Use 0-10V + dir on 7i77 instead of +-10V out

I have 3 Lichuan AC Servo that I was told that they should work with +-10V. They kind of do but not in a correct way.
The +-10V sets the speed so the speed for +5V = speed for -5V, but in the same direction!
The direction is set by a separate pin.
On top of that the servo vibrates a lot when feed with negative voltage, but it runs smoot on positive voltage.

One solution is to build an interface with some opamps that derives the direction and rectifies the output voltage so it's Always positive.

One other solution that I suppose is possible is to do this in a hal-file or something.
There are a couple of statements in the machine hal-file that indicates it is possible.
net x-output => pid.x.output
net x-output => hm2_7180.0.7i77.0.1.analogout0

I'm completely new to Linuxcnc so any help where to start looking would be appreciated.

The actual Servo but an other seller.

Its entirely possible to do this is HAL (using the absolute component),
but its not likely to work very well as the direction signal will not be synchronized
perfectly with the analog output (VFDs often use 0 to 10V + direction but its not a good way to drive servos)

Are you sure there is not some setup option to get true +-10V operation from the drive?

For the moment I'm afraid that it doesn't work with +-10V. I certanly hope I have missed somthing!

Ok, then it sounds like the best solution is to build an extra interface that generats a direction signal and converts the speed signal to 0-10V if it is not possible to make the servo work as a "standard servo"!

I will contact the seller of the servo and hear what they say about this.

perra_e wrote:

I have 3 Lichuan AC Servo that I was told that they should work with +-10V. They kind of do but not in a correct way.
The +-10V sets the speed so the speed for +5V = speed for -5V, but in the same direction!
The direction is set by a separate pin.
On top of that the servo vibrates a lot when feed with negative voltage, but it runs smoot on positive voltage.

One solution is to build an interface with some opamps that derives the direction and rectifies the output voltage so it's Always positive.

One other solution that I suppose is possible is to do this in a hal-file or something.
There are a couple of statements in the machine hal-file that indicates it is possible.
net x-output => pid.x.output
net x-output => hm2_7180.0.7i77.0.1.analogout0

I'm completely new to Linuxcnc so any help where to start looking would be appreciated.

The actual Servo but an other seller.

There are several things that you will need to do. First post the model number of the servo drive.
Second, you will need to set the correct parameters in the servo itself. Most new servos allow programing to set the correct interface, i.e. +-10V, step