RC Valve Controller Development

Following a conversation with Mario at the Champs today, and previous conversations with Will, it seems that there is a lack of cheap and simple RC switches suitable for controlling pneumatic solenoids / relays etc. I developed one as a one off for Manta a while back, so I have all the code and basic designs for one. What I'm wondering is whether it would be worth developing a proper PCB and getting them printed professionally to sell. (The reason why I'm willing to sell these rather than my FW and HW ESCs is that they are much much simpler and easier to put together in large quantities)

The kind of specs they're likely to have is as follows, based of the one I built for Will:

12A continuous max
Up to 80v input
3 channels (eg, throttle stick up for flip, center for dump, down for self right)
Gear channel to arm/ disarm
Small (likely to be quite a bit less than 5cm square, though until I've developed the layout I can't give an exact figure)
~£20 - £30 (again, not sure on exact components so price is subject to change)

What I need to know is how many channels do most people usually use (the less the better really)? Also, would people actually want the arm and disarm feature? It is mainly in there as the throttle channel does not failsafe correctly on some receivers, so the Gear channel is checked for the fail safing. However it is also used to arm and disarm the RC switch so that there's no chance of accidentally firing it by knocking the throttle channel whilst arming (I have it on Richie's lifter and I find it really quite a useful feature).

I've got exams coming up so am unlikely to get anything for sale until the end of June at the earliest, but I just want to do some market research so to speak to see how many people would be interested, and what features people would want..

Think it would work as a simple on/off relay for weapons such as lifters too? That being if you could up the A rating a little?

When i designed a low pressure flipper based around a high pressure architecture, the issue I would have wanted to overcome using some kind of control system is the amount of time the valve is open as I would never be able to control it effectively by sight which would mean a loss of co2. Therefore, would it be difficult to add in an adjustable "open valve" settings which would reset when the control returns to its normal position?

Just thinking out loud, the problem with having the dump in the centre for the throttle stick is that on most tx's the throttle channel doesn't spring back to centre, and if down is self right you need to be carefully not to manually pull the stick down too far as you will then just fire the self right valve. Ideally I guess you want it on a stick that does spring to centre so you can just quickly flick it up to fire and release and it will dump.
Not sure how the heavy guys get round this, I've always used the rudder channel to control flippers but it is a pain in the arse as I always forget which direction is fire.
In conclusion, I haven't really helped at all but hopefully food for thought!

Apparently there are equivalent products available for a similar sort of price range such as these RC Relays.

All mine would be is a smaller solid state version for a similar price with maybe a few extra features, so the market for custom controllers seems to be substantially smaller than I originally thought.. They'd be useful in Beetleweighs and Featherweights which are pushed for space, but unfortunately I can't see enough need for it to be worth getting a professional print run done, unless anyone can think of a suitable application?

I looked earlier at the technobots stuff.

The Battleswitch, even if I had good experiences with 1 in the past, nothing else indicates that those are useable in our machines, I didn't meet a roboteer yet that can say it's a good switch.

The Technobots switch can't take more than 30v, and that seems the cut off for most available switches.

It is perfectly possible to use a lower voltage RCswitch to switch another relay that switches the main relay. But that introduces more points of failure.

In short, every machine that uses a solenoïd to control something needs an RC switch, and as long the voltage isn't higher than 6S, commercialy available stuff will do the job. It's when you go over that 6S it gets tricky.

What Team RCC is searching for is a dependable source that delivers the goods.
For Hannibalito 3,2 it's simple. Just an on/off switch. Throttle up, 36V goes trough, valve activated. Throttle centre or down, valve closed.

Bullfrog. Stick center or down, valve 1 activated. Stick up, valve 2 activated.

Another machine needs a switch that is a bit more tricky. But the one as used in Manta can do the task. Stick neutral, nothing happening. Stick up, solenoid 1 activated, stick down, solenoids 1 and 2 activated.

As long as it doesn't need more than 2 separate output channels, it should be fairly simple to make one that is less than 20x20x10mm. It'd likely have an output of 5 to 8A at well over 36v depending on what suitable components are available, so that should be ok to drive 2 Burkets in parallel.

Next time I get a load of stuff professionally printed I'll see if I can tack a few dual channel switch boards onto the order (they'd also be used for single channel switches; just leave some of the components off)

I've had noting but trouble with the Battleswitches, they constantly misfire. I've used alot of the Technobots RC Relays and they've been fine.

