Active braking system

[RiotSC]

Some people were interested in the active braking system (which prevents gears from overspinning), so here is one of my earlier implementations with the V2 mechbox.  The concept is quite simple.  When the trigger switch is at rest, it shorts the motor, thereby turning the motor into a generator. The kinetic energy from the gears' residual momentum is then turned into heat.  Note that without using a FET device, V2 mechbox would not have enough room for the wiring.

Here is a close-up of the modified trigger switch:



I used sheet nickel as the active braking contacts.  I made one contact on the switch block (contact 1, shiny silver colored metal on the left), and on the switch block retainer one (contact 2, shiny silver colored metal at the center).  Contact 2 is bent into L shape at the rear to provide contact for contact 1.  Contact 1 is connected to the switch contact on the right (contact 3) via a copper braid (shown at the top).  The split wire from the motor positive line is then connected to contact 3 (shown at the right).  Finally, the split wire from the motor negative line is connected to contact 2 (shown at the lower center).  The thin blue wire at the bottom is the gate wire for the FET device.

Here is a view of the wire layout:



The motor wires are both split to provide the necessary wiring for the brake contacts.  The wires at the bottom of the mechbox should be tucked to the side neatly and thereby not interefere with the motor pinion gear.  Depending on the thickness of the wire coating, it may be possible to use wiring up to 14AWG.  16AWG is probably the highest for most though.

This implementation also requires modifications to the right shell of the mechbox.  Some material has to be removed to provide room to accommodate the extra thickness from contact 2.  The black electrical tape shown in the picture is to prevent the mechbox from shorting with the contact.

For people with systems that offer spring release function, you can simply add a diode to the motor, as shown at the bottom of the picture.  While the diode does nothing for overspin, it does provide active braking when the gears travel in reverse (when spring release function is activated).  This will prevent the sector gear from striking and damaging the rear of the piston catch tooth.

Final note: the implementation shown was one of my first attempts at the active braking system.  Later versions are less crude and do not require mechbox shell modifications.  The only reason I showed this initial version is due to this specific AEG required servicing.  So I snapped a few pics while it was open to give people some ideas.

Well, that's it.   Have fun.  :wink:

[Joaz]

Hm.. my question is: At which position will the gears rest on semi? If would be bad, if it stops too early, especially on semi-only guns.

[RiotSC]

The motor starts braking the gears as soon as the trigger switch block returns to resting state.  That is solely determined by the mechanical interaction among the sector gear cut-off cam, cut-off lever, and the switch block.  Assuming you have not modified the positioning of the cut-off cam (most aftermarket gears would not have allowed you to do so anyway), the braking would happen as soon as the last tooth on the sector gear releases the piston.  Of course, if the switch block return spring has weakened, the braking would also be delayed.

[cjfield1129]

Would a Version 3 mechbox be setup differently? You mention that later braking versions are less crude requiring no gearbox modifications. Could you elaborate or snap some more pics of your lastest braking installation. This all seems like a ton of work/modification to fix something that really shouldn't be a problem. I guess I'm not too knowledgeable to discern if this is just a pure physics issue or that of QC in the cutoff lever/switch block assembly.

[RiotSC]

Basically, to utilize this design, you need to install additional contacts behind the trigger switch block to use the return mechanism for creating a motor short circuit during the system's rest state.  This will prevent the gears from overspinning by braking them when the switch is cut-off.  So the nozzle and the piston would no longer be retracted after every semi shot, which is really useful for semi only AEGs.

It is indeed a ton of work to fix something that shouldn't be a problem, but Tokyo Marui never inteneded for people to modify its guns.  When people start using aftermarket parts and high voltage battery packs, the mechanical system reacts differently.  As I explained in the earlier posts (might be other threads), overspinning is caused by the gears' high residual momentum due to higher rotational speeds caused by high ROF and/or high power setups.  In high ROF setups, the high voltage battery pack in combination with low ratio gears make the gears spin at such speeds that the gears simply retain enough momentum after cut-off that they'll just continue spinning.  In high power setups, the built-up energy from overcoming and releasing the piston creates enough acceleration for the sector gear during its no-load half-cycle that it also retains enough momentum to continue spinning after cut-off.  In both situations, the gears continue to spin until the energy is dissipated, either on their own or by the main spring's resistance.

As for V3 mechbox's actual implementation, the main difference is that I routed the nickel contact for the trigger block (contact 1) from the bottom side onward to the block's rear.  For the retaining contact (contact 2), I routed a wire from the front to the left wall next to left switch contact.  At the very bottom of the left wall, the wire is connected to the retaining contact, which extends all the way to the rear and makes a L-bend at the trigger switch block's fully extended rest state.  There is no mechbox shell modification for the V3 implementation.  I know it's difficult to visualize without some picture aides, but you'll need to wait till I open up a V3 for maintenance for those photos.

There is an easier approach by using a normally closed automotive relay.  The main problem is actually finding a reliable one.  As some of you know, many of those automotive relays can actually turn on due to accidental bumps or shakes.  That can cause major problems for our applications.

[kchmon]

Could you draw up a simple schematic. I'm having a hard time trying to visualize where all you are connecting contacts and whatnot.

My motor gear shorted my FET and melted it the other day. I thought if I'm going to put in a new FET, (and a FUSE this time!), I might as well try a braking system because this is a sharpshooter rifle.

