ShotShow impressions on SystemA's upcoming Revolution gearbox

[MadMax]

Wallace and Kumi were happy to show me their upcoming Revolution gearbox. I was asked not to spill the beans before they made proper press releases through various means. A lot has already been said and man their released studio shots are a lot nicer than what I could take at the show so I've not got much new to say. Still, I got to handle the parts directly and see what they were up to.

SchAWinggg!

Overall a very interesting gearbox. SystemA is using a similar piston to their PTW line of AEGs if not exactly the same one. I can't recall how many teeth the PTWs have, but if you've got one open take a look.

Anyhow, SystemA is using very similar avionics to their PTW. Looking through the sector to detect sector gear orientation to execute counted burst and semi. One thing I do like is that their FET board is very well heatsunk to a heavy strip of aluminum which is closely clamped to cast aluminum gearbox. Chuck in a bit of heat transfer goo and their Revolution gearboxes could have better heatsinking than their PTWs.

The gears look to be nicely milled. On closer inspection, the teeth look to have nice involute profiles, but without magnification and proper gauging saying much more would be uninformed hand waving. I forgot my hardness tester at home (dammit left it in the other jacket) so I couldn't measure hardness. Gear flanks were smooth though.

It looks like they've planned to accurately engage the motor bushing instead of letting it rattle around in loose fashion. The grooved motor hole makes me think that they've got a shoulder ring machined into the bearing motor bushing so when the gearbox is closed, it registers the bushing radially and axially. I suspect that the motor will directly mount to this gearbox instead of through a removable plastic grip a la TM.

Overall, the gearbox looks to be a very nice aluminum casting. No heavy flash and the sides close up nice and coplanar.

[MagnetoPenguino]

Seems interesting...

I need a new V2 box for my G3 rebuild... I was going to just get a King Arms or something cheap, but since I don't really need the gun anytime soon, I may wait for this instead.

Think it'll be any stronger than the normal V2 box? What kind of springs would you say it would reliably support?

Basically, this thing kind of has a SW-Computer built in, correct? Would be kind of nice to not have to worry about wire up a separate component elsewhere in the system.

[MadMax]

I have a strong feeling it uses the same springs available to PTW users (but not in a cylinder assembly). Smaller diameter than standard AEG diameters.

Gandalf's SW-Comp is probably more ruggedly built than PTW electronics. SystemA has taken a bit of a techno hammer to their electronics problems by directly observing sector gear orientation with optics. They require a massive retrofit (entire gearbox) where the SW-comp is drop in for almost any mechbox build. Judging from the regular failure of PTW electronics, I feel that Gandalf's lightning tolerant stuff will survive the clumsy tesla sweaters that most AEG mechanics would wear. While direct sector observation is a very definite way to control the motor, it's prone to blockage related malfunction.

That being said, I haven't done enough of my own assessment of the SW-Comp's open loop control scheme. I don't know if it will have half cock issues as it accumulates uncounted sector rotation operating in 3rnd burst. The open firmware allows some pretty interesting possibilities though. I suspect it might be a simple issue to run the motor at low voltage between shots to get it to wind to the same spot on the piston (say early partial cock where the piston just engages the piston). Gently rotating to piston engagement after shots or bursts could be used to set up the next shot when the motor turns on at full power to assure clean first tooth pickup. It would also work out accumulated sector rotation so you'd be starting with the same sector orientation every time.

The SW-Comp offers more flexibility for the kind of airsoft hackers here and it's a heck of a lot cheaper to boot!

[Gandolf]

Hi,

A very cool thing there!!  I was looking at the heat sink and thought they might be using thermally conductive plastic at first.  But I am glad to know that are not "that" far ahead of us 

I assume they are using dual main FETs without active breaking.  In their system, they can probably get away with that just fine.  I have found that about 250 peak amps is the limit for MOSFETs.  Beyond that the heating and all just kills them.  Only powerful LiPO systems can touch those currents, but LiPO is showing up a lot now!!

I think magnetic sensing will take over in a little while.  It is just too easy and too robust.  We just have to find time to work out the R&D there 

Gandolf

[MadMax]

An induction coil close to the sector teeth might be useful in detecting it's position. Riot suggested using optics to count the teeth, but they'd also be muck prone. I keep thinking that there might be a use for the now obsolete semi auto latch that disconnects the switch block (pre FET era). It still rides on a lobe on the sector gear. I think it should be a simple matter to detect it's position with optics. Coupled with timing tricks and an assumption on RPM it should be useful for inferring rotation angle on the sector gear. It might also put the sensor further away from the goop slinging off the gears.

