Hello,
I want to control a linear actuator salvaged from an Ikea adjustable standing desk (galant).
I also have a control box that came with the desk and I manage to find the manual and data sheet for it.
The desk consisted of two actuator of type db7.
When I connect them both to the control box it works fine to run the actuators both up and down with the buttons. I want to use only one of the actuators for my project but some "smart" function in the control box prevents me from doing so. If only one of them are connected to the controller it doesnt work. According to the data sheet there is an "auto detection of the number of units connected" that I assume is the reason..
Each actuator are connected with a cable consisting of six leads and I thought it must be possible to fool the control box in some way..?
This is however not clear to me and I would really appriciate any suggestion or advice. I have both mulitmeters and an oscilloscope.
Thanks in advance!
You know that you are in an Arduino forum?
Can you tell us what your question(s) have to do with Arduino hard- and software?
If there is no Arduino involved or you are not planning to use one instead of the existing control box, then you should look for an adequate forum, maybe something "close" to a CNC forum, to post your questions there.
Each actuator are connected with a cable consisting of six leads
-> mightbe indicating unipolar stepper motor.
To prove this assumption you can measure the resistance of the motor(s) like so.
If you want to replace the control unit or powering it with an Arduino, we could help you with your questions after you have verified that the actuator(s) are driven by unipolar motor(s).
rpt007:
You know that you are in an Arduino forum?
Can you tell us what your question(s) have to do with Arduino hard- and software?
I'm planning to use an arduino for controlling the system and my first thought was to control the actuators using the arduino but since the controller with build in power supply seems to work fine for my purpose is feels like it should be easier to use that one and manipulate it by connecting the arduino instead of the buttons.. This requires however that it also works for a single actuator.
I've now opened one of the actuators and the motor doesn't look like a stepper to me, more like a DC motor with a gear box.
Pics of the control box migt say more to you than to me:
the two cables connected in the middle of the PCB are the ones going to the motors and the small PCB to the right in the first pic is the buttons to control up and down motion.
I anyway measured the resistance and got the following result:
wire # - wire # resistance
1 - 2 1.29 Mohm
1 - 3 OL
1 - 4 1.29 Mohm
1 - 5 1.29 Mohm
1 - 6 OL
2 - 3 OL
2 - 4 0.8 ohm
2 - 5 13.2 kohm
2 - 6 OL
3 - 4 OL
3 - 5 OL
3 - 6 0.8 ohm
4 - 5 13.2 kohm
4 - 6 OL
5 - 6 OL
I assume there is no use of this information more than to conclude this is not a stepper..?
the lables on the controller and the motor:
Any idea of how to proceed?
OT:
To further emphasize that this is indeed an arduino project I should give some background and mention the purpose of the whole project..
As I started with electronics and specifically arduino about two years ago my interest in the topic grew exponentionally and soon my small apartment was over loaded with arduinos, breadboards, electronic components and messy as never before. Last week I decided to do something about it and dedicate some permanent space in the apartment for electronic work and build some suitable storage for all the stuff.
I used old pallet wood and this is what I ended up with:
The computer screen is attached on drawer slides and it can slide up and fit perfectly behind the sortiment boxes with electronics above the desk:
Lifting the screen manually and put stuff under it to support it is not optimal so I guess its now clear what the actuator should be used for and its relatively close connection to arduino...
If the wires listed in the table above correspond to the 6 motor box leads, it is hard to say what it might be.
A two-phase stepper would have two coils of rather low resistance, but I would not expect the coil resistance to be as low as 0.8 ohms (wire pairs 2-4 and 3-6). Each coil would draw about 14 amperes from a 12 V supply, which seems unreasonable considering the motor plate (7 amperes).
I have no idea what the other resistances might represent, but I would expect there to be limit switches in the slide mechanism.
Unless you are rather good at electronics, it is nearly hopeless to reverse engineer that controller. It has a microprocessor and some high power motor control electronics.
If my eyes don't cheat me, each motor is connected via 6 wires (see the picture with the pcb and the 2 connectors at the right side in the middle).
The measured values are a bit strange and very different from what I would have expected.
Maybe you are right and it is a DC motor and the additional wires correspond to a position controller; but that is just a 1. speculation. It would get clearer if you had an oscilloscope and drive one motor and measure the high ohm wires with the oscilloscope. But - again, I am just guessing around what it could be.
If I find time today I will see what Aunt google can tell me.
Additional information taken from your pictures and asking Aunt Google:
The motors seem to be DC's with a lot of power (looking at the information on one of your pictures, where the datasheet says: 12V/max 7Amps or so.
In a brochure about the CBD4 control unit it says, that it is able to correspond with hall sensors (imho as limit switches), although not all of the actuators are equipped with hall sensors.
But they built the CBD4 control box as "one for all" purposes.
