Going Brushless?

Having got r/c working I am on to the next step of motor control. I'm heading for a module about 1m LOA, so I opted for the G2 hydra15 combo, which seems to integrate with my Rx/Slave, and currently using a 12V battery with the thought that the current would be less for the same power than with a 6V battery.
Future problem is that the motor is a 1100kv unit so I'm intending a 2-stage belt reduction ( 2 x 3:1) to get down to the target of 1000 rpm, since that seems to be compatible with the prposed shaft seals.

I've noticed that there seems to be no definate 'neutral' and the motor tends to twitch both ways, though at the moment I'm not sure if that is noise pickup at the analogue X-Y joystick inputs generating variable servo codes or just dither in the PWM module and/or controller. Anybody else had this?

Also, the instructions on the controller talk about setting the min & max positions based on red & green LEDS that I don't seem to have, though I do have a 'beep' on switch on. Anybody managed to configure these?

As a final question, if I select max fwd or max rev then the motor spins up OK, but then stops, is it some kind of cut-out as the motor has no load at the moment?

Simon,

i presume you have an oscilloscope to view the rx pwm outputs with?

i would start by looking at what the pwm output of your rx actually is. some ESCs don't like values higher than 2ms or lower than 1ms, with neutral being 1.5ms. a value higher or lower than those extremes may be considered invalid, shutting down the system as a safety precaution. most ESC will have a pretty wide "deadband" around the 1.5ms setting to give a stable neutral, so for that, i would look at how stable or "twitchy" your 1.5ms value is.

lastly, since most scale subs are slow lumbering beasts, i doubt if a 6v system will draw too much amperage for the 15 amp rating of the ESC... it might make things easier on your reduction system design.

Thanks for that, a great help.
I haven't got a 'scope on it yet, but I did go back and check the nominal pulse widths and realised that I had characterised the SG90 servos as providing 180deg of movement with a range 0.58msec to 2.5msec which I had left as the default range in my rc system.
Changing to 1msec to 2msec reduced the mechanical range to 90deg but also made the motor and servos stop twitching, presumably they are now hitting the extended neutral position.

I had to add a small offset of about 0.05msec to get a true netural from the motor but thats all added digitally. As its a digital Tx I create the servo positions as 0 to 255 with 128 being the neutral, and convert from the digital value in the Rx where I can accomodate specific individual hardware.

I shall another think about 6V vs 12V, it would certainly help mechanically.

Simon,

calibrating the ESC to your system's neutral, max, and min settings will eliminate the need for the neutral offset, and give you "true" min and max for the ESC/motor combo.

I tried contacting Mtec about how to configure the ESC (G2 Hyra15) to my Tx/Rx and to get mechanical details of tha max penetration depth for the mounting bolts, but I'm still waiting. Seems a good product byut not very good documentation or customer services.
However, with the zero offset set up in the Rx software and a guess at the max bolt lengths I now have a prototype gearbox which seems to work.
Since the Hyrda15 is suppled with a 1100kv motor it should give about 12000rpm @ 12V, far more than the 1000rpm which is my aim, as that seems to be the limit of the seals I am thinking of using.
I used MXL drive belts & pulleys as they seem to be the smallest pitch (0.8"), smaller than the T2.5 (2.5mm) so smaller diameter pulleys.
I opted for a two stage design each matchng a 12-tooth with a 36-tooth to give a total of 9:1, a bit smaller than I wanted but it fits within a 68mm diameter envelope (I intend to use 75mm OD tube for the WTB). The 12-tooth was the smallest I could get bored for 4mm shaft.
The drive belts are 54-tooth (109mm) which gives a centr-centre distance of 29.5mm, though my use a plastic rule and making each frame individually means that the actual distance is more like 29mm or less; lesson - do it properly not quickly!
There are 10-tooth available but they would have required me to bore and add the grub screw.
I have pinched the idea of the bored brass bolt from this group, though at he moment they are simply plain bearings, I have yet to counter-bore to fit 8mm OD ball race bearings.
The use of tw equal stages means the motor and output shaft are concident so a bit easier to make the frames, the middle frame of course has a clearance hole for the motor as it as outrunner.