I think the heatsink block looks black because the faces have some tape on them. I can't exactly recall, but I seem to remember that hunk actually being black. I wish I looked closer at the heatsink. It may be electrically insulated because it's screwed into the FET tabs. A lot of photos of Revo gearboxes seem to show the block covered in black film.

Isn't the drain continuous with the sink tab on many power FETs? I would think that it would be ok to ground this.

I think the current requirements of Revo gearboxes may be significantly lower than typical 3 gear AEGs. The extra gear probably could give them an additional 2:1 reduction (take some guesses on gear diameters and shoot a guess Riot?). Maybe one FET is for forward and the other for braking. Maybe this is why they need electrical isolation for the gates?

[RiotSC]

It does use PTW springs, so don't count on being able to just purchase a cheap AEG spring and throw it in.  Due to the bumpered cylinder head design, it should be able to take a bit more punishment than the standard V2 mechboxes.  A lot of stress would still be put on the mechbox spine though.  Let's hope SystemA's cast alloy is better than the run of the mill zinc/aluminum blend used in most aftermarket mechboxes.  Otherwise we'd still be seeing popped mechboxes as people find ways to up the spring rate.

There is definitely enough space to allow another 2:1 ratio reduction.  The question would be why bother?  Unless the motor that is paired with the Revolution mechbox is tuned purely for RPM, it does not make much sense to step up the ratio further since the ROF would just go down.  With the Magnum motor, SystemA already proved it could make motors that offer vast amounts of torque and RPM at the same time.  Considering that SystemA uses the same 3 gear layout as regular AEGs in its MP5 PTW line, it is quite odd for them to denounce the 3 gear design. 

As for electronics, SystemA's page does state that it does braking as well.  So I guess they noted our progress with H-bridge FET switches. 

[MadMax]

A ratio reduction would reduce current demand. 1/2 torque roughly equates to 1/2 current which puts loads into the realm of single FET switching. You also reduce heat that has to be dissipated in the motor and brush wear. All in all, I really do like the idea of running lower input torques.

The requirements for small diameter (space req'), high torque in a DC motor are conflicting. The easiest way to juice torque in most electrical motors is to up the diameter to increase the torque moment in the coil and magnet layout. Small motors like to run high RPM lower torque. Trying to run high torque requires high flux magnets, and high current carrying capacity. At some point you top out the available space for brush contact and you run heavy wire with low turn coils to crank current capacity and resulting torque. The motor designer rapidly runs into a high current corner where the brushes start to smoke and the armature can't dump heat fast enough (same size means same area for convective heat xfer).

The best value solution (save making a larger motor dia) is to run with more reduction and up the voltage. The funny thing about higher RPM operation is that you get more airflow through the motor.

If we're stuck with brushed DC, the most elegant torque improvements come with higher gear reduction. If we can make a design leap to brushless, we can move the current conductors to the can where we can make much faster conductive heat removal and remove the brush losses.

Back to the mechanical side, I'd like to see a quarter twist plug at the rear of the gearbox so you could pop out the spring guide and spring without having to open the gearbox. I can think of a few schemes to make a removable plug that threads into the spring guide that a M4/M16 stock could thread into. That would be a cute add. Pop off the stock to change the spring.

[RiotSC]

I highly doubt that SystemA would require the kit to run at a voltage that is other the standard voltage range (7.4V - 11.1V for LiPO and 8.4V - 12V for NiCD/MH) unless they wanted to increase the costs further by adding yet another proprietary accessory (the battery).  With our standard battery packs, a higher gear ratio would just mean a lower top speed.  So to counter that effect, I would just have to assume the bundled motor is RPM tuned with the torque matched to the resulting load for optimal performance.  Of course, without knowing the actual gear ratio used inside the Revolution mechbox, it is kind of difficult to draw a reasonable comparison.

For high torque motors, high current is not a requirement though if the torque/current efficiency ratio is high too by increasing the magnetic flux and number of windings.  That is why high speed motors tend to draw more current compared to high torque motors under the same load.  Brushless motor is indeed the way forward, but it does not look like any manufacturer is willing to take the first dive.

