K8055 - Dual Wiper Motor Control with Variable Speed

nope - only when I was using a 100K pot for feedback... and changing to a 10K pot put that right

It seems pretty stable...

Tim

Hi Bigtalltim,

Thanks again for your quick reply.
Big Thanks for your post explaining speed control with PWM signal, and Direction control with Relay. Are you a teacher in the life (smile) ?

I would like to show you my prototype with a video, but I fear that I'm not able to post a video very soon. I will do my best...It should be a good solution, it is easier to understand everything with pictures and videos.

When I say that is not what I expected, I mean I have no variable speed until now.
But my first problem for the moment is : I would like to stop the motor when it reaches the posisiton, as I can see on your video when you are moving the slider 21.

See you

No problem Adbgg... I have been meaning to post an explanation for ages!

Dont worry if it takes a while with the video... I have another request for you (and Riton!):

Please could you post your current profiles so that I can have a look at trying to adapt the original K8055 profile to drive your boards?

The more I think about it, the more I believe that this is the way to get basic speed control out of your setup which does not use the PWM output... adapting this original profile should give 3 movement speeds in each direction through the use of more or less pulses per routine as with the original setup which was for pnuematic actuators (I think)

One last question for you: is the position pot having any effect on the movement at all, or is the setup completely ignoring the feedback data? - this will help me to work out whether this is a tuning issue or a potential profile rebuild issue...

Tim

PS. I have been many things over the years, but yes... I spent a few years as a technical trainer!

PROTOTYPE PROFILE

Hi Adbgg

I have had a look at the profile, and the setup in your post Small Prototype / VM110 + Sabertooth 2 x 10A

I have tested this on my spare K8055, and it seems to work well - the pulses are visible through the pot LD9, so I can see it working... of course, I have no way to actually test it here...
You may like to test with just your K8055d before hooking up the Sabretooth

Have a look through the profile, and test if you are happy with it - if it works, we may be able to develop the idea further.
(I can't think of any reason that this should cause you problems, but make sure you agree that it is OK before turning it on!...)

PLEASE NOTE - this is for a card with an address of 1, so you have to take the jumper off SK5, and leave the jumper on SK6, or profiler will tell you that an interface is unplugged!

PROTOTYPE PROFILE K8055d and Sabretooth.rn2
In theory, this should give you pulse based speed control (CRUDE pwm) with your setup...
Here is how it works:

To move in a positive direction, the Analogue output is set to:
0%(Full speed forwards), then 50%(stop), then 0%, then 50%, then 0%.
Once the finished position is reached, stop kicks in

To move in a negative direction, the Analogue output is set to:
100% (Full speed backwards), then 50%(stop), then 100%, then 50%, then 100%
Once the finished position is reached, stop kicks in

To stop, the Analogue output is set to 50%.
This is neither one direction or the other, so the sabretooth stops... in theory!

The original RocknRide profile had 3 speed settings that simply use different numbers of pulses... slow speed has more, fast has less.
To test the concept, I have removed one of the axis, and two of the speeds... they can be easily be put back in if this works.

Tim

THANKS Bigtalltim,

Unfortunately, I can't load your profile. I've taken out the jumper SK5 and let on the board the SK6, but X-SIM is not able to load the profile. I try the other combinaison (SK5 on the board, and SK6 out, it can't help me). It is curious, because with the Velleman software, I'm able to connect the board.

Anyway, I've done other tests this morning and I can say now that I'm very close to the solution.

BUT YOU ARE RIGHT.

I must configure 2 or more speeds in the menu output of X-SIM.
I've already tested with only 2 speeds, as follow :

1) The first speed is set at 100% (positive), 0% (negative), and 50% (stop).
2) The second speed is set at only at 60% (positive), 40% (negative) and 50% (stop).

My potentiometer is not very precise, but everything works as expected :
- The motor moves quickly when it is far away from the position to reach.
- Then it moves slowly near the position,
- And, Last but not least, the motor STOPS when it reaches the position. That's a good a news.

AT LAST, IT WORKS !!!

As you can imagine, I'm really happy today to realise that I've found the solution.
I hope it is really the solution, but it seems that is the good one.

Of course, I will put a video (or picture) when I have time to do.

The problem seems to be solved : Adbgg + Riton + Bigtalltim => K8055 + Sabertooth = SOLVED.