What kind of extra battery life would get be getting since the motor is acting like a generator? I'm thinking only 10% at the most.

[RiotSC]

Hopefully that helps.  (I also hope Jay doesn't mind my borrowing and modifying one of his article pics...  )  I basically added additional contacts at the rear of the switch block to add the short circuit path.

Also, you do not get any additional battery life out of this setup.  All it is doing is turning the kinetic energy from rotational momentum into heat, thereby acting as the brakes.

[PsyphyerVII]

I'm confused by the diode you placed on your motor. If the motor is shorted when the trigger is at rest, doesn't it provide active braking for both directions? Wouldn't the diode become useless since electricity can travel both ways on the short?

[RiotSC]

Originally the reversed diode was meant to protect the MOSFET against the surge from the motor at shut-off.  However, since the reversed diode also shorts the motor when the motor spins in reverse, it provides an easier option for guns with spring release function than to implement the whole trigger switch modification.  While the reversed diode does not provide braking during normal cycling, it does so when the system travels in reverse with the spring release activation.

[PsyphyerVII]

Actually I ment to say "why do YOU still have the diode now that you've added the brake contacts"? The short from the brake prevents overspin in both directions already. With the diode, if you were ever to connect the batteries in reverse polarity (somehow), and pulled the trigger, there would be a short circuit.

Anyways, I've been wondering, wouldn't using a p-channel mosfet be easier for the active braking system? The p-channel mosfet is active when the gate is at ground and unactive when the gate recieves positive voltage. Please verify if this will work or not:

[RiotSC]

Didn't you post this on Sunday?  When I had time to answer it on Monday, it was gone.  Sheesh, and I thought I was seeing things due to fatigue.  

Anyway, the diode is soldered on and serves its original purpose of surge protection against the motor at shut-off.  So I left it alone.  There is a short time gap for the switch block between disengaging gate contacts and engaging brake contacts.  So better safe than sorry, considering electric field travels at near light speed.  :wink:

As for the second MOSFET idea, I initially thought about a depletion mode (normally-closed) MOSFET, but they are generally low current and slow.  Looks like your circuit might work though with the P-channel enhancement mode MOSFET.  The only downside is it'll continue draining your battery during rest states, so better unplug when not in use.  You'll also need to make sure that its operation times are faster than that of the main power MOSFET.  Otherwise you'd be looking at one heck of a short circuit.  

[PsyphyerVII]

Hehe, you got me there . I deleted my post couple of minutes after I posted it because I was still thinking of other possible methods. But reposted when I couldn't think of anymore.

As for the depletion mode mosfet, don't you mean it's normally open? Once it detects a voltage (-V for pnp, +V for npn) the center material gets depleted, creating a gap which effectively closes the mosfet. But you're right, depletion type mosfets can't handle much current, and are thus are rarely used or sold.

As for the battery drain, unless I'm mistaken, mosfets are designed to be activated by voltage, not current. An example of a gate that requires current to activate is JFETs, but the main problem with current activated is as you said, they continue draining current. MOSFETs where designed to only require voltage, therefore getting rid of current leakage. This would allow the battery to stay connected without a problem.

In the schematic, the p-channel mosfet's gate is connected to the black wire/negative voltage, which turns it on and shorts the motor. When the trigger is pressed, positive voltage from the red wire takes over and turns off the mosfet, preventing a short circuit. Hopefully my li-poly battery doesn't exploded when I try this...

[RiotSC]

You are correct that MOSFET ideally passes no current through the gate, but leakages do occur during operation.  There is also the subthreshhold leakage between source and drain.  So technically the MOSFETs will continue to use power at rest state.  That is why it is always a good idea to disconnect your battery packs from the AEGs when not in use.

As for the depletion mode MOSFET, they are indeed normally-closed, AKA normally-on or shorted.  Normally-open is the exact opposite.

Well, do keep us updated with your experiment.  If you find a suitable P-channel MOSFET, do share the info as well...    I'd try with a cheap NiCD/MH pack first though since damaging a Li-Po pack is expensive and potentially dangerous.

[PsyphyerVII]

Augh, I'm always making that mistake when speaking about transistors, saying open=on and closed=off, because I tend to visualize it as a real gate hence the open=stuff goes thru, and closed=nothing goes thru (although it DOES make more sense that way)  

I'll be ordering a couple of p-channel mosfets tomorrow and will be testing this out. Only thing that bothers me is the nano-second short-circuit that might occur if the n-chan mosfet is activated before the p-chan mosfet. Someone recommended a "phase-splitter transistor", which can "delay" a current, but I've had no luck finding such a thing.

[PsyphyerVII]

SUCCESS! The p-channel mosfet did the trick. With my mosfet relay + active braking circuit, my piston always ends up at the rest position instead of the usual 50% compressed position in my aeg (magnum motor, m170, 11.1v). Same result was achieved using different battery/motor/spring combinations. Below is my finalized design:


Protects against switch burnouts, overspin, reverse overspin, and increases ROF
Perhaps I should add an resettable fuse mod and the capacitor mod to make an "AEG enhancement chip".

The mosfet part numbers were the ones I used in the design. There are better mosfets out there, but these were cheap and thus suitable for the "test".

If you plan on following my schematic, make sure your mosfet and battery polarities are correct or else it you will have a short circuit upon pressing the trigger (mosfets cannot control reverse current).

Edit: updated the image