What might be useful is a switch with a timed output, so it just pulses instead of staying on, would be good for electric axes and the FP pneumatic flippers.

Beast2 has a techno switch relay and is no problems firing 2 burkets

Well, I had to make a Farnell order up to £20 for the free shipping, so I ordered some parts to try out these valve controllers. Put together a little dual channel one on prototyping board, and this is what I've got:

Attachment 5306Attachment 5307

(Just shows top and bottom sides of the board with an antweight motor for scale; currently ~22mm square)

All seems to work fine during the limited testing I've done so far; It'll run a 550 motor at 12v under load without getting too warm, and seeing as it should be good for 8A at 100v I'm guessing it'd fire a Burket valve just fine.

If I did get enough interest to do a production version, it'd likely be ~50% smaller and wrapped in heat shrink. The firmware is also fairly crude (thrown together for testing), so that'll need various improvements before it's suitable to use.

Upgraded the firmware to make it useable and added some heatshrink. It's currently set up so one channel is on when the stick is up, the other is on when the stick is down, and both are off in the middle. I'll be bringing it a long to Portsmouth on Sunday, so if anyone there has a Burket they can let me test it with (or better yet, a robot to stick it in!), then that'd be awesome :)

Ed Wallace would have tested that at burgess hill as he needed one ! By farnell do you mean cpc farnell ? As if you do a web order it's free delivery after £10.

Would you be able to make this controller with centre off, up will be on and the down could be a timed on switch adjustable between 0.1 seconds and 1 second ?

... Could've sworn it was £20... Maybe I'm getting RS and Farnell mixed up. Either way, I now have a prototype valve controller and parts for 6 more :L
(EDIT: Farnell is £20 for free shipping, CPC is only £10... might have to start looking at CPC again..)

And yeah, would be a fairly small firmware change to make a timed switch, though to make it adjustable could be more of a pain :L

It would have to be adjustable as it would open my valve just enough for a self right flip.... And I think that's around 0.25 second press of my tx button !

I mean, it's perfectly doable, but would require a different microcontroller that can take an analog input from a potentiometer. The prototype will just be running the basic code for the time being, but I'll look into adding adjustments when I get hold of some new chips.

On a similar note, I am also looking into developing a board that converts a lifter into a giant servo. It would just involve fitting the board between the Rx and the lifter ESC and then attaching a potentiometer to the lifting arm or whatever it is you're controlling to provide servo like control. It would also have another potentiometer on the board to allow it to be tuned to the setup being used. Is there anything like this already available? I had a quick look and couldn't see anything..

I'm thinking of designing a few different types of these little boards, which I can then send off on one production print run. Currently got the valve controllers, and the servo adapter... any other suggestions? I've been asked to do a mixer that failsafes correctly before, but I think those are already available...

Got asked to make some single channel ones, and so was going to build them on veroboard like the first one, however I realised that I could re-purpose my N2HV ESC boards to build the valve controllers on:

Attachment 5312
Attachment 5313

Sadly I forgot to take the prototype board at the weekend so couldn't test it (Maelstrom needed one as well so could have got some proper battle testing :roll: ), So I still don't know if these will actually fire a Burket, but theory says they should.

In other news, I've got a rough draft of the servo control code working; just waiting on some better chips. If these work well it'll hopefully make the flappy lifters like on Tormenter 3 properly controllable :L

I would be very interested in servo-style converter boards. Am I right in thinking that they would have a similar effect on claw mechanisms, so long as the potentiometer is mounted to a point of rotation?

Yeah, they'd work fine for that assuming you fitted the potentiometer correctly.

Been working on a load of designs to be sent off for professional printing. Currently got:

Feather ESCs
2 channel valve switches
Servo controller boards
Nanoweight ESCs
Signal mixer with built in Gyro and auto signal inversion (for when invertible robot's are flipped)
Antweight ESC with built in Gyro and signal inversion.

Any suggestions for bits of electronics hardware welcome; I still have a bit more space on the board. I will only be printing 10 of each, but can print more if people are interested in one type, though I will likely only sell limited numbers of the FW ESCs and Gyro boards as they'll probably be a right pain to build :L Also, most of these are currently just untested layouts, so don't be too surprised if I pull items off the list after the boards arrive..

I had been considering using an Arduino to make an electric Axe controller with a couple of modes, switched between using the AUX 1 on the transmitter. Mode 1, you use the rudder stick to manually fire back and forth. Mode 2, you use the throttle stick to control the speed the axe fires automatically back and forth. I was going to use some small switches near shaft to act as limit switches so the axe knows where it is. However I am not sure its actually that simple, but I will give it a try regardless, I am sure you can come up with something better.

