Well to start off, I wanted to throw out some caveats. This article is from my experience with my sailboat, a 1975 Hughes/North Star 35/1500 and converting her from gas to electric power.
There were quite a few people that asked my range and how I was dealing with cooling. One important piece of data is that when I originally bought the boat the shaft alignment was not great. I found that the shaft when centered was very difficult to turn. After investigation I found that my cutlass bearing support strut was slightly bent. I took some time to straighten that and I went from struggling to turn the shaft to wondering how I can add a break to it to stop it from free turning under sail. I can’t in words explain how much easier it was to turn after I realigned it. As you can imagine, the effort to turn the prop shaft dropped significantly which also in turn drops the heat considerably. I believe I went from 1500 watts to 750 watts to get my 3 knots. It was that significant.
The second thing for me is that I never motor anywhere. I motor out of my slip and raise the sails. I have learned to trust my gear and I sail almost into the last 250 feet left to motor. As I originally didn’t have any liquid cooling, I was reliant on air and as I motored I typically left the the cockpit door open and as such ambient air was my cooling. I have had a couple of shut downs, once blown off anchor during 58 knots winds and once going through the Hamilton canal/Burlington Lift Bridge, that was interesting…
I also test crossed the Hamilton Harbour at 3 knots quite a few times. Lets just round that at 3 miles/5 km across. I never had any real issues aside from I knew the controller got hotter the faster you go, so I just kept it at my 750 watts, and crossed at 3 knots (3.4 mph/5.56kph). I usually made it across within an hour.
A big part of my decision to go slow is that my batteries could only handle a 25amp draw rate. If I took power out of the batteries any faster than that I would have damaged my batteries. This year, I changed to Lithium LiFoPO4 batteries and I am now finding myself with cooling alarms, and hence my testing as I have run the motor now at 75 amps for quite some time (to the point of overheating the controller).
Controller vs Motor Temperatures
Gear:
One important thing to note is that the controller gets way hotter than the motor. By the time the motor is hitting high temperatures you would have fried the controller. So I started my liquid cooling by concentrating on the controller.
I used a pump, radiator and cooling plate. I placed links here for convenience as well to the transmission cooler.
One Pump or Two?
I decided on putting two pumps in, mostly for redundancy but also because if the cooling system shared coolant then the higher temperature controller heat would be artificially increasing the motor temp. Each system is closed loop, one for the controller and one for the motor.
Quick Conclusion
For all those that just want results, I decided to go with the computer radiator qty two(2) with three(3) fans on each one. I am looking for 120mm industrial fans and Noctua is being evaluated with an IP67 version.
The computer radiator with 3 fans was simplest, used the least amount of power and provided adequate cooling.
In real world testing, a quick backing off of the throttle drops the temperatures quickly as well. Adding in a generator gets rid of any battery issues and distance is dependent on how much gasoline is carried. A 3500 watt generator is what I intend to use as backup power.
Testing
I did all tests with a simple point and shoot test gun that reads the temperature.
So for my initial tests, although I typically run at 750 to 1200 watts depending on the wind, I chose 500 watts, 1800 watts and 3500 watts. 500 watts is my trolling rate, get out of the marina and moves me at about 1.5 to 2 knots. 1800 watts is my highest I have ever really used. It gets the boat moving at what I consider the point of diminishing returns, where the energy in is just being consumed and there are no equal notable gains in boat speed. 1800 watts gets Northern Comfort up to about 4 to 4.5 knots. At 3500 watts the boat rudder is starting to really feel the impacts of the prop walk. It’s quite significant. Speeds are 6 knots+. All these speeds are wind dependent (heading into the wind or away from the wind).
I did all testing tied up in the marina and my experience shows that this causes more effort than moving. When the boat is moving it seems that it requires less effort. I would have thought that 500watts moving or tied up to the dock would be the same but it’s not for some reason. I can go further than tied up to the dock under the same working load.
The controller is set to downgrade the power at 100 degrees and shut off at 150 degrees Celsius. That in my opinion is very hot but Golden Motor says that it is normal to run hot.
Initial Test Data
No Liquid Cooling – Just Air
At 500 watts it seems that I could not go on more than 40 minutes. At 40 minutes the controller reached 100 degrees Celsius and shut off.
1800 watts only lasted 12 minutes.
3500 watts lasted 5 minutes.
I also was confused by the 500 watt load temperature readings so I added a simple fan blowing on it and I never hit the cutoff so I think moving air at a minimum is required.
