Jason McNeil Posted August 19, 2016 Share Posted August 19, 2016 Got a Flir Infrared Camera to test control-board temperatures under the grueling back-and-forth motion. Inmotion say that their boards can go as high as 80° before doing the extreme >20° 'get-off' tilt-back. I wanted to create this condition and determine: - How long would it take to overheat the board? - What is the peak power and the corresponding voltage drop? - What is the intensity and localisation of this waste heat? - What happens when this 80° is reached? Are there some areas that exceed this? Can you damage the board or will it force you to dismount? - How effective is the heat-sink? Link to comment Share on other sites More sharing options...
Hunka Hunka Burning Love Posted August 19, 2016 Share Posted August 19, 2016 That's very cool! Are those two rectangular things on the right side of the board the capacitors? I didn't think they would heat up that much, but I can see why there may have been some issues with some capacitor legs on other brands. I wonder if there's some way to keep the caps cooler or whether there's a more sturdy capacitor that could be used. I'm just trying to think what can make these control boards bullet proof. Say someone is climbing a really steep mountain hill on a V5F+, what can be done to avoid tiltback at 80 degrees? If there is a large enough heatsink/fan, couldn't that help keep temperatures down even more so that the board never reaches an overheat situation? Ideally I would like to see an EUC that can never be made to overheat even under heavy duty conditions. It must be possible somehow to get that done? I guess it's like those old Model T's that you see blowing their radiators and overheating. We've come so far now that rarely do you ever see cars do that any more. Maybe we need some sort of ethylene glycol/water radiator to help move heat away more effectively. Link to comment Share on other sites More sharing options...
HEC Posted August 19, 2016 Share Posted August 19, 2016 17 minutes ago, HunkaHunkaBurningLove said: That's very cool! Are those two rectangular things on the right side of the board the capacitors? I didn't think they would heat up that much, but I can see why there may have been some issues with some capacitor legs on other brands. I wonder if there's some way to keep the caps cooler or whether there's a more sturdy capacitor that could be used. Did you mean hot? Those might be capacitors (see picture below and rotate clockwise 90 degrees) and the small coil / transformer. Capacitors can be indeed equipped with (half-)round radiators with cooling fins if needed. Link to comment Share on other sites More sharing options...
esaj Posted August 20, 2016 Share Posted August 20, 2016 50 minutes ago, HunkaHunkaBurningLove said: That's very cool! Are those two rectangular things on the right side of the board the capacitors? I didn't think they would heat up that much, but I can see why there may have been some issues with some capacitor legs on other brands. I wonder if there's some way to keep the caps cooler or whether there's a more sturdy capacitor that could be used. I'm just trying to think what can make these control boards bullet proof. The problem wasn't the heat, but bad solder-job on the capacitor legs. Even cheap electrolytics are usually rated for something like 105 degrees Celcius. Quote Say someone is climbing a really steep mountain hill on a V5F+, what can be done to avoid tiltback at 80 degrees? If there is a large enough heatsink/fan, couldn't that help keep temperatures down even more so that the board never reaches an overheat situation? Ideally I would like to see an EUC that can never be made to overheat even under heavy duty conditions. It must be possible somehow to get that done? I guess it's like those old Model T's that you see blowing their radiators and overheating. We've come so far now that rarely do you ever see cars do that any more. Maybe we need some sort of ethylene glycol/water radiator to help move heat away more effectively. Larger heatsink and "better" placement of the mosfets on the heatsink would probably help already. If you look at manufacturer datasheets / application notes, they usually suggest a star-like placement of the mosfets to keep temperatures steadier. Also using materials with better thermal conductivity than aluminum (like copper), but that's more expensive. Placing the mosfets nearer to the center of the heatsink would seem like a logical first step, typically they're at the edge of the heatsink, and the heat tends to "flow" in all the directions, so it "hits" the edge at one side. Also there must be insulation between the heatsink and the mosfets, as the metallic back"plate" of the mosfet is also (typically) connected to the middle-leg (which usually is the drain, at least for N-channel devices). One another thing I could think of would be to allow the high-side mosfets to make direct contact (without insulation) with the heatsink, as they're all connected to the battery voltage anyway. The downside is that the heatsink will then be at battery potential (up to around 67.2V ), which can create a health risk if you're fiddling with the internals of the wheel and the battery isn't disconnected... Personally, I'm not that excited of the idea of using water (or other liquid) cooling, as it makes the structure more complicated and can get damaged in a crash or such (imagine the liquid starting to leak). Basic air- (convection) cooling can still be made better in these. Link to comment Share on other sites More sharing options...