Axe controller's not a bad idea actually; I need one anyway for my next feather.. I'll probably just change the firmware on the Servo controller boards for that though as they're fairly similar in hardware requirements. It would likely be a time based system where the axe fires at full power for an adjustable amount of time based on the potentiometer on the board. I'll see about having it switchable to manual control on the rudder channel too as that's something I'd quite want for self fighting... May also be able to do limit switches, but I'm not sure I'll have enough IO pins for that!

Just a thought on what Terry said about adjustable timing, you said you couldn't have a potentiometer as you lack an analog input pin, but I guess you could have an ordinary push switch as the input, then you could for example have it so you press and hold the switch for 2.5 seconds or 25s and that would correspond to the valve being open for 250ms, or have it so you press it multiple times with each click being an additional 10ms of open time. Obviously this is harder to make precision adjustments than with a pot but perhaps could be an option?

Not a bad plan. However I just bought a PIC with analog inputs instead :P

The controller boards arrived back a while ago but I'm yet to properly do anything with them; I've made up a few of the servo controller boards to do several things other than control servos (the one I used to do the endstops for Richie's lifter is shown below), one of the 3 channel valve switchers was sent for testing in a heavy without much success yet (though I haven't had it back for diagnostics), and the other boards I haven't had the time nor bought the parts to populate them.

http://www.fightingrobots.co.uk/atta...tid=5538&stc=1

What I have been doing is working on the new heavyweight controller prototype, which is just a scaled up version of the featherweight one with 2 Fets per channel rather than one, and much better and safer driver circuitry. I say prototype as just after sending them for printing I realised a load of flaws in the design that need improving, but it should still work fine albeit not at such a high current rating as I'd hoped. This version also includes the facilities for testing out current limiting, which should help keep everything from catching fire. Having looked at the ragebridge controllers I've been impressed at what a powerful tool current limiting can be (60A controller running a heavyweight fine).

As such, I realised that it could be possible to create a plug in board for other controllers to add in adjustable current limiting. This would plug in between the power supply for the ESC, and also the signal line. It would be mainly aimed at the TZ85a, and would have an upper current limit setting of around 80A. It could be used to reduce the risk of flaming your motors on stall or a large overvolt, as well as possibly safely driving bigger motors than you otherwise could (see 60A ragebridge running 1000A startup current motors). Just a possibility; curious if there'd be much interest in me making such a thing..

A small/lightweight in-line current limiting board would certainly be interesting. I'd like to current limit the Dewalts in T3 so I can go to 6s and keep the smoke in, but don't have anything like enough space for an RB2.

What method are you thinking of using to limit the current?

It would be a closed loop system, so would take current readings with a low resistance shunt, and then use the readings from that to adjust the maximum RC signal to the ESC accordingly. Unfortunately it won't be possible to have a very high refresh rate seeing as the RC signal updates at a maximum of about 100Hz, but it should still be plenty enough to stop things frying so easily.

Ok, so back on the original topic of this thread, the valve controller has now been successfully tested for a few events in Dystopia (the valve controller worked fine, even if the rest of the machine occasionally didn't :L ). This was running 2 valves on separate channels at 36v, so fairly typical usage in a heavyweight, and there have been no issues with it. I've just bought in a stock of components, so these things should now be ready to sell.

Prices are looking to be:
£15 for a single channel switch
£20 for a 2 or 3 channel switch (uses different board)
£2 UK Postage (will have to check price for further a field)

Single channel default operation will be on for up, and off for centre stick and down.
Two channel default will be channel A will be on the up stick, channel B will be on the down stick, and both will be off on centre stick.
Three channel default will be channel A will be on the up stick, channel B turns will be on centre stick, and channel C turns will be on the down stick.

I can set up the channels for any combination of up, centre, and down that you like, so just tell me what setup you want. Do note though, I haven't tried firing multiple burkets on one channel with this at once, and although it would probably be fine, if possible it would probably be safer just to use two channels set to fire at the same time, so bear this in mind when ordering.

I don't have a new picture, but I'll just repost the one of the single channel switch from earlier in the thread as they haven't changed at all:

http://www.fightingrobots.co.uk/atta...tid=5573&stc=1
If you're interested, either PM me or email me at rapidrory@gmail.com. Payment is via PayPal to that same address.