Tuesday 30 April 2013

BHD Lights 2.0 Teeces V3.2 lights demos

Well, I ran it for a few more hours, tweaked a few weightings to make it more likely to do what I wanted and I'm really pleased with the results. :)  The random nature means that it does what it wants, but the probability of stay or change is weighted. :)  Here is a demo of the PSIs in action:



And the triggering is working fine, no de-bouncing so a slow setup of the combo lines may cause issues, but single lines are fine, and multiple lines are best done using a microprocessor.  Here is a vid of the total kit.



The sketch is here:  Google Drive  Look for BHDLights2_0.zip  in there are 2 versions, 1 for V3 and one for V3.2 PSIs.

I still need to test the HPs, but don't have any spare 2 pin female leads, so may need to tweak the code if i am triggering the wrong LEDs, but i guessed at the HPs based on the last boards, so might be OK.  Easy fix, as soon as i have leads i will test those too.

Monday 29 April 2013

Teeces Rebuild for pin triggering and 3.2 PSIs

A quiet week for My R2, sort of.  :)  I have been working on Teeces kits.  I am doing an all blue one for my R3 dome and a couple of std coloured ones for spares. :)  Should help a few UK/EU builders and save them paying import duty. :)  But typically everything is cheaper in the US, even LEDs!  So the price to me for all the bits will be quite a bit more than the JoyMonkey run, so it's not worth selling any as kits, but i can get a couple finished and tested for sale to make it worth doing.  And with a few nice additions of my own.  They should be pretty sweet and will just need power. :)

I have just added the code for the 3.2 PSIs.  The sweep is in there and can go from either colour in either direction, and I also added the half and half option in both directions too (Stuck slider).  Unfortunately to achieve this I had to remove the independent control of the PSIs, so front and rear do the same thing now, but as they cant both be seen at the same time it's not too bad.  But the new effects are worth the effort.  It looks great behind a diffuser, cant wait to get mine installed.   The standard random mode can swap colour (red / blue / purple), do the sweep or the half and half, and its all done with weighted randoms to make the frequencies more suitable, so no patterns. :)

Then I added a 3 state line triggering system to my code.  A bit like Rotopod's Periscope code.  This should allow for easy triggering, it will allow 7 triggered states and the default random.  The single Gnd and 2 Open modes could easily be triggered by servos or similar, simply pulling that pin to GND.  But it should also be simple to control with another arduino and combine with servos etc then too.

I have applied these 7 modes:
Open, Open, Open  -  Random mode, so that with no triggers it just runs randomly.
Gnd, Open, Open   -  Short Circuit Script- Timed to match MP3
Open, Gnd, Open   -  Cantina script, like a graphic eq on all screens for MP3 length
Open, Open, Gnd   -  Leia Message script - Text and then sine wave till for length of MP3
Gnd, Gnd, Open     -  Wowie script - Hearts and then heartbeat, intended for use with Wowie.mp3
Open, Gnd, Gnd     -  Random flash on all HPs On/Off
Gnd, Open, Gnd     -  Scroll R2 icon across all logics
Gnd, Gnd, Gnd       -  Text of choice scrolled (Han Shot First!) or packman, or X wing, or Tie Fighter. :)

These are the ones I use the most.  But I can also use them to turn individual HPs on/off etc, so there are other options.  If you have any ideas of the best 7 let me know. :)  I'm happy to tweak it.

If you need an assembled Teeces kit, with triggering built in.  Please let me know, i may have one spare. :)

I would prefer not to ship these to the US, they will end up expensive with shipping and import costs and there are others over there who do the same thing and id rather not tread on any toes, so UK/EU only.

For my setup, I intend to run my PSIs from my magic panel.  To allow space for a voltage divider to be used on my Teeces logics.  So mine will be setup differently.  But I think this approach will be a flexible method for the rest of you. :D

I would like one more night running the code to make sure it is as i want it, but im almost ready to upload it.
I will finalize it and upload it tonight if anyone fancies upgrading their kits. :) There will be a V3 and a V3.2 version for the PSIs, but both will have the triggering.  You will just need long pins on 4, 5 and 6 of the pro micro to connect the lead to trigger.  My kits will have long pins fitted on 2-6 for the input lines and the option of using I2C, so you can use it with R-Series or Ard2-P2 (with the right code installed :) ).