Rehab1 Posted August 20, 2016 Share Posted August 20, 2016 @Jason McNeil I must admit you go beyond the call of duty!! Unfortunately with all of the talent on this forum offering logical and reasonable solutions to reduce overheating of mosfets why aren't EUC manufacturers implementing them? Link to comment Share on other sites More sharing options...
MarkoMarjamaa Posted August 20, 2016 Share Posted August 20, 2016 No need for water cooling. Just put some pipes (copper?) inside that has outside air flowing. Maybe so that they get air flow from apparent wind. Plastic is not good a heat conductor, maybe just change some parts of cover with some metal? There are two problems, getting the heat away from capacitors (inside the box) and getting the heat outside. If you can lower the heat, it should help how long tha capacitors last. There's was some simple equation in net how heat affects lifespan of capacitors. I also have Flir, took some pictures of my tv and put some usb-powered Papst fans to hotspots outside just to be sure. Maybe get some heatpipes from old notebooks? Link to comment Share on other sites More sharing options...
Rehab1 Posted August 20, 2016 Share Posted August 20, 2016 Why can't the allen screws holding the mosfets onto the strip heat sink be removed, followed by installation of individual finned heat sinks? What about a mosfet cooler?Would'nt these improve heat dispersion? It appears the current heat sink for all 6 ganged mosfets offer minimal heat dispersion. Link to comment Share on other sites More sharing options...
lizardmech Posted August 20, 2016 Share Posted August 20, 2016 27 minutes ago, Rehab1 said: Why can't the allen screws holding the mosfets onto the strip heat sink be removed, followed by installation of individual finned heat sinks? What about a mosfet cooler?Would'nt these improve heat dispersion? It appears the current heat sink for all 6 ganged mosfets offer minimal heat dispersion. There's no reason they can't do it. EUC manufacturers just refuse to spend the $5 or $10 per unit it would cost for a proper heatpipe based cooler. Link to comment Share on other sites More sharing options...
Rehab1 Posted August 20, 2016 Share Posted August 20, 2016 @lizardmechIt appears to be an easy modification. I am going to order 6 and experiment. I have a Flir that just sits on a shelf gathering dust that was originally used to thermally inspect for heat loss in my home. @Jason McNeilCan you please provide me the dimensions of the mosfets used on the V5F+ control board..including the drain tab? Thanks Link to comment Share on other sites More sharing options...
Keith Posted August 20, 2016 Share Posted August 20, 2016 I'm not sure I'm understanding the direction this thread has gone in? As far as I can see, Jason did the worst thing possible in terms of generating excess heat and the Inmotion board stayed well below 80 degrees, at which point it should have gracefully encouraged the rider to stop until it cooled down, however it didn't get hot enough to see that. compare it with this guy, on a cheap Pinwheel doing the same thing as a "test" who couldn't understand why the supplier stopped talking to him after he blew his second control board! http://hoverboardfails.net/electric-unicycle-fail-pinwheel-t1f-during-a-simple-test-2nd-motherboard/ Link to comment Share on other sites More sharing options...
lizardmech Posted August 20, 2016 Share Posted August 20, 2016 4 minutes ago, Keith said: I'm not sure I'm understanding the direction this thread has gone in? As far as I can see, Jason did the worst thing possible in terms of generating excess heat and the Inmotion board stayed well below 80 degrees, at which point it should have gracefully encouraged the rider to stop until it cooled down, however it didn't get hot enough to see that. compare it with this guy, on a cheap Pinwheel doing the same thing as a "test" who couldn't understand why the supplier stopped talking to him after he blew his second control board! http://hoverboardfails.net/electric-unicycle-fail-pinwheel-t1f-during-a-simple-test-2nd-motherboard/ They really should be able to run at full load indefinitely. Motor inverters would be close to 90% efficiency. Assuming 1.5kw output you would have to shift 150W of heat, that's less than components in a PC and some video game consoles. Link to comment Share on other sites More sharing options...
