| Author |
 Topic  |
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
 Posted - Apr 25 2009 : 4:39:13 PM
|
A TL084 and many other normal quad opamps have an output of 30V p-p at 100kHz.
The low power LM324 has an output of 28V p-p at about only 1kHz. It can't produce 28V p-p at a higher frequency because the output is slowly ramping instead of switching quickly.
A 2N4401 and a 2N4403 have a max allowed current of 600mA and would make a good Mosfet gate driver. |
 |
|
|
boron
Mad Scientist
Canada
223 Posts |
Posted - Apr 25 2009 : 8:40:18 PM
|
Well we tried a test this afternoon in which we replaced the lm324 with the TL084 and tried it without the transistors, the output of the TL084 was directly fed (through a resistor this time) into the gate pins of two parallel IRF540N's.
The result was that the scooter accelerated under load when throttled up. The mosfets worked lovely and did not blow. The driver was able to accelerate and decelerate no problem. After the ride (which lasted maybe max 30 seconds) the mosfets were slightly warm (although contact with the heatsink wasnt perfect and has since been corrected). So now that it works very nicely, will adding the transistors bring any extra benefits? ie. will the transistors maybe keep the mosfets a little cooler?
Thanks a lot for all your help audioguru. I hate to say it but you were absolutly right about the lm324  |
Edited by - boron on Apr 25 2009 9:40:51 PM |
|
|
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
Posted - Apr 26 2009 : 07:41:20 AM
|
I am glad you fixed it.
The transistors will allow the Mosfets to be a little cooler but might not be worth it. Try them to see. |
|
|
|
boron
Mad Scientist
Canada
223 Posts |
Posted - Apr 28 2009 : 08:03:34 AM
|
I'm afraid I may have done something else stupid this time but do not know what... 
After that successful test, in which the mosfets were held down to the same heatsink with an elastic band, we decided to make the mosfet assembly neater and more permanent. We took two new mosfets and drilled holes into the heatsink (its like a small cpu heatsink)we then took two screws and screwed the mosfets through the hole in the top of their casing into the heatsink (gently) so they would not move and be secured. We then placed thermal paste between the backs of the mosfets and the sink. We realizaed that the screws and sink would cause the backs of the mosfets to be linked together through the sink but since they are in parallel figured it didn't matter. We then soldered with the appropriated gauges of wire the two mosfets in parallel and tested them with a multimeter before mounting in the scooter.
Here is what happened: When the rider gets on and tries to accelerate, it seems as if he has lost all power, moving along barely with any speed. If the back wheel is lifted (no load) the pwm works very nicely and the wheel seems to spin fast. With the rider, the mosfets do not blow, they don't short or smoke or anything, but as I said there is almost no power and the rider moves at a pitiful walking speed, not even.
The only difference between the test that worked and all subsequent tests was the mosfet mounting... but even after undoing that the problem now seems to persist...
I am now very confused as to what I did to cause the power to drop off like that? I have checked wiring and circuit boards and all seems fine... and its not the batteries because they were recharged and can propel a rider no problem. Any suggestions you have will be greatly appreciated yet again.
Thanks. |
|
|
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
Posted - Apr 28 2009 : 09:50:15 AM
|
The old Mosfets worked but the new Mosfets are very weak. Then something is wrong with the new mosfets?
Or you have a poor electrical connection to both drains? |
Edited by - audioguru on Apr 28 2009 09:51:01 AM |
|
|
|
boron
Mad Scientist
Canada
223 Posts |
Posted - Apr 28 2009 : 2:30:17 PM
|
Yep that's what I thought. I'll double check all the connections and mosfet's next time I get a hold of the scooter.
We may also think that potentially the diode across the motor may have blown which would cause the current to be diverted, reducing power, but the pwm would still function. I shall try to check it asap. If this is the case, what should be the minimum specs for this diode? |
|
|
|
audioguru
Nobel Prize Winner
Canada
4218 Posts |
Posted - Apr 28 2009 : 3:04:32 PM
|
| The current rating of the diode must be the same as the max current of the motor. |
|
|
|
CurrentOverflow
Mad Scientist
Canada
311 Posts |
Posted - Apr 28 2009 : 4:12:27 PM
|
| And it must be able to switch as fast as the pwm frequency right? |
~Mike~
Theres 10 kinds of people, those who understand binary and those who don't |
|
|
|
boron
Mad Scientist
Canada
223 Posts |
Posted - Apr 29 2009 : 8:45:14 PM
|
| Alright so the diode is still in perfect working order. (turns out I bought one capable of handling wayyy more then enough current) However I believe the culprit has been identified... we were using alligator clips for a test and the clip connecting the middle pin of the pot to the input of the op amp was creating a poor electrical connection, it was loose and conducting/not conducting with the slightest bump... hopefully that's the whole reason for this, I'll try it out again tomorrow... |
|
|
|
Topic |
|
Electric Motor and PWM