If you have a separate power supply (5V and at least 2-3A) you could try to connect the "0.8Ohm wires" to that power supply (without having the motor connected to the control unit) and see if the motor moves or at least you can feel a stroke - immediately shut the power off, when that happens. So you can test if you got the right wires to move the motor.
I was offered a non working one of these i think but could not get it in the car.
I think that it is a simple 2 wire dc motor, the other being sensors.
Possibly similar to a servo with a gear driven pot /limit switches , when turned do any of the values change ?
I found a little bit more about the control box (another type, same company, but I guess it will also apply for all of their CBD control units.
Attached find the picture which I found at a German site.
As you can see they produce "one for all" models and it depends what actuator they use if there is nothing, limit switches or hall sensors inside the mechanical actuator case.
Highflower:
According to the data sheet there is an "auto detection of the number of units connected" that I assume is the reason..
Each actuator are connected with a cable consisting of six leads and I thought it must be possible to fool the control box in some way..?
I doubt its trivial - on the desk moving one actuator without the other would bend the desk, so the controller
simply doesn't allow it.
This is however not clear to me and I would really appriciate any suggestion or advice. I have both mulitmeters and an oscilloscope.
Thanks in advance!
Best regards
Olle
Think you either drive both actuators together, reverse engineer the control box, or buy a suitable motor controller.
If you are lucky and figure out the connections you may be able to feedback the position signal from
one actuator to both inputs on the control box, but I'm guessing - open up the box and get some photos?
thanks again for all good answers and the time you've spent on this!
I've done some more testing and also taken the "gear box" on the actuator apart to get a better understanding of it. (I had to brake three small plastic clips to get it apart but nothing that cannot be fixed with some epoxy)
In the end of the screw (close to the motor) there is an mecanical switch which is pressed when the nut is in it's end position. Only three of the six wires are connected to the motor and the remaining wires are connected to a small circuit board in the plastic box connected on the motor. another small pcb is also inside this case.
Some pics:
Does this reviel any new information?
I've also measured the resistances again and I found some differences this time. The high ohm combinations were not measured this time and I guess I had a bad contact last time I measured. What is even more interesting is that most of the resistances were changing when the end switch was pressed, some of them just a bit, others a lot.
As can be seen in the table above two of the wires (the ones I chosed to call #3 and #6) are not connected to anything but eachother ("over load" for any other combination involving them). This result is repeated both with the swich closed and open.
New resistances for the remaining wires switch not pressed (wire numbers in rows vs wire number in columns):
Wire no ->: 1 2 4
Wire no:
5 19.2kohm 11.02kohm 13.2kohm
4 9.2kohm 0.8ohm
2 9.12kohm
with the switch pressed the following was achieved:
Wire no ->: 1 2 4
Wire no:
5 23.0kohm 11.76kohm 20.4kohm
4 9.2kohm 9.2kohm
2 17.9kohm
note that the small resistance between wire 2 and 4 is changing to a high value when the swich is pressed and that the resistance between wire 4 and 5 as well as between wire 1 and 2 is almost dubbled.
The other smaller differances was clearly visible when measured. I connected the multimeter and pressed the swich while measuring and the value changed for all combinations except between wire 1 and 4.
I had an idea to just connect one of the actuator to all the six different connectors in parallel on both places on the PCB and thought the controller might interpreat it as two different actuators but after a closer look at the controller PCB I realized this was not a good idea..?
As might be possible to see on this pic, the lower right pin in the upper connector is on the PCB connected to the lower left pin on the lower connector, i.e. the connectors are not identical. These two pins (the lower ones on both connectors) are the ones that I excluded from the new table with resistance measurements (wire 3 and 6).
Is next step to connect the oscilloscope and what to measure in that case? Or is it maybe trivial for you experts out there to figure it all out now? If you think its easier to get a new motor driver and control it with an arduino I'm definately gonna do that but then I need to know what to buy..?
In worst case I can connect both actuators to the computerscreen or hide the other one as a "dummy" behind the desk but it doesnt feel like an optimal solution...
The little circuit board in the gear housing is almost certainly a Hall effect sensor that counts revolutions of the worm gear. That tells you the overall extension of the lead screw. It would be useful if it can be reverse engineered, but is not necessary for manual operation.
Trace the actual motor leads and tell us what you find.
Well you can see the 4 H-bridge MOSFETs clearly on the main board, shouldn't be too hard to figure out
what's driving their gates if you trace the signals.
My guess is that 3/6 thick connections are to the brushed D.C. motor, apply +/- 12V to run the desk up or down.
You could use a Pololu motor driver for this.
Red=+5 V power to the Hall sensor, thin brown = sensor ground.
Yellow = count pulse output.
Green = limit switch.
You can find 12V, 16A XBox power supplies at thrift shops for about $10, or on line for a bit more.
What is even more interesting is that most of the resistances were changing when the end switch was pressed, some of them just a bit, others a lot.