Rehab1 Posted August 20, 2016 Share Posted August 20, 2016 @Keith If we took Jason's experiment and moved it to an outdoor setting with an ambient temperature of 90 degrees F /32 degree C would that make a difference? I road yesterday with a heat index of 100 degrees and was a bit worried practicing back and forth routines. And is 'back and forth' the most grueling test? Link to comment Share on other sites More sharing options...
Keith Posted August 20, 2016 Share Posted August 20, 2016 5 minutes ago, Rehab1 said: is 'back and forth' the most grueling test? Once a motor is running it acts also as a dynamo (or more correctly alternator with BLDC motors) this is what limits maximum speed, at some point the RPM generates a voltage equal and opposite to the applied voltage (the back EMF) and current (therefore also power) drops virtually to zero. On the other hand, at zero RPM the only limitation to maximum power is the internal resistance of the motor, controller and battery. Therefore accelerating hard from a standing start will pull more power from the motor than climbing a steep hill at (say) 10MPH because at 10MPH you have a significant back EMF limiting the maximum power. So, Yes, zipping back and forth is like constantly accelerating hard from a standing start and is the most gruelling test. I sometimes do it to try to improve my balance and revers riding but it is the only time I make damn sure I'm also running the app to monitor the temperature, I stop when it hits 50 degrees. The only thing that will draw more power than back and forth is trying to accelerate as hard as possible up a hill that is so steep all you can do is stand still whilst leaning as far as possible forwards. Link to comment Share on other sites More sharing options...
MarkoMarjamaa Posted August 20, 2016 Share Posted August 20, 2016 When your riding back and forth just for testing, there's also less apparent wind to cool chassis and electronics. Link to comment Share on other sites More sharing options...
Rehab1 Posted August 20, 2016 Share Posted August 20, 2016 Thanks for the knowledge! So much to learn? Link to comment Share on other sites More sharing options...
Xima Lhotz Posted August 20, 2016 Share Posted August 20, 2016 Thank you for this informative movie! To add to the discussion of cooling. Why not use peltier coolers? https://www.amazon.com/s?ie=UTF8&page=1&rh=n%3A172282%2Ck%3Athermoelectric peltier cooler So if temp > than X value start cooling. They are vibration proof, withstand high temperatures and can cool well below ambient temperature and there are no moving parts. They do consume power but controlling them with a PWM signal they can be used optimal. So if temp is below X degrees ex don't use them. And then progressively turn up the power (to cool the board). Or they can be used as a emergency measure if you have ample cooling on the board under normal conditions. This togher with "normal" cooling with heat sinks I think could be a good solution for board safety. Link to comment Share on other sites More sharing options...
Keith Posted August 20, 2016 Share Posted August 20, 2016 26 minutes ago, Xima Lhotz said: Why not use peltier coolers? Yes, but what exactly do you do with the heat you've removed , it's still inside the casing? Seeing as they move the heat less than 1/2 centimetre away and have efficiencies around 10-15% so will add a bucket load more heat of their own, I suspect they will make things worst rather than better. Link to comment Share on other sites More sharing options...
esaj Posted August 20, 2016 Share Posted August 20, 2016 Peltiers might maybe work if the "hot-side" was attached to a very large heatsink, that leads to outside of the casing... It could be more efficient than "normal" air-convection, as the element has more surface area to dissipate the heat? I've only fooled around a bit with a single 12V/6A peltier I've got just for fun, and the "hot-side" does get REALLY hot Link to comment Share on other sites More sharing options...
Jason McNeil Posted August 20, 2016 Author Share Posted August 20, 2016 A couple things I probably should have added in the original comments: The Wheel actually stood up well to this test, did a similar exercise on the E+ and it did not fair as well. I will be doing a comparison of this test with some different model Wheels later... A large heat sink is definitely important, but if the control-board is built to operate at higher temperatures, an extreme temperature gradient between the active heat-generating regions and surrounding surfaces, means that there is less likelihood of going into over-temperature condition in normal 'steady-state' riding, thanks to the second law of themrodynamics Here in Miami, the ambient temperature is sometimes >37°C. I've never experienced an over-temperature in the 750KM I've clocked on the V5F+ so far The picture Rehab posted is of the board after it background removal software lopped off the heat-sink, thinking it was part of the background... Link to comment Share on other sites More sharing options...
