K8055 - Dual Wiper Motor Control with Variable Speed

CONTENTS OF THIS THREAD:
As the thread has grown in size it has become increasingly difficult to find content quickly...
Here are the locations of some of the key points, demos and design considerations:

Page1/Post1: CIRCUIT - 12V Relay circuit design
Page1/Post1: VIDEO-DEMO - Explanation on relay setup
Page1/Post1: VIDEO-DEMO - Demo of multiple wiper motors
Page1/Post1: PROFILE - Basic build of relay setup (No position feedback)
Page1/Post8: CIRCUIT - Frakks MOSFET version of Relay Circuit
Page2/Post8: CIRCUIT - Track layout for 12V Dual Relay circuit
Page2/Post8: EXPLANATION - Description of Relay circuit build
Page2/Post9: VIDEO-DEMO - How to set up K8055 for Position Feedback
Page2/Post9: CIRCUIT - Wiring diagram for connection of Pots to K8055
Page3/Post1: VIDEO-DEMO - Frame - Initial Proof of Concept Test
Page3/Post5: CIRCUIT - How to connect K8055 to Pololu High Power Motor Driver 24V20
Page6/Post2: VIDEO - First test of Sarrus Linkage Actuator
Page6/Post2: VIDEO-DEMO - Comparison of RolandVanRoys solution and my own including design concepts
Page7/Post2: VIDEO - Second test of Sarrus linkage demonstrating good positioning response
Page7/Post9: EXPLANATION - How to use Test Tablet
Page11/Post1: EXPLANATION - How to connect multiple K8055s to a single PC
Page11/Post5: EXPLANATION - PID Control concepts and link to further reading (wikipedia)
Page12/Post9: CIRCUIT - 24V version of Relay circuit (see also Page13/Post3)
Page14/Post4: CIRCUIT - ADBGGs circuit layout for K8055 and Sabretooth 2x10A
Page14/Post9: EXPLANATION - HOW THE K8055 and RELAY SOLUTION WORKS!
Page14/Post9: CIRCUIT - Track layout for 24V dual relay circuit
Page15/Post7: VIDEO - Further tests of Sarrus linkage type actuator
Page17/Post1: EXPLANATION - Screw thread actuator concept - fail
Page17/Post5: EXPLANATION - First steps towards Chain Drive setup
Page17/Post6: VIDEO - Chain Driven Linear Actuator - Proof of concept test
Page17/Post9: VIDEO-DEMO - How to wire a switch to change motor direction
Page17/Post9: VIDEO - Extreme Stress test - Demonstration of power available through gear reduction
Page18/Post5: EXPLANATION - Tronics solution - How to wire Automatic Direction Recovery Diodes
Page18/Post7: VIDEO-DEMO - How Auto Direction Recovery Diodes work
Page18/Post10: VIDEO-DEMO - Switches at Extremes of position with Direction Recovery Diodes
Page22/Post2: EXPLANATION - How to connect up to 3 motors with bi directional control or up to 6 extra motors to one K8055
Page22/Post2: CIRCUIT - H bridge to use spare digital outputs and enable a 3rd bi-directional motor with one K8055


For all those who have asked about the Velleman K8055 and it's use for wiper motor control...

Here's how I have done it

This setup allows variable speed through the use of Pulse Width Modulation (PWM) and bi directional motor control of two windscreen wiper motors. Other solutions I have seen (including the basic preloaded profile in X-sim for the K8055) do not use PWM, and have to have complicated pulsing routines built in to slow and change direction...
This is way simpler than all that!

Also, this uses relays so there is practically no power loss to heat unlike MOSFET H-bridge alternatives.

Better than any other single benefit, this setup is cheap - super cheap!

The complete double wiper setup shown in the video has cost me no more than £75...
I costed the components for Tronics AMC alone at about £73 locally to me, and then you have to buy the hbridge bits... and then the motors...
Way too much for me to spend right now! - besides... I already had a K8055

Below is a list of my spend so far:
Velleman K8055 prebuilt £30
Wiper motor from ebay - £14 no postage
wiper motor from ebay - £12 + £6 postage
DPDT relay £4.99 each (x2)
Relay Socket £1.69 each (x2)
1K resistors £0.14 each (x2)
4 diodes (from a lucky bag... about £0.10 each)
2 NPN Transistors (from a lucky bag... about £0.15 each)
PCB sheet £1.54


I already had some bits lying around (so there was no cost unique to this project):
Power supply from an old dead PC
Etching fluid
Solder
USB Lead for K8055
Old wires to connect it all together
PC Fan to cool - not required for test, but may be needed for the transistors under full load
Connectors to hard wire into the board to connect to power supply
2 Junction blocks to make connecting the wiper motors easier

The circuit solution I found in the velleman forum:

all credit to paledin micro for the circuit design

From this circuit, I designed my own PCB to use cheap locally available relay parts.
The circuit is simple and takes little time to make - even for an inexperienced PCB maker like me!
To use other relays the design may need to change, but if anyone is interested, and I will draw it up and scan it in - I'm afraid the photos I have aren't great (as you can see here! )



Here is a video showing the two motors working independantly
http://www.youtube.com/watch?v=NH3vymnngT4


Here is a video where I give a description of how it works
http://www.youtube.com/watch?v=m5wlM9Irxnk

The profile is extremely simple - only two lines for each direction per axis:
BigTallTims version for Dual Relay setup.rn2

Enjoy!

