Switching to L2 the temperature rose to a a peak of 136F or 56F above ambient temperature. So an extra 8F degrees of heat build up for L2 charging.

Jay if I'm understanding your description of the motor bay hatch insulation on your 2012 it actually touches the top of the charger unit? The 2014 hatch definitely has some air space between the hatch sound insulation and the top of the charger. I would guess somewhere around 3/4 of an inch. The insulation is molded on the bottom side to provide this space. Still since heat rises I'm sure the air temperature under there while charging hovers significantly above ambient temperature. I think a well placed Nidec blower (12 volt 0.6 amps 25 cfm) would lower those numbers by around 20F. My experience is that all electronics run better and longer if kept as cool as possible. 136F isn't all that bad, in my experience, but the internal components are hotter than the casing.

Something to consider for sure considering the out of warranty cost and/or hassle of replacing a charger dc/dc. Another thought what if the coolant pump ran full time instead of just 25% of the time while charging???

Aerowhatt

 

[siai47]

The charger, inverter and motor are in the same cooling loop. There is a pump that circulates coolant through all three during charging. If you put your hand on the coolant hose between the inverter and the charger you can feel the circulating pump cycling on and off. And yes, the motor gets hot during charging as it's windings are used as part of the charging circuit. The coolant flows to the radiator at the front of the car and there in is the problem. Although the radiator heats up, with the car stationary, there is no airflow to cool the radiator. If the cooling fan would start during charging, the problem would be solved. However, it doesn't. I have used a fan blowing in the radiator opening and with the hood open a whole lot of hot air comes out of the radiator. The components in rear cool right down. So, try a way of turning on the radiator fan when charging and solve the problem or put a fan in front of the radiator and see if this works for you. In the summer in central Florida I always used this method when L2 charging.

 

[jray3]

Thanks siai47, so glad to see you still dropping by to contribute. A reed switch on the pump circuit to activate the radiator fan relay could be a way to maximize that cooling without excessive runtime.

 

[Aerowhatt]

The radiator fan has two speeds. The low one is very unobtrusive and almost unnoticeable. I noticed it inside the garage with the AC on medium. It will come on to cool the AC condenser periodically. This fan speed would be perfect for getting the job done without using a bunch of power or making much sound.

Me too: thanks for checking in Siai47!

Aerowhatt

 

[kiev]

Maybe someone with a CDROM version can check for the US cars, but there are two relays shown in the online workshop manual, Lo and Hi, for the fan. Maybe a jumper wire with a switch can be made to turn on the Lo relay whenever you want it.

 

[Aerowhatt]

Sure enough, Siai47's information is correct yet again. I moved my remote temp sensor to the return line of the coolant loop a few inches from the connection to the coolant tank in the back of the car. Under L2 charge it peaked at 113F or 33F above ambient. So we are cooling the components with what would be considered nice hot water coming out of one of our kitchen sink faucets. Of course it is worse if the ambient temperature is higher than my fortunate 80F garage is!

Also noticed that the radiator cooling fan only has a two wire connector. Which means that the low speed voltage is achieved somewhere else in the cars systems. It is not built into the fan motor

Realistically the radiator (with airflow) should be able to return coolant just a few degrees (<5F) above ambient air temperatures. I think a 25F to 30F reduction in operating temperature is likely worth the time and effort for a modification.

Aerowhatt

 

[Aerowhatt]

I put 50 CFM worth of blower airflow through the condenser and radiator from the front. After an hour the coolant temperature stabilized down ~9F to 104.4. Removing the motor bay hood there was a lot of really hot air trapped in there. Feeling the inverter it was still too hot to hold on any surface (can't really reach the bottom).

I moved the blowers to the charger bay, blowing air over the top and across both sides, more or less. After an additional hour the coolant temperature stabilized at < 100F. The charger can be touched and held on any surface without discomfort. Given this rather unexpected result I'm going to go with two Nidec blowers connected to the 12V DC harness. By using a properly rated heat switch it will cool when needed and be off the rest of the time automatically. Mounting will be a bit tricky but looks workable.

Conclusions from these observations:

1. The OEM cooling strategy for the charger/DC-DC is underwhelming.
2. Increasing the cooling at the radiator provides some improvement.
3. Direct forced air cooling of the Aluminum case of the charger makes a big improvement. Not only is the case much, much cooler, but it being so, is allowing the components to dump so much heat to the case that the coolant temperature drops ~13F or more, The obvious conclusion from this observation is that so much heat is being pulled of the unit by air cooling the case, That even the fluid cooled plate at the bottom of the unit has significantly less heat to transfer to the coolant and the coolant stabilizes at < 100F

As soon as I get this mod fabricated and installed I will post pictures, etc., and recheck the coolant temperature steady state at L-1 and L-2 rates

Aerowhatt

 

[JoeS]

I was shocked, I tell you, I was shocked, as up until now I had considered the engineering on our i-MiEV to be excellent. It simply never occurred to me that, with liquid cooling, there would be any thermal issues, and certainly did not expect high case temperatures. I had simply assumed that Mitsubishi had calculated that a radiator fan was unnecessary... DUH. I haven't removed the rear access cover in years, and I bet a layer or dust/dirt doesn't help the thermal situation either.

Aerowhatt, thank you for performing the measurements. Like I said, shocking! siai47, thank you for still keeping up with us and offering your wisdom.

Like Don, my first reaction was to go back to charging on L1 (120vac).

For the time being as a quick-and-dirty partial fix attempt in this summer heat, I've stuck a 120vac fan in front of the car at the radiator inlet and one underneath the back of the car aiming straight up to at least move some of the air back there. Edit: I activate both fans with the timer for the EVSE.