Hunka Hunka Burning Love Posted August 20, 2016 Share Posted August 20, 2016 I guess it depends on what extreme duty these wheel will be going through and what ambient temperatures they experience. In Dubai, I hear that temperatures go up to 40-50 degrees during the day. In Canada, we experience some 30+ C days. I like hearing that the V5F+ can withstand these upper temperatures, and maybe 80 C is a good enough upper range? I don't know. I do mostly steady state riding with not many hill climbs, and being 165 lbs in weight it would likely be more than enough for me. For those into stunts, who are heavier in weight, or who experience heat waves and hilly areas, it would be nice to have a wheel that just can't be pushed over the limit. With CPU's running those intensive PI calculation tests, the calculations stress tests the components for hours. They don't start throttling down or crashing if the cooling system and components are in good shape. Usually there's a peak temperature I think that occurs. I wonder if there's a maximum temperature that these EUCs can be held at to prevent overheating situations in all conditions or do they just keep heating up no matter what heatsink/fan combo there is? Link to comment Share on other sites More sharing options...
Rehab1 Posted August 20, 2016 Share Posted August 20, 2016 @Jason McNeil hey, that was one fine piece of artwork you removed? Link to comment Share on other sites More sharing options...
Jason McNeil Posted August 20, 2016 Author Share Posted August 20, 2016 3 minutes ago, Rehab1 said: hey, that was one fine piece of artwork you removed? The metal block from behind where you circled. Link to comment Share on other sites More sharing options...
electric_vehicle_lover Posted August 21, 2016 Share Posted August 21, 2016 @Jason McNeil thanks for sharing!! Great detail!!! I would like to share the information I have about the subject for the MicroWorks 30B4 board: https://github.com/EGG-electric-unicycle/documentation/wiki/MicroWorks-30B4---30km-h-controller-board-with-bluetooth#Max_temperature_protection Max temperature protection When the temperature achieve 65ºC, the EUC will make 2 quick beeps and tilt back the pedals. Here is a screenshot showing the app and the temperature on that situation: There is no temperature sensor on the board. I believe the system uses the temperature value returned by the MPU6050. The STM32F103 also have an internal temperature sensor but seems this sensor would need to be calibrated on production line. Link to comment Share on other sites More sharing options...
Xima Lhotz Posted August 25, 2016 Share Posted August 25, 2016 On 20 augusti 2016 at 5:42 PM, Keith said: Yes, but what exactly do you do with the heat you've removed , it's still inside the casing? Seeing as they move the heat less than 1/2 centimetre away and have efficiencies around 10-15% so will add a bucket load more heat of their own, I suspect they will make things worst rather than better. Well there are several options, but one would be to place it between the outside and inside heatsink in the new V8 for example. As an extreme situations cooler (or just use it when breaking energy gets to high), so waste it (energy) in cooling the board. And this would be good for countries where its a lot hotter (Dubai) as its a heat exchanger and it can then cool below air temperature. Sub 0 if you want it to. The coldest temperature can only be surrounding air so if you have 40 degrees of air temperature, the board won't get cooler than that with conventional cooling. But, of course a conventional cooling solution with no components and moving parts designed properly to handle all solutions is the first choice. Link to comment Share on other sites More sharing options...
Xima Lhotz Posted August 25, 2016 Share Posted August 25, 2016 On 20 augusti 2016 at 6:31 PM, esaj said: It could be more efficient than "normal" air-convection, as the element has more surface area to dissipate the heat? Yes, you can get a delta T of 50 degrees under optimal conditions. So you can go below freezing if you set it up that way. Going below 0 might not be optimal though, moisture and other not so nice things might build if its not controlled in the right way. Replace fan with the wheel turning inside the casing moving air around the heatsink (in the V8 for example) Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.