MOSFET High Resistance

[Retroanaconda]

Struggling to get my head round this one chaps.

Situation as follows:

• Star L85A2
• Standard (ish) spec - TM EG1000 Motor, M110 spring, Guarder gears/shims
• 8.4V 3300mAh NiMh battery
• 16AWG silicone wiring
• Deans connectors
• AirFET non-braked MOSFET

I installed the MOSFET, ran a low-gauge wire from the battery positive connection on the Deans into the gearbox, through the trigger switch, and back out to the MOSFET. The larger motor wires do not go into the gearbox at all, positive runs straight from the Deans to the motor, and the negative from the Deans to the 'FET and from there to the motor. This was done so as to A: avoid the need to run the heavier gauge wiring through the gearbox and B: to reduce connections to a minimum to avoid unnecessary resistance and the associated voltage drops.

The problem I've found is that the rate-of-fire is actually lower now the MOSFET is in the circuit. It seems to add a lot of resistance to the setup. I have recorded a video to illustrate the problem as describing it just in text is difficult:

http://www.youtube.com/watch?v=nQ7W-zUI7Fg

The first part of the video shows me firing the AEG using the trigger to power the MOSFET. Then I short out the pins on the 'FET so as to provide a direct current to the motor (note I do this close to the body of the MOSFET so as to include my (not pretty, but functional) soldering in this circuit) and you can see there is a marked increase in the ROF. Measuring the voltage at the motor during firing, using the trigger gives just over 5V. Shorting it directly as above gives over 7V.

An interesting point is if I take a jumper wire and bridge between the battery positive and the trigger (gate) pin of the MOSFET, bypassing the trigger switch, it fires at full speed. Does this mean that my trigger is not supplying enough current to the MOSFET for it to then supply full current to the motor? There is only a resistance of 7.5 Ohms across the trigger contacts. The switch is a microswitch, could this be a symptom of the switch bouncing, causing the MOSFET to turn on and off very quickly resulting in a poor connection? The 'FET does get pretty hot.

Any help graciously recieved. This is my first foray into MOSFETs, if it turns out to be incompatible with the microswitch trigger then I'll just pop a new microswitch in and call it a day, but I'd rather have the 'FET for efficiency's sake.

[Kingluke23]

Id remove the contacts and give em a clean with some wet and dry. I would have thought there should be a dead short across the contacts once made. Could be wrong. Cant hurt though. My L85 had severely burnt contacts when I got it so I made some slightly thicker ones out of copper. Bit of a pain but it never happened again.

Sorry just read it was a microswitch....bang a new one in or pull it apart and clean the contacts.

Sent from my GT-I9300 using Tapatalk 2

[Gecko]

Re: MOSFET High Resistance

The best wire is 14 awg PVC very low resistance

[North West Outpost]

Re: MOSFET High Resistance

If there are three wires, the single thin wire will come off the Negative centre pin of the transistor. If you want to use another thin trigger wire you will need to splice it from the Positive thick wire. When I make my fets I do it at the same place the corresponding pin from the transistor connects to the Red wire. I guess you are not wanting to strip the shrink tube off the fet so as near as possible should still be fine.

Then just use the two thin wires to trigger contacts and thick 16awg go to the motor. Nice soft silicon coated is fine, if you can fit 14 it may be better but if not super tuned doubt you will notice the difference, except trying to get it through tight spaces.

[Gecko]

Re: MOSFET High Resistance

With the new MOSFETs I highly recommend not using the middle pin the FETs are very sceptable to heat so I normally use a crimp for the drain wire and screw it into the back, and 14 awg PVC is smaller than 16 silicon so it's alot better.

[Retroanaconda]

Re: MOSFET High Resistance

Thanks chaps. The wiring is not the limiting factor as bypassing the 'FET (by shorting the source/drain pins) gives a very good rate of fire.

The issue is only apparent when using the trigger contacts to trigger the MOSFET. If I take a wire from the battery positive (spliced at the Deans) and touch it to the gate pin of the MOSFET then I get the same good rate of fire as if I bypass it completely as above. So I don't think the 'FET itself is at fault, nor is the wiring or soldering.

Another thing I noticed was that when bypassing the trigger contacts as described above I can fire it for ages without the MOSFET even beginning to heat up. Firing for maybe 10 seconds using the trigger results in a hot MOSFET! I think the answer lies here:

http://extreme-fire.com/Micro-Switch-Bounce.pdf

This would explain the poor perfomance and heating up of the MOSFET, as the bouncing of the microswitch contacts causes the 'FET to switch on and off very quickly (I know that document pertains to AB 'FETs but the same theory applies to standard ones just with less effect). I will try sourcing a very lightweight switch as per the above document (0.1A rated current - as opposed to current 15A rated switch) and see if that solves the problem. As some of you guys have suggested it could probably do with a new switch anyhow as the current one is getting on a bit.