Excellent! :clap:

Using the preset ranges (like 0%-100% and 40%-60%) as a way of controlling speed is inspired!

Way more elegant as a solution than pulsing routines - good work!! :thbup:

I'm looking forward to the video :

Tim

The latest test on my own setup went well today - I am now totally convinced that the K8055 is suitable for at least a 2DOF using the Analogue inputs on the board for positioning feedback

The videos below show one K8055 driving 2 separate actuators...
I have locked the lower arm in place to stop them flopping around all over the place - this type of actuator really needs to be locked into a triangular frame, but I have only 2 built so far!

Ignore the fact that the actuators are not perfectly in sync... the pot on the left one is a bit dodgy,
and the motor speed needs adjusting (by twisting the cases)...

...That said, I am very pleased with the results!

Same direction using slider 21,
but driven on seperate axis with separate position feedback pots:
http://www.youtube.com/watch?v=MzkmZyJEEm8


Different directions using slider 21,
and driven on seperate axis with seperate position feedback pots:
http://www.youtube.com/watch?v=z6fnlhbK-NE

The next test (hopefully tomorrow) is to drive the two actuators from two K8055s - one on each board

Tim

For anyone following this thread for the Sabretooth option, please click on the link below to the thread by adbgg for full instructions on setup, wiring and information on how to set up the profile:

small-prototype-vm110 + sabretooth 2 x 10a
(VM110 is the code for the pre-built K8055 )

Connection diagram for quick reference:
[img600px]http://nice.maurice.free.fr/X-sim/Test/K8055-Saber-potar.jpg[/img600px]

Many thanks to Adbgg and Riton for their work with this hardware option! :thbup:

Tim

Like Menix I seem to have fried a K8055 board. I connected the circuit up correctly, checked it three times to make sure, turned the power on (I check the output of the PSU before I did this), the K8055 was already plugged in and functioning. I am using the circuit with the extra transistor as protection but I must have done something wrong. There was a broken trace on the board and the PIC is dead. I repaired the trace and replaced the PIC with another from my second kit and it is working fine but I am not going to connect the circuit again until I find out what I did wrong.

Perhaps I can post pictures of the breadboard and you might see what is wrong?

Hi Historiker

Please do post pics of the breadboard - it might be difficult to spot the problem, but two pairs of eyes are always better than one!
Please also post the layout of the circuit that you were building, and if possible the footprint layout of the relay...

A couple of questions for you:

Did you get the K8055 built up or as a kit?

Where was the broken trace on the board, and was it a burned out track, or already broken? (that might help identify the source of your problem)

Tim

Hi Tim,

The K8055 in question was pre-assembled. It worked out of the box very well until I hooked it up to my circuit. I had also purchased another K8055 in kit form to use for my third motor.

The trace was burned, there was no indications of power or recognition from the computer until I repaired the trace. After that I got power to the power LED (LD110) but not other lights (LD8 did not flash), the computer recognized that a USB device had been connected but did not recognize the device.
I replaced IC3 with one out of my other kit and the K8055 lit up (LD11 on and LD8 flashed three times as it should). Windows recognized the card and I could run the demo program and get it to work in X-Sim as I had before I tested it with my circuit (and the subsequent burn up). I used your profile in X-Sim, works very well. Of course I could only test via software as I had not connected the relay circuit up yet.

The circuit that I used was the one that you posted for 24vdc use (page 12).
Components:

NPN transistor TIP3055 for Q1

PNP transistor 2N3906 for Q2

R1 and R2 are 1k

The relay coil voltage is 24v, 37mA current and 650ohm resistance, DPDT (LB2-24D-S-R)

Q1- the emitter is connected to ground, the collector is connected to D1 and the relay switch (pin 5, green wire) and the base is connected to R1 and PWM1(red wire) on the K8055.

Q2- the emitter is connected to D2 and the relay coil (pin 8, another green wire) the base is connected to Digital Output 1(green wire) on the K8055, and the collector is connected to Ground on the K8055.

The other relay coil lead (pin 7) is connected to 24vdc and the other relay switch (pin 6) is connected to 24vdc.

The coiled wire (red and black) is the plug from my PSU (24vdc 47A).

Hi Historiker

I have only had 2 minutes to look at this, so have not been able yet to work through the whole circuit...

First impressions are difficult because I have never used one of these boards to connect a circuit... but...