Tim 8)

Thank you Tim for sharing! :

I would like to see how hot these lucky bag components get...

Let's put some thought into these components and it will be TOP WORK! :clap:

This can also be connected to the AMC to replace the dMSHB for those who are looking!


oh, and the best part is: It is so simple you can build it on a perfboard, even without copper on it.

With no load on the wipers, with it running for approx 5 minutes on a slow speed, I could still touch them...

On the medium or fast speed, they are too hot to touch with no fan - the fan worked really well to cool them.
I'm afraid I am no expert on components - they are marked TFK 9606 BUX 127 VDO

Tim

is that better then thanos controller?

obisk: yes and no...

why yes and why no?

Because I'm not your electronics teacher...

Google is your friend!

Maybe Wikipedia?
http://en.wikipedia.org/wiki/Relay
http://en.wikipedia.org/wiki/Mosfet


And better in what? Price? Performance? Simplicity? Size?

Here is a diagram if you want to try it with MOSFETS:



Make sure to use LOGIC LEVEL ones if the PWM voltage from the K8055 is ~5V.

Current rating should suit the motors, about 15A continuous should be good for those wipers. Don't forget heetsinks.

Superb - Many thanks for this! :

I will try to get these parts and build this as well!

When I get the components and have this built, I will post a video of this alternative side by side with my own version to compare the performance

Tim 8)

Very nice speed controller!

But... position feedback potentiometer? where you connect that?

The PID positioning has to be done in software using some analogue inputs, or an external hardware controller. This is why I said it can be used with the AMC!


On an other note:
Do you have time to work on the firmware and make improvements on it?
One thing I could give you a tip on: Use Data packet sent at simulator startup in the USO to set up the PID gains. Eliminates hyper terminal and save memory.

Is it better in Performance then thanos Controller?

Depends on what components you are using.

Relays can usually carry a lot of current! The bottle neck is the PWM MOSFET and the recirculating diode (if used).
It will be cheaper to buy 1 MOSFET than 4, so you can go even larger on it.
You can get more current out of this bridge for the same money and much simpler build. It is more reliable IMO, even though the relay contacts will wear out over time.

The disadvantage of the relays in performance: You cannot change the motor's direction as fast, and you hear the clicking when it does.
This could be a good thing too, to protect from high spike currents when they are quickly reversed.

Thank you

The position feedback pots go onto the 2 analog input channels on the K8055 - just like JimNZs flight sim.
Just as Frakk says, the position processing will be software... so the AMC will still have the edge!

Tim

okay thank you frakk...

i test first thanos controller an later the K8055 ;D

I already made new firmware version on this matter... but will be able to change the PID gains on the run too if needed... send back to xsim the position of the pots and.... Shhh... its a surprise!!!

Sounds good!

For this relay motor driver it might be necessary to add a little delay time to PWM when changing directions to let the relay switch properly.

Only testing will tell!

Couple of images for anyone interested:

As Frakk says, this could be made on pretty much any experiment board... now I didn't think of that, and chose to make a PCB... :

...but thats ok, cos I'm gonna say that it was so I could make the tracks extra big to take the current! 8)

In case you want to try a pcb (or use the same relays) this may help... My layout for the double board was as follows:



My layout is designed around a DPDT relay from Maplin. Their code is N08AW. The relay has spade connectors, and fits into a socket - this makes it easy to change when you go past 10 million or so operations and the contacts wear out!

NOTE - the max number of operations is stated by the manufacturer as only 100,000 at full load, but as the motor is stopped when it changes with my setup, it never gets changed at full load! - 10 million it is!!
BUT I guess that is something to take into account if using this relay bridge with Tronics or any other controller

The relay goes into socket N32AW, which is soldered onto the PCB.

It could probably be layed out better, but not by me... : ...so I had to use bridging wires to complete the circuit
(these are shown in colour on the circuit layout picture).

(if you look carefuly there are 3 wires, though 2 of them are one on top of the other so it looks like only 1 around the relay socket...)

Now the Power.
I used connectors from old PC expansion cables (the ones that allow you to power additional drives).
I did this to benefit from the fact that the power sockets in a PC power supply give you both 12V and 5V - both of which I need in this set up. (Yellow is +12, Red is +5V, Black =Gnd) and also that they often have fans wired in already!



The 5V wires on the other connector (they are not shown in this picture, but will be included later when I have the frame built) are used to power the two position indicating pots on the K8055 analogue inputs.

Using a PC power supply seems to work fine - I may decide that it needs more power when I have the frame finished and can test it under load, but for now I like the convenience of getting two voltages out of one unit, and both motors seem to be running nice and fast.

I have the base of my frame welded together now - I have made this from the scrap steel sides of an old bunk bed, so this has cost me nothing so far except a few welding rods... I believe I am on track for a full motion sim built for less than £80...hopefully I will have the motors mounted and the platform moving by this weekend...

More will follow ;D

Tim 8)

Connecting Position Indicating Pots to K8055...

Nearly done now - just the top of the frame to finish! :happy:

Here is a critical piece of information for anyone wanting to try this build

I have recorded this as a video because it involves changing a few settings in hardware...
...much easier to show you instead of try to describe it!!

http://www.youtube.com/watch?v=gamF95vtads

The layout to connect pots to the analogue input is as follows:


Tim 8)

Good progress Tim! :

I hope your simulator will move very soon! Getting close there!