K8055 - Dual Pneumatic Cylinder Control

Hi to all!

I have been playing with the K8055 and Pneumatic options, and thought it would be useful to share setups, control circuits and air loop diagrams, profiles and of course, successes and failures!

It is my intention in this thread to investigate all previously tried and some untried setups created from standard pneumatic components, and demonstrate where possible the differences in control that can be achieved; I will be comparing different setups where possible within the same frame - however, to avoid having to buy large numbers of (sometimes quite expensive) components, it may be that only one cylinder of the two in the frame is active.

I will be looking at all setups, including more expensive components such as proportional valves, but as with my relay solution, I am particularly interested in achieving the best possible setup with cheap components, and will dedicate significant time to improving existing setups and profiles (if improvement is possible!)

I will not (for reasons of cost) be prototyping with miniature components, so all builds and examples will be directly representative of actual motion platform performance, and will be under load.

I appreciate that much of this investigation has already been carried out, but not in recent years...
I am keen to investigate recent advances - particularly in PWM control for which there are many papers online, but also advances in technology with new components (where cost permits!)

For each setup, I will post setup diagrams (electrical and air), pictures, videos and example profiles for discussion.
Where possible, simplified overviews of available component types will be also be included.

As with the Relay solution thread, it is likely that I will edit this first post later to include an index.

Tim

As I will be working with many different pneumatic valve configurations, I have designed this transistor array as a convenience - the terminal blocks are much easier to manipulate than soldered wires!

For those that followed the K8055 - Dual wiper motors... thread, the way that this board works will hold no suprises!
For everyone else, here is a video explaining it at a very high level

http://www.youtube.com/watch?v=ogUamuON2FY
oops - I said relays at 1:28... meant transistors! :

I will draw up a circuit diagram for this later on in the week
(I will be back in Amsterdam with work until the end of September, so should have some time to kill in the evenings).

Unlike the relay circuit which already existed, I am proud to be able to claim 100% originality on this configuration and circuit design... yes, it is a simple enough circuit, but it is super cheap to build, very effectively controls all 8 outputs, and is very convenient when changing setup configurations.

I have a design also for a circuit to incorporate PWM / Dynamic Voltage, which will be necessary for proportional valves; I will build and test that later, and share the design when I am satisfied with its performance... for now, this board is specifically designed to support experimentation with valve configurations and improved pulse routines (if improvement is possible).

Here is another quick video demonstrating it working with two 5/3 valves
(also giving an overview of the way in which the 5/3 normally closed valve works):

http://www.youtube.com/watch?v=6ieyQ9R3YVo

Tim

As promised, here is the circuit layout for my K8055/Pneumatic Valve test board
Apologies that this is hand drawn - will replace as soon as I can with a digital version...
this was drawn on the plane on my way home!

The component view is from the top, with important tracks highlighted in colour.
Component numbers and values in the top and bottom margins.



Regards

Tim

PS. I have finished building my Pneumatic test platform today - will post some pictures and design notes (possibly in a video) later.

Test rig

Hi to all!

Here are some pictures of my test rig. It is simple enough, but if anyone has any questions regarding the design, I would be happy to explain

The platform itself is sitting angled down in the pictures - this is because it is designed to be level with the pneumatic rams in the centre position.
The back leg will eventually be replaced by another cylinder, hence the spare space behind it. (I have a cunning plan on how to control that )

The frame articulation is made possible by the use of castor housings (just with the wheel taken out). 4 small ones for the pneumatic cylinders, and 2 large ones to control the central axis around which the platform rotates.

The platform is able to move up and down, and rotate around the central axis (imagine this axis being front to back between the centres of the 2 silver castor housings) but cannot move from side to side; I must admit to being a little suprised at how incredibly stable the platform is - before I added the hinged section in the middle, the top was flopping around all over the place... with the hinges screwed down, I genuinely cannot move it from side to side!







Making a frame that could be easily packed away was very important, as I dont have too much space, so here is a pic of the frame in its normal storage position... not too offensive - even for my wife!!