It is possible that this transistor is in the wrong way... check the datasheet - i think you have collector to D2 and Emitter to ground... havent had time to download the datasheet but as I remember, the leg layout is e-b-c when viewed from the underside with the flat section at the top.
(I am not sure of the implications if this is so...)

Please could you post a pic of the burned track on the K8055 - the position of the burn may help to identify the source of the problem

I will look in much more detail later on tonight.

Tim

Thanks for the quick reply Tim, you are correct. I did have it wired backward. But would that cause a feedback into the K8055 that would result in damage?

Here is a graphic from velleman showing the pcb layout for the K8055. I put a light green circle around the trace that was burned out. It runs between the two solder joints.

let me research that one... any circuit error could be dangerous, but unless this resulted in the current back flowing through the base I'm not sure...
(by the way, I'm not sure if that can actually happen!... but if it could it would probably fry your K8055)

There may be other problems... check everything again just in case - also, consider that if the transistor has been damaged, it may be possible for the same problem to occur even if has been turned around...

Tim

Good point Tim, I did actually replace all parts of my circuit just in case. Now I am wondering if the transistor (2N3906) had been in the correct orientation when the board blew and did I put it in the wrong way when I changed the components? As I said, I triple checked my circuit (breadboard) before powering the system on the first time. I would like to think that I did not miss it three times in a row But I would not put it past me, lol. Oh well, it is all a learning experience. The cost of a new K8055 hurts a bit but the components themselves (for the relay circuit) are cheap and easily replaced. I wonder if I could get a replacement PIC (IC3) from Belgium if I sent in the one that I burned out? I will post that question on their forums.

Is there a reason why you changed to a PNP Q2 transistor? It will work, but the PWM-Output will be inverse of NPN.

This holds true for PNP:

NPN: Current flows from Collector to Emmiter. Biasing current flows from Base to Emmiter.
PNP: Current flows from Emmiter to Collector. Biasing current flows from Emmiter to Base.

The burned track is just a thin connection between the two ground layers. These two ground layers should be connected by other, much thicker connections as well. With that being said, the track was burned because of high currents and it turned into a fuse.

Can you post a picture how you fixed the ground plane? You should always separate the high current ground path (going straight back to the supply) from the digital ground path.

My first bet is that you connected D2 in reverse which shorted out the supply when the transistor turned on (as soon as the circuit is powered, since 0V turns on a PNP), causing high currents through the ground plane and the very thin track.

Hi Frakk, thanks for taking a look at my mess

I am a little confused on your question about Q2 though. The circuit design that Tim posted showed a 2N3906 (PNP) which is what I used. Should that transistor have been an NPN?

I think you might have hit the nail on the head though with You should always separate the high current ground path...from the digital ground path I have ground from the PSU directly to the K8055.

D2 was connected correctly, that was one of my main worries when I put the board together, getting the diodes in the proper orientation. Could the mistake be in the circuit design and D2 should be reversed because Q2 is a PNP transistor?

panic not - the parts for the K8055 itself are also cheap and apparently quite easily replaced!

I suggest you have a look at this thread...

http://forum.velleman.eu/viewtopic.php?f=3&t=3392&view=previous

If that isn't the chip you need, then a further post in that thread wouldn't be out of place!

Tim

PS. @Frakk - do you think that the design of the circuit has an error?
The circuit below was posted in the Velleman forum by the designer of the original relay circuit, and includes both PNP transistor to switch the relay, and direct connection of -VE battery to ground...
If these you feel that these are risky inclusions and there is a safer option, I would like to amend the diagrams urgently, and bring the change to the velleman forum also!
I'm afraid that my own knowledge is not enough to question the design... :

Sorry for some reason I thought NPN was used as they are more common.

There is nothing wrong with the circuit, PNP or NPN doesn't really matter. The only difference is that the direction logic will be inversed.

The battery also needs to be grounded to prevent any floating voltages across the devices connected.

ps. The pull-up resistor on the PWM line should be connected to ground for a pull-down. I can't see any reason why you want to have the motor turned on if the K8055 is missing...

Hi Frakk, thanks for clearing that up. Could you elaborate on grounding the battery. I am using a power supply, where would I ground it to? There is a connection from negitive on the PSU to the ground tab on the K8055 as per the diagram above. Is there another step I should be taking?