Aerowhatt, looking forward to seeing what you come up with.

 

[wmcbrine]

But, if not, recharging L1 at 12 amps (1440 watts) will cut your waste heat by more than half and it's a completely free option, costing you nothing and no mods needed.

L1 charging is less efficient, so there's a cost in terms of electricity. (I'm not sure how much.)

 

[Aerowhatt]

I have a prototype cooling blower mounted in my car and two charges monitored with it. First let's review.

First all of, the numbers are are based on my 80F ambient charging environment. It is reasonable to use these numbers adding or subtracting as needed depending on your ambient charging location temperatures.

As designed the OEM cooling system delivered the following results.

Under L-2 the case temperature peaked at 56F above ambient temp.
Under L-1 (12 amp) the case temperature peaked at 48F above ambient temp.
So charging L-1 only delivered a ~15% decrease in charger temperature rise.

Switching to coolant inlet temperature as the comparison metric.
Under L-2 OEM setup the inlet coolant temperature was 113F (33F above ambient)
With 50 CFM of ambient air blowing directly through the radiator this was reduced to 104.4F (charger case still too hot to hold on to)
With 25 CFM strategically blowing on the charger case, the case (nothing on the radiator) the coolant temp reduced to 99F (19F above ambient) (charger case easily held on all sides and top).

The prototype is a single Nidec Gama 30 blower strategically placed in the motor bay. As a prototype the the bracket is pretty ugly, but that's why we prototype. Nothing like doing it to show one what needs improved. Like the prototype bracket, the final version will require no drilling of the car whatsoever, and will be fairly easy to remove and reinstalled should a warranty trip to the dealer come up. Even with the hatch lid completely installed and the sound insulation, etc., plus all the junk I keep in the trunk . I'm getting the same temps as I did in the cover open - blower cooling the charger trials.

Funny how realizing your >$4,000 charger/DC-DC is running uncomfortably hot motivates one to get things done. I plain didn't want to charge again without a solution in place. Wife always get nervous when I'm working in the garage late into the night

Aerowhatt

 

[wmcbrine]

I was shocked, I tell you, I was shocked, as up until now I had considered the engineering on our i-MiEV to be excellent. It simply never occurred to me that, with liquid cooling, there would be any thermal issues, and certainly did not expect high case temperatures.

I don't think it's been proven, yet, that there are thermal issues -- only that greater cooling is possible. But, what matters is what temperatures are in spec for the components, not whether they're hot to the human touch.

 

[Aerowhatt]

I don't think it's been proven, yet, that there are thermal issues -- only that greater cooling is possible. But, what matters is what temperatures are in spec for the components, not whether they're hot to the human touch.

I've been charging EV's for clients for 18 years. Recommending and installing the gear to get the job done. Just my experience and opinion here. Manufactures spec their components to beat the competitions components. So specs are more often than not generous in the direction that OEM buyers like to see, to be competitive. The EV's on which I was able talk the owners into supplemental cooling for the chargers would get at least twice the MTBF (mean time between failures). They spent a lot less money in the long run too. All electronics work better and longer at lower temperatures IMO and experience. They always have and always will!

Imagine charging outdoors with 100F ambient temperature. Case temps would be 156F at L-2 rate. The case is always cooler than the components cranking out the heat (sometimes by 40F). I'm going with actively cooler since specs have always been off the mark for longevity in my rather extensive experience. The proof when it comes is too pricey in my estimation!

Aerowhatt

 

[JoeS]

...Funny how realizing your >$4,000 charger/DC-DC is running uncomfortably hot motivates one to get things done. I plain didn't want to charge again without a solution in place. Wife always get nervous when I'm working in the garage late into the night

Same here!

Regarding failure rates, one need to simply google "mtbf vs. temperature" for just about any type of electronics component and check out the graphs to see the dramatic component life decrease for increasing temperature. For semiconductors, it all goes back to maintaining as low a junction temperature as possible.

Without sitting in on the design reviews of each of our major subassemblies, we simply don't know to what depth each individual component was analyzed and what coordination existed between the circuit designers and mechanical engineers, especially from a thermal analysis and control standpoint. For example, different potting compounds can act as thermal conductors or thermal insulators - BIG difference! Liquid cooling normally implies that all significant power-dissipating components are tied into that heat exchanger, and I would have expected minimal contribution to cooling by the chassis. For reasons already ably discussed by Aerowhatt, I find both the lack of a temperature-activated electric fan in front of the radiator and the high chassis temperatures during charging to be disturbing.

Aerowhatt, what did you pick up as your fan's 12v power source, activated only during charging?

 

[Aerowhatt]

Aerowhatt, what did you pick up as your fan's 12v power source, activated only during charging?

Right now it is temporary, separate 12 ah AGM. The DC to DC output is right there and easy to connect to. The negative will go to ground (likely to the bracket for the blower). Fuse (3 amp) right at the positive connection to the DC to DC post then to a surface temp switch on the opposite end of the charger case. It should turn on when the case on the leeward side gets to ~120F then off when it cools to ~110F. From what I'm seeing that should have it on only when charging.

I need to do two setups. One for each car, so I might make a jig for the bracket. If so, then it's a short step further to put together a kit that others could use (if there is enough interest). Even if it was slow to shut off in a very hot climate it wouldn't be much of an issue. These blowers are very efficient and only use 7 to 9 watts running.

Aerowhatt

 

[JoeS]

Aerowhatt, glad to see your entrepreneurial spirit alive and well, and count me in for a couple of 'kits', if you decide to do it.

I haven't taken my front bumper off lately, but IIRC there was lots of room in there to place a fan. Too many other projects for me to play with this right now, and my external 120vac box fan is certainly least-effort for now.