[MadMax]

Damn, I wish I had counted the teeth. Wallace probably wouldn't have been too happy with me doing that though.

IRC, torque in DC perm magnet motors is proportionately related to the number of turns per coil, current, and diameter. Turns * current is a simple number of wires carrying current in the same direction across the flux issue. Diameter gives the reaction force leverage to apply more torque. However, diameter is fixed so we should chuck that one out. AEG motors are small diameter so that doesn't help unfortunately.

You'd think you could pack in the turns and use hemoglobin ripping magnets, but I think there is a conflicting objective in high torque motors. Cranking the flux also increases the back EMF generated by the armature turning. You can develop high torque at lower RPM, but you rapidly top out and can't drive high rev's. Upping the coils requires finer wire and reduces current carrying capacity.

Still, all this discussion is hand waving without dyno testing. At the very least, we should measure a low RPM (say around 100rpm) current on our motors and a free running RPM and current so we can draw a torque-rpm behavior. From that we should be able to measure actual running current and see if our motors are running at max power (1/2 stall current).

Still, my gut feel says that there are compounded benefits for low torque higher rpm operation:

-more airflow over armature
-reduced tooth loads on the gear with the fewest teeth (pinion)
-lower I^2r losses

TM went with their 3 gear arrangement to provide reduction for low energy springs. I conjecture that their lower constant springs give rise to an optimal reduction with their standard ratio for stock AEGs, but non optimal reduction for high constant springs with the constraints imposed by limited motor diameter.

[RiotSC]

From the pictures I could find, I was able to figure out all but 3 gears' number of teeth.  (For those not in the know, each compound gear has a minimum of two gears per gear unit.  So there are actually 9 gears in SystemA's 5 piece gear set, including the motor pinion gear.  Well, 10 if you count the piston rack gear.)  Using the estimated gear center distances and OD information, I came up with the guesstimates for those 3 uncertain gears as well.  It would appear that the resultant gear ratio is 40 to 1...    If I were even anywhere close, that'd put the gear train in the realm of "Ultra Torque-up" gear set.  That'd need one heck of a high speed motor for sure... 

[JCW8]

With the removal of some components and the move to electronics, I'm wondering why they didn't go with a brushless motor...

[Whiskey11]

snip

Please STOP! Lol, I am just finishing up this chapter in Physics at the Uni and I must admit it hurts my head to think about it this often.  When we went over this I instantly realized why there is such a difference in the Systema Magnum and the Turbo3000.  IIRC they use the same type of Magnets (god knows they are hard to turn) but the Systema motors are all handwound as aposed to machine wound per the T3k.  It's quite amazing as it makes the T3K a little less torquey then the Systema Magnum.  It is strange how fast Airsoft and Calculus based Physics crossed!  Something as simple as a DC motor no longer holds any value in my mind!

To JCW8.  Probably cost?  I dont know, it makes sense not to have them but think of the infastructure change needed to produce those motors.  Besides they may buy their motor parts elsewhere and just assemble them.  Again cost of infastructure change. 

I've not really been a fan of this gearbox, I see too many area's that need some changes for me to accept it.  The idea while cool does not justify the probable cause.  Really I can only see them going two ways with cost.  Really high (400-500$) because it's "top of the line" and intermediate pricing because of the signifigant reduction in pricing.  Knowing Systema's past marketing habbits it will probably hit the 350-400 range to be priced above their Magnum/Turbo GB's as this will be their "Flag ship" gearbox if you will.  I guess it will come down to time.  I'd rather just buy another SW - COMP from Gandalf

[JCW8]

...or, maybe they're going to release one later with a Brushless option, but want to sell some of these first so they can get repeat business from the guys who always have to get the latest upgrades...

I think the best thing about this gearbox is that it simplifies things, and having a computer that supposedly will park the piston in the same spot every time is nice.

[Gandolf]

I think they are simply going slow and not bitting off more than they can chew.  The new electronics seems to be keeping them very busy as is.  Trying the brushless motors is almost harder than what they are doing now.  They just don't have the bandwidth to do both.  Maybe next year.

Gandolf

[RiotSC]

Considering that SystemA does not use brushless motors even on its PTW line, I find it unlikely that we'd be seeing brushless motors for regular AEGs from SystemA any time soon.  At least not until a competition releases one first and proves there is a profitable market.