My initial thoughts are:
Use 1.5mm aluminium not 1mm, less flexing, especially the middel frame that has a large clearance hole for the motor.
Using M4 tie-bolts allows adjustment but is so fiddly to get the motor aligned, but the use of a guide block during assembly would help, or use spacer tubes. However the use of tie-bolts does allow 2 frames to be assembled and rigid then add the last.
Having 3 frames was an idea to shorten the layshaft and possibly use the space 'above' the motor for something, say a servo, but in fact I think it is too small. I think I will keep 3 frames as it provides a nice space for the motor wires to exit.
Whilst the vibration & noise is low the nuts still keep vibrating loose so I need loctite or nyloc, as I don't fancy star washers as they might upset the alignment.
Use half-nuts to retain the bearings and reduce the bolt length as they take up less space.
The max current is about 1.7A @12V max speed, though no load of course, but OK.

For the real thing the output bearing would be a longer bolt and have a seal on the outer end, but I'm not quite fixed about that yet, and of course I need the 2 control rods, again probably a drilled bolt, but not sure about bellows (simpler but longer) or O-rings (more exact maching but shorter).
The whole gearbox I intend to fix to the WTB bulkhead via the prop-shaft & control rod bolts, with the others being in recesses machined in the bulkhead. I was thinking of holding the WTB together with tie-rods outside, so lougs cast into bulkhead as there would be less penetrations to leak.

Next step is to fit ball races to the layshaft bearings and see if I can get in 2mm control rods.

simon,

apologies, as i appear to be a little late to the party regards to the gearbox, but the brushless outrunner motors can be disassembled, and the motor shaft turned round so that the extended portion of the shaft comes out what i will call the base, instead of the bell.  that way you could have a gearbox with two bulkheads instead of three.  The motor would be switched around so that the motor would get bolted to the middle bulkhead of your gearbox, and the front gearbox bulkhead, where the motor is now mounted, would no longer be needed.  in actuality, because of what you have already built, the middle bulkhead of the gearbox would get replaced by the forward bulkhead of the gearbox, with the motor bolted to the outside of the bulkhead, and the shaft going through a small hole in the bulkhead.

Looks like very nice work from Simon. And a very interesting comment from Tim. I didn't realize that you could (easily??) dismantle a brushless motor and alter the shaft configuration. That makes what is already a very compact system potentially even better.

And while I am on, I have not found brushless speed controllers as easy to use as brushed ones.

Just at the moment I have a Mtroniks G2 Hydra 15 Brushless Speed Controller which I am having trouble getting to go into reverse. You know it is not really "working straight out of the box" if you have to get the Scope out!

LATER EDIT: I have quite a few Hydras now and they are basically fine. I did try a Hobbyking progammable one which I found disappointing. (I was trying to eliminate the reduction gear by programming but it was too restricted in setup.)

I hadn't thought of dismantling the motor either, but as it happens the motor plate has a hole there as there is a slight raised boss on the motor back plate so clearance is needed anway, so I might have a look when I start on the balll-race bearings and have to dismantle it all.
Note that the pulleys I used have miniature hex key grubs screws, not sure what size but < 1.5mm, but RS to the rescue.

Whilst I had problems with the ESC initially, once I reduced the actual pw to the 'normal' range (the servos seem to be able to accomodate a bigger range) it seemed to work in both directions. The neutral pw may be a bit off, but that may be due to my use of miniature thumb type joysticks not returning to exactly the same point.
I did have a software bug in the Rx mind you that resulted in a software underflow, so at almost max reverse (due to the neutral offset) it underflowed to max forward; not a good move, I also got smoke due to the mounting screws being too long, along with some odd chattering, but it survived!
I am still waiting for a reply from Mtronics about if the Hydra15 can in fact be tuned to the Tx/Rx as the instructions supplied are no help as they talk about red and green LEDs that I don't seem to have, I assume absent as it is already tuned to the motor.

Simon, David,

i cannot find anything anywhere that shows where the LEDs are on the Hydra G2 escs.  i guess they must be in an obvious place if simon says he has never seen them (sarcasm).

reading the manual, calibration requires that you be able to see the LEDs.

I have just had a look at my Hydra and the LEDs ARE there - next to the red cable.

They are not very visible partly because they have to shine through some "potting".

David

PS. This speed controller now seems to be working fine in forward and reverse without me doing anything. I think that the initial setup is always a bit haphazard.

Thanks for that, so they are, if you turn the lights out you can just about see them! though mine seem to have a label over them as well.
Mind you the 'as supplied' settings have seemed OK so far so I shall probably leave them alone untill I need to change them.