I am really happy with the changes, they look really good.  I will try and post a video of the new PSI actions  ASAP. :)

Thursday 11 April 2013

2-3-2 Takes another step forward

Finally some positive 2-3-2 results.  :)

The shortened leg rods have made a big difference.  One is still too long, but he is already much more stable and balanced in 2 leg mode.  Aiming to redo that this weekend.  The actuator mounts, the redone shoulders and the tightened setup have all helped, and i am almost there now.  Starting to feel that with a couple more refinements i can trust it a lot more.  I haven't even adjusted the positions in the code yet. :)

There is still some lumps in the action, this is caused by the caster wobble, you can actually see when they turn, so adding the feet drives will cure that, i will have a go at that soon.  

But the wobble and walk are fine and he doesn't want to fall over and the lack of play in the rods now means he doesn't build up momentum in the same way when wobbling. :)

Id still like to redo the ankle pivots, there is some play there that can be removed.

But here is a quick demo of how its looking. :)



So last few bits on 2-3-2 ASAP, then the door servos (putting this off, my hands are too big to get in there easily), a few refinements to the control code and a few servo spot adjustments and i will be ready for CE2. 

Sunday 7 April 2013

LFS, Drive Wheels and Leg Rods

Been ill the last 2 weeks now, really bored of it.  But got a bit done the last few evenings and this weekend.

The LFS was next one the list, so i had a look at it and decided not to use the designed rotation of the top in relation to the base.  My thinking was that i could then do full rotation if needed and the wires wouldn't get pulled about so would always look better.  So i glued the screen on, and glued the fiddly bit to the top of the base and used a small piece of alu tube to connect them.  This leaves a good size space in the centre for connecting the solid core to the flex.

The flex i have isn't right, it's a 4 core telephone handset lead, but it doesn't quite work, and i need to replace the solid core, may go with grey, black or blue instead too.  So i ordered a 3.5mm headphone extension that is coiled, i hope it's about the right size. :)  Waiting for that before soldering up the board.  I have the option of passing the main wires down the central tube and using the flex just for the look, i don't think I'm going the aerial route, not sure yet.



Then the leg rods.  I stripped off the battery boxes and removed the ankle locks.  Will leave these off till it all works, they were a bugger. :)  While I had the drives off I had a look at the wheels, they were rubbing slightly, so maybe that was the cause of the static?  Even if they weren't they would be costing me power.  So I got some new blades for the Stanley knife and started cutting.  I cut about 3-4mm off each side.  Same foot print, just the edges trimmed.  Seems much better, if the hubs slip i will add a few screws to hold them in position, but wont do that till it needs it.



I shortened the rods by a cut width, about 1mm, the aim was to align all the elements at the upright point (legs, feet and body).  And so allow more back movement to help with balance.  I stripped them out, cut them with a dremel and braced them with a steel plate.



Then tightened all the rod fittings and reassembled.  The position is better and it all seems better, but there is some play in the JAG ankle axis, perfect for 3 leg, but for the 2-3-2 and 2 leg driving i think less play there will make a big difference. So that is James' next job.  Cheers dude, couldn't have got this far without you.



The actuator positions now need a fine tune, think i will pop it all together and test it with the current ankles, It will be another step closer to a reliable mech. :)

Once the ankles are sorted i will be able to strip the legs and rod setup, threadlock and final fit it. Hope that will stop it coming lose and causing issues. :)

Last night I weakened and ordered the Magic panel for the dome (the alu panel that goes red.  Thinking of adding the new 3.2 PSI boards to run from that, and not the logics, that may free up enough space to do the voltage divider on the Teeces. :)  Should be a nice addition. :)  Will rig up the PSI boards and run them on a Nano till the Magic Panel arrives, be good to do some testing and triggering with those and maybe the HPs on their own pins. The Magic panel has 6 outs, so 3 for the HPs, and 3 for the Chain to the PSIs. :)

Started playing with the periscope control tonight too.  Just a 5 button pad to trigger the 8 modes via the 3 input lines.  I have the old code installed, so no off mode, but works fine.   I may just edit the modes and use it this way from the same controller as the lifter.  But need a breakout board to change the code on it, so that is now ordered.  Will need one for the magic panel too, so will come in handy. :)