I just procured a pressure vessel for the purpose of test casting Motobrains under pressure. I expect this to solve my gas entrapment issues during the casting process. I just poured one with 20PSI. I let it sit under pressure for just a few minutes. At 20 PSI with 6 sq/in of surface area I was pushing 120 pounds down on that goo. Any air in there should have found it’s way out I would think. I invested no time or effort in the pour relying solely on the pressure to make a good part. I’ve been told that if I pour very slowly I can get pretty good results. I don’t want to spend a lot of time on this process so I am hoping that the job is easy with the right tool. We will find out tomorrow.
I just pulled the first demo casting from the mold. I’d say it looks good except for the large void. I am going to have to try casting it under positive air pressure to get rid of the void. Removing it from the mold was extremely simple. I needed to do little more than give it a stern look and the fake Motobrain leapt from mold. Many followers of this project offered advice to move toward a silicone mold and I am pleased to have listened. This is a great process!
I poured the silicone mold this morning. The process is quite fast. We shall see tomorrow how it looks. I am cautiously optimistic. My vacuum pump had a little performance anxiety today. While my tests showed it could achieve 30 inches of mercury of vacuum, it only got to 25 inches today, so that may be an issue. It looks fine though, not too many bubbles. This is just a test run anyway because the mold master is not perfect. I should learn plenty from this and frankly this only needs to be adequate. Perfection is something that my manufacturer needs to deal with. I am making samples and they can just be good.
The mold master I ordered arrived yesterday along with enough resin for a dozen more Motobrains. The silicone will arrive early next week and I will create the mold and try it out. I am not satisfied with the quality of the master. I used a new vendor and their quality isn’t as nice as I had hoped (good but not worth memorializing for all time). It was about half the normal cost so you get what you pay for. This is not a big deal because the part I have is adequate to learning silicone mold making and casting. Once I’ve got a process I like, I’ll order another master taking into account anything else I learn and we will be off to the races.
Expect some test castings next week. Perhaps even a real casting in the new mold if everything goes well.
I am going to make a silicone mold as the polypropylene mold has been a pain to work with. Most of the parts have been ordered. This is the master I’ve drawn up. It is much taller than is necessary, but it does allow me to make the pour a lot less of an exercise in dexterity. I can pour in the correct volume and then left gravity settle it down to the right volume. I am going to do it upside down and a friend suggested. This allows me to put some heat sink features and a legend directly into the Motobrain which ought to be super cool at the end of the day. Not including the tabs, Motobrain is 3.5″x2.5″x1.5″. The tabs can be sawed off.
I did another several hour test to see how a Motobrain cast in thermally conductive resin responds to high current loads. The test was similar to last one where I started at 50A and began taking regular readings and when I saw the temperature was stabilizing I turned it up 10A. The results are below. The maximum temps are much lower now than with the previous resin. This stuff is clearly allowing more heat to radiate off the device keeping the max temps in check (10 degrees lower inside and out with this resin compared to the previous) while also reducing the thermal gradients across the device which will improve MTBF in the long run.
The top was measured with the FLIR camera. The side was measured with a Fluke thermocouple. The internal was measured with a sensor on the MCU.
Here you can see the lab in the IR spectrum. The wires are straining under the current load. You can even see the two AC power lines I an running to the power supply to reduce the fire hazard.
This is a photo of the fake ignition switch I use. It is really a couple of series diodes that do a fine job convincing Motobrain that an alternator is running or not. All the current also flows through this and boy was it unhappy. 85C is HOT!
That is as hot as the Motobrain got over 2 hours of testing up to 90A.
I planned to do the test up to 100A but the power supply started to make “creaking” noises and I got nervous and shut it down. I am a bit intimidated by what Motobrain can do. 1300W is a ton of power! Still I did turn it up to 100A and took snap shot. It did the deed and didn’t burst into flames. Kudos to you Motobrain. The internal temp was 44C.
I even did a little cold temperature testing. I put it in the freezer for 20 minutes and then powered her up to see what would happen. No drama at 4C. I’ll need to do more of this but the preliminary testing I’ve done (down to -80C) has all been positive.
I finally have a little success in casting a Motobrain. The Motobrain looks great and the mold is intact! Unfortunately, I made two significant changes compared to the last attempt so I do not know for certain which change was responsible for the better outcome. I changed both the resin (because I ran out of the preferred resin) and the release agent. As I said last night, the release agent I used was cheap floor wax (and AquaNet hair spray). The resin is not preferred because it is harder and not thermally conductive. The customer I will send this unit to doesn’t require high current capacity so I decided to make use of this resin rather than waste it.
I am still going to give a silicone mold a try but I won’t have to wait on it which pleases me greatly!
I waxed up my remaining mold with some cheap floor wax and just finished casting another Motobrain. I am hoping that I can remove the Motobrain from the mold without destroying it (unlike the last two Motobrains I cast) so I can keep making these in the near term while I come up with a new silicone mold. I have been advised to go ahead and figure how to work with silicone. I’m planning to go ahead and do it.
In other news, I’ve been refining the firmware and am really getting pleased with its function and reliability. I have heard also from the first recipient of a Motobrain and they are liking it so far. That is great news I’d say.
I have no good reason to become proficient at casting. I want to be able to do it when necessary but I have no grand intention of becoming an expert at it. Still, I ought to be able to cast a damn Motobrain without it being a pain in the ass. Alas, it was once again a total pain.
The bottom of the mold popped off without any fuss leaving a mirror finish and I was excited to pull the sides off and see a pristine new Motobrain come out. No amount of prying helped release the sides though. I ultimately had to chisel the mold off at the cost of 2 hours and $350 for the mold which is currently in more than 400 tiny pieces. In two attempts, the new mold release did not work at all on this mold material. I’m done with that stuff. It is worthless, expensive or not. I’m going back to the $8 Home Depot floor wax. That stuff actually seemed to work at least.
Chiseling the Motobrain out of the mold did a bit of damage to the resin but it is still in good enough shape for my needs. I didn’t do a great job making sure the cap was completely in the resin and you can see a few voids just below it. I did try but I couldn’t seem to push it down far enough and I didn’t want to make a mess by pulling it off and trying again. I still have some more trial and error to suffer through before I can get these guys packaged up nicely. I feel like I am on the right track but I am not quite there yet.