[dave38x]

Re: MOSFET High Resistance

You can use a capacitor to "de-bounce" the switch if needs be, I couldnt tell you the circuit but I remember studying it once.

[Retroanaconda]

Re: MOSFET High Resistance

That's an option, as is building an astable circuit with a 555 timer to created a timed pulse for each trigger pull.

I've got another micro switch laying about here that I bought a while ago, a 20A one I think, I'll pop that in just to rule out the worn contacts in my current switch. If the problem still exists then it must be the bouncing induced by gearbox vibrations, and I will look into de-bouncing.

[North West Outpost]

Re: MOSFET High Resistance

ARGH....My minds gone onto overtime, thinking up ideas for different bits and bobs.

[Retroanaconda]

Re: MOSFET High Resistance

Well I put the new micro switch in, but unfortunately in the process of disassembly I've managed to snap one of the legs off the MOSFET. The good news is that the new switch is a lot more sturdy and has a far more positive switching action, whether this makes it any more vibration resistant I don't know. Time will tell.

Ordering a new MOSFET. I wish I still lived near an RS trade counter, major ball-ache having to order stuff online

[No1_sonuk]

Re: MOSFET High Resistance

If there are three wires, the single thin wire will come off the Negative centre pin of the transistor. If you want to use another thin trigger wire you will need to splice it from the Positive thick wire. When I make my fets I do it at the same place the corresponding pin from the transistor connects to the Red wire. I guess you are not wanting to strip the shrink tube off the fet so as near as possible should still be fine.

I don't understand the bold part.
A non-braking MOSFET usually uses an N-channel MOSFET and goes in the -ve line of the motor circuit, and shouldn't have a red wire at all.

Unless.. Do you run the red wire through the MOSFET heatshrink and take a tap off it for the second trigger wire? Essentially making a 6-wire unit.

As for the switch contacts:
MOSFETs are voltage switched, not current, so as long as the voltage is higher than the threshold, current is immaterial.

[Retroanaconda]

Re: MOSFET High Resistance

Well in theory the MOSFET could be in the positive supply side rather than the negative ground side. I don't think it would make much difference functionally.

On mine the small trigger wire is fed from the battery Deans connector, that way only the light gauge wiring is going into the gearbox so it's nice and simple.

[No1_sonuk]

Re: MOSFET High Resistance

Well in theory the MOSFET could be in the positive supply side rather than the negative ground side. I don't think it would make much difference functionally.

It would make a lot of difference.

The Gate line of an N-channel MOSFET needs to be pulled to the threshold voltage (Vth) ABOVE the source voltage.
e.g if Vth is 5V, and the source pin is connected to the battery -ve (drain to motor -ve, motor +ve to battery +ve), a voltage of 5V or more on the gate will turn it on.
When it is on, the source voltage is as near as makes no difference battery -ve, so the MOSFET stays on.

If you put the N-channel above the motor (drain to batt +ve, source to motor +ve, motor -ve to batt -ve), your required threshold becomes batt +ve + Vth. So for a 7.4V battery, the gate would need to be at 12.4V minimum.
In practice, you'd use a high-side driver to do that, which makes the circuit more complex, or use a P-channel MOSFET.

The P-channel MOSFET would go on the high side, and use a -ve gate voltage, so connecting the gate to the battery -ve would turn it on.

BUT. P-channel MOSFETs don't have as low on-resistance as N-channel. That's why N channels are used low-side.

Braking MOSFETs use an N-channel for the normal motor on running, and a P-channel to short the motor for braking (connecting the battery +ve to motor -ve after turning off the N-channel).
If the N-channel doesn't get turned off quick enough, the P-channel going on will cause shoot-through, or "fuse-test mode" by shorting the battery. This would not be a good thing...

[Retroanaconda]

Re: MOSFET High Resistance

Ah I see, fair enough. My knowledge of MOSFETs is based on 6th form electronics which was a good while ago now

My new units arrived from RS today so I will plumb one in this evening and see what the result is with this new microswitch.

[Retroanaconda]

Re: MOSFET High Resistance

Back together and it cycles nice and fast now. So I guess the old switch with it's far weaker bias spring was allowing the contacts to bounce which was messing up the MOSFETs operation. No heating of the 'FET now on full auto so it seems to be solved. For now

Thanks for the help all.