As for motor temperatures, no temperatures ran above 40 degrees Celsius. As the controller failed in all tests, the control cut off the power way before any high temperatures could be recorded.
If you want complete data on each test, I have put this on our Patreon page. We could use some help with our journey and it’s a small ask, a cup of coffee worth, to help us out.
Adding Liquid Cooling to the Controller – 360mm Cooling Rad and 3 Fans
I had some spare 120mm fans for computers as part of my mining rigs, so I decided to try my luck with it. I am not certain the exact cooling specifications aside from the fact that the computer fans are 120mm x 120mm and have 75 CFM rating at 12v.
The radiator I chose was a copper radiator 360mm size which is 39.5cm x 12x x 2.5 cm / 15.6″ x 4.7″ x 1.0″ dimensions. As the radiator surface area can hold 3 fans, they call it a 360mm (to hold three 120mm fans). It has some great cooling properties and as the controller has almost identical heat dissipation qualities as a computer CPU I decided to try it.
The pump was a 12v 19w pump with g1/4 screws. These are not NPT threads, just straight and require some barb connectors with o-rings to keep the water out. The one I ordered had a 110mm acrylic liquid cooling tank. The reason I wanted this reservoir on it was to assist with filling and providing an easy way to let air escape as well as keep a visual eye on the glycol levels.
Hey lets throw in a transmission cooler as an option:
As a last test, I wanted to see if a transmission cooler with a huge 8″ 1200CFM fan would make any difference. I sourced from Mismoto their transmission cooler kit. This kit was fairly cost effective considering fans for the computer radiator are $25 each and I needed 6 of them.
What did I test?
As the above tests showed a relatively linear comparison, I decided to stick with 1800 watts. If you look at each temperature curve, they are pretty predictable with a little amount of data. Also, as previously mentioned, my ‘normal’ usage was between 750 watts and 1200 watts depending on the wind, so testing 1800 watts would be my ‘normal peak’ usage.
Test Results: Liquid Cooling options vs No Cooling
Observations
Although not completely scientific I tried to be as consistent to get my readings. I used the same measuring tool and used the Victon GX Touch screen to get the wattage within 1800-1900 watts (taking into account the existing power being consumed before starting).
I started the pump, took the DC consumption voltage, added 1800 to it and tried to hit that number using the throttle.
I was tied to the dock at the time of testing.
Controller @1800 watts
The controller with no cooling didn’t last 10 minutes.
I was surprised to see that 3 fans faired about as well as 6 fans.
I was very surprised to see that the computer radiator out-performed the better FAN on the transmission cooler. I suspect that this is because the computer radiator has more heat dissipation being designed to work in tight spaces, that technology really has come a long way whereas the transmission cooler just has a winding tube through some heat dissipation fins. Even though the fan was 3-6 times more powerful, the heat dissipation efficiency is quite a bit better in the computer 360mm radiator.
I would state that the controller curves were very similar and stabilized between 60 and 70 degrees Celsius no matter what liquid cooling was used.
The last Dual Pumps was worse but I attribute that to the fact that I had the 2 radiators (both) running at the same time, and one was in front of half of the other so I think there was some performance loss as a result.
I also wanted to note that all tests were done with e cabin open and the companionway ladder out of the way exposing the motor and controller. the wind was not consistent and some level of wind cooling would have affected these results. The Dual Pump test had the computer 360mm back on the controller and the transmission cooler was on the motor. I expected the controller to reduce slightly but it did not. The temperature was also about 5 degrees hotter so perhaps ambiant air contributed? I will test again in the fall.
The temperature gun as well would be inconsistent, however over the duration of the test would be +/- the same error to some degree across all results, so I think not in play as much as the wind.
Motor @1800 watts
The motor behaved as expected. I tested by taking the measurement on the dead center pin. This seemed to be the hottest part of the motor which would make sense as the bearing would be right on the center shaft.
Without any cooling, the temperature climbed fairly steadily and although in 1 hour was just over 60 degrees Celsius, I’d predict that within under 2 hours the motor would have been dangerously hot. I also observed the same temperature increase at a rate relative to the RPM of the motor, the faster the RPM, the faster the heat increased. At 500 to 750 watts, it would likely be a longer time to reach peak temperatures but not in scope of my tests.
With cooling, @1800 watts, the motor settled in at 40 degrees Celsius using the Mishimoto Transmission Cooler.
Recent Comments