For testing under load, I will simply strap the kids static frame onto it for now - complete with the spare car seat from my VW Sharan:



well... OK - I made it for me, and it's a bit big for the kids, but they insisted and now won't leave it alone!! - heres a pic of my youngest son clearing the nurburg ring in an M3 in under 8mins... at 4 years old!... with a box under the pedals as he couldnt reach them!! - go-karting lessons all round I think...



The final build? who knows... it is likely to be similar to the fourstar f1 simulator (http://www.youtube.com/watch?v=qIe9O529MR0) - I figure why not? everyone has seen it and drooled... we may as well try to get as close as possible!!

It's a shame I didnt get more time this weekend to actually connect the frame up to the air, but at least now that I have the control board and frame prepared, it's something I can get straight on with when I finally get home!

Tim

Congrats tim,

Looks like you finally got there. I followed all your topics peeking in.
For me theres nothing worse than being stuck in the brain work category of the build altho it can be fun and rewarding sometimes too. But all I really want is to just get playing on the dam thing. I am a perfectionist so I wont cut any corners and it took me a long time to come up with a configuration with electronics and hardware. Will you still complete your wiper motor setup or is that abandoned now?

Bart

Hi bart

I am still working on both - I have built a number of wiper motor sims, but until I cracked the gearing/chains setup they were too weak to work for me, and I didnt bother to post them... I very definitely havent given up on that front, and the linear actuator is still under test... keep an eye on the other thread for that in coming weeks

This is just the start of a different line of investigation to see how far we can improve K8055 pneumatic setups

Tim

Hi Tim, well done with your endless testing of the K8055. I am still keen to see a result, especially with wiper motors. I may have missed it in the thread though. Look forward to seeing the cylinders in action.
Cheers, David.

Hi Bsft

Check the other thread for the full answer, but you didnt miss it - I just didnt post it yet! Have been busy

This thread is an extension of the K8055 concept - very similar, but I am attracted by the high power and speed available to pneumatics... just have to nail the control!

Tim

Hey there, some new infos? :happy: :happy:
Thanks Charles

Hi Charles

I'm currently putting all of my efforts into the design of an interface board that will drive either the Relay or a single 5/3 pneumatic valve

Nothing new here, but in brief the feedback from K8055 to pc software is an issue with a system that is capable of such fast movements, with overshoot of position a huge problem - not so bad with wiper motors, as they are quite slow to react.
I dont really want to have to over-regulate the exhaust to slow things down, so am thinking outside of the box and going back to first principles.

My circuit currently in test uses the analogue out (0-5V) to drive a circuit which will find its own position, and automatically correct if it has overshot. This will be performed using analogue components and logic chips, and so will happen in real time rather than every 20ms as with the K8055 analogue input. Crashing will not be a problem, as the last signal sent will stick, and the board will continue to hold position even though the K8055 itself is temporarily unable to assist.

I have had a number of issues to overcome:
The circuit concept requires positive and negative voltage (a simple chip and capacitors used to provide both)
The output from the K8055 is only 0-5V - ideally I need to double this (a simple Timer circuit is used to boost the signal to 0-~10V)
The circuit power amp outputs a single voltage - this could be positive or negative but will determine the direction of travel (PNP and NPN transistors are combined to define which valve opens or which direction the motor spins)
The circuit operates with 5V, 12V and 24V demands, so power is rather complicated - I'm almost certainly going to have to use a computer program to design the pcb... I think this one is a bit beyond trackboard!! (though that is how I am prototyping)

I have a prototype nearly finished. All of the bits of the circuit work fine individually, but I have yet to connect it to a motor or valves and I am back to Amsterdam for work tomorrow until Thursday. Will solder the rest together on Thursday night, and get testing asap!

I can predict how the circuit will behave, so I can use the evenings this week to write the profile... it's just going to take a while to get this one into a video and uploaded!

Watch this space!!

Regards

Tim

Hi Tim,
I am very curious about the video.
I've been a long time working on building my simulator pneumatic cylinders (with 3 speeds)
my project is almost finished, and very curious about your solution. :clap:
best regards Charles
sorry for my bad english