MOSFETS overheating

[mcguyver]

That wasn't directed at you Mr. C, but rather at Kos-Mos.

In "theory", a mechanical switch will exhibit zero loss, as there is no semiconductor present, and metal-to-metal contact has virtually no DCR. However, enter into the mix contamination from lubricants, arc damage, poor contact pressure and the resistance builds up pretty fast. I don't know if you could really do an accurate test, as each gun will change from brand new over time, and no 2 guns will ever be the same, especially when the mechbox has been upgraded, relubed, etc.

But it's fair to say from new with the lower current draw of a stock gun (300ish fps) and no switch contamination, the standard mechanical switch is not bad enought to warrant the installation of a FET at all, and with lower voltages as you mentioned, is of more harm than good.

[Naglfar]

Originally Posted by mcguyver

But it's fair to say from new with the lower current draw of a stock gun (300ish fps) and no switch contamination, the standard mechanical switch is not bad enought to warrant the installation of a FET at all, and with lower voltages as you mentioned, is of more harm than good.

Quoted for truth, this is the exact reason I decided not to venture into designing FET circuits. Though I would like to finish work on designing an electrical 3-round burst setting.

[WingZER0]

Originally Posted by MadMax

Stuff the Guru said

I don't know if they're actually better persay. Main reason I add FETs in is to prevent the switch from arcing (my P90 had severe issues with that, end result being repeated double or triple fires when in semi automatic, and the corrosion on the switch was pretty severe). That's why I perfer FETs, despite lost voltage. 3 FETS in parallel, ironically enough, was to prevent any 1 FET from overheating due to high current load primarily.

And the gun has been upgraded, just that I run it with the stock spring for testing. Plus it's also for incase the switch itself is ever damaged or loss. With a fetted switch I can simply get an over the counter 1/4amp rated switch 2 minutes to pickup from a local store instead of having to go to CA for another 14A switch that for some reason I cannot find normally (if ANYONE has found one of those switches that I can source right from Canada at non stupid prices, please let me know. I'm still looking).


And they're right: Brown black bown gold (10 x 10^1 ± 5%), and Red red red gold (33 x 10^3 ± 5)

[eel one]

Originally Posted by MadMax

RC cars and airplanes depend on FET switching to accomplish PWM speed control, so do cordless power tools, but we don't need this feature.

You might not what the feature. But some of us do. In the latest unit I installed, I have PWM speed control to lower ROF if I need to (not all gaming sites condone high speed guns) + the unit use the PWM to soft start the motor in order to prevent arcing of the brushes and have less of a shock start on the gears. To the user there is no noticeable delay in trigger response even on a 8.4v battery pack and a EG1000 motor. It even saves juice on the battery giving you more shots per recharge.

By using a switch computer you can also get rid of that old fuse and use a more intelligent solution that actively monitors the power drain of the system.

Now I won't say that MOSFETs ain't voodoo to most people. But even on low ROF, low power guns I have not noticed a drop in performance by installing a FET. But I have old guns with worn switches so that might explain why.

[mcguyver]

So, eel, how do you "soft start" a DC motor required to do cyclic loads and be able to do repeated "soft starts" in semi? Unless you don't use semi-auto?

The only thing I can think of is varying frequency, but there is a frequency of 0 Hz at motor start. This doesn't work like an AC motor, so normal AC soft start systems would not be effective I think.

[CDN_Stalker]

My minor two cents worth, but I'm thinking he used three MOSFETs (which I have VERY little remaining knowledge or experience with) in parallel in order to make up for the lower ratings the ones he found. In my mind as far as an analogy goes, it's about the same as using three #22AWG wires (rating of 5A max in a confined space EACH) in parallel to be able to safely handle 15A of current across the three, net result is overheating but still being able to pass the current.

[mcguyver]

Originally Posted by CDN_Stalker

My minor two cents worth, but I'm thinking he used three MOSFETs (which I have VERY little remaining knowledge or experience with) in parallel in order to make up for the lower ratings the ones he found. In my mind as far as an analogy goes, it's about the same as using three #22AWG wires (rating of 5A max in a confined space EACH) in parallel to be able to safely handle 15A of current across the three, net result is overheating but still being able to pass the current.

Basically. But if 1 of those wires fails, then the other 2 must carry 7.5 amps each, or 50% over their rating. This is why parallel wiring is forbidden in the Canadian Electrical Code, except for service feeders.

[CDN_Stalker]

Originally Posted by mcguyver

Basically. But if 1 of those wires fails, then the other 2 must carry 7.5 amps each, or 50% over their rating. This is why parallel wiring is forbidden in the Canadian Electrical Code, except for service feeders.

Good, I fully understand his issue then.

[eel one]

Originally Posted by mcguyver

So, eel, how do you "soft start" a DC motor required to do cyclic loads and be able to do repeated "soft starts" in semi? Unless you don't use semi-auto?

The only thing I can think of is varying frequency, but there is a frequency of 0 Hz at motor start. This doesn't work like an AC motor, so normal AC soft start systems would not be effective I think.

Like I said, by use of the PWM speed control. It is used to cut down the current to the motor. We know that the start up of the motor is the point in the cycle where there is the most drain from the battery. Now the normal switch system of an AEG is off-to-full power. That is like taking the engine on your car to max revs and dumping the clutch, not an optimal solution. So by cutting down the power to the motor to a lower level, until the motor starts to turn and the drain drops, you also significantly cut down arcing. Ok truth be told the SA response time is 1/10th of a second longer than usual.

For more info go to airsoftmechanics.com and read this thread:
http://forums.airsoftmechanics.com/i...hp?topic=648.0

At about page 9, soft start begins to be debated and documented onto page 11.

Remember that everything on this unit is computer controlled.

[mcguyver]

Everything is a trade-off. To give the same start-up torque, you would need to increase voltage if you decrease current. Simple Ohm's law. And, if you factor in a reduction in output, then this is feasible. But to have equal ROF and reduced current without increased voltage is not possible. So, you would need a step-up power supply, hence your pulse-width modulation.

But, that level of sophistication is way beyond your basic user, and even all but the most serious of tinkerers.

[eel one]

That is why I am a beta tester of the unit, befor it gets released for sale. I come up with ideas, give them to Terry at www.extreme-fire.com he picks what is usefull, programme the unit and sendt it to me. I'll then test it and give feedback. Terry then edit the programming, download it to a new chip and ship it to me. I then swap to the new control chip and redo the testing (and around we go). I have no knowedge of programming in C or putting together the unit. BUT if you want to give it a try. Terry have it all layed out as open source, blue prints, component list, programming script, everything. Go knowk your self out:
http://extreme-fire.com/SW-COMPUTER/SW-COMPUTER-2.4/

[mcguyver]

I can barely retrieve my e-mails, let alone program anything.

I did manage to program my VCR back in '87!!

[eel one]

Well all I can say it that the soft start works as well as all the restrictions that are possible through the unit (lower ROF, 3-round burst, semi only, 3-r-burst to full auto). Also all of the safety mechanisms work, like high voltage cut off, high current cut off, high peak current cut off, low battery cut off, stuck trigger doing power up also results in the gun refusing to fire. I have not been able of get the unit to over heat but it is supposed to also shut down in that event.

[Kos-Mos]

what chip is he using (I assume some MicroChip PIC microcontrollers, since it is the easier to get)

Right now I only have hands on a PIC16F876 and a custom programmer for it.
Plus the size factor prevent me from using it in an airsoft gun.

I am working on a very simple, pot-controlled PWM to use with some 11.1v Li-Po and a FET half H-bridge. I am looking at some very basic microcontroller that is small enought to do that job. (Maybe something DIP-8 package or smaller, like SO-8)

[eel one]

Atmel ATTINY85 microcontroller