No assist when warm
|
| |
Four of the guys reported that with the hot weather, their MIMA systems were loosing assist. Careful observation by Nemystic, and some testing by Hafnhaf with some freeze spray identified that it was a thermally related issue. I could not duplicate the problem on my test fixture, so I made a mount for the MIMA board where I could access it while driving. Nemystic brought his card down, so I would have one of the bad boards to work with. With a can of freeze spray in hand, I quickly found that Q6 the switching transistor that drives the MAMODE1 20KHZ PWM signal was the thermally sensitive component. The problem was that the MAMODE1 signal showed no measurable change between working and not working. Further testing showed that the gain of the 2N3704 transistor being greater than 100 would cause the assist to stop. Cool the transistor which drops the gain, and it works fine. I pulled out my heathkit semiconductor curve tracer, and was able to clearly see the gain change. I went through all of the transistors in my stock, and found that more than half of them had gain higher than the 100. I selected one of the transistors with a gain of 70, and replaced the part on Nemystics board, and the board became stable with no problems even when baked to 122F in the closed car. To finish the confirmation test, I selected a transistor with the highest gain of over 200, and put it into a new board that was not having the problem, and sure enough, assist stopped working as soon as the temp got to 80F. Now I need to come up with a board patch that will allow any transistor to work reliably. Further testing with the new boards I just built confirmed that this fix works well, so if you have this problem, and want to fix it your self, just cut the emitter lead, bend it up, and after cleaning the hole in the board install the 150 ohm resistor and you should be good to go.If you want me to fix it, send the board back and I will turn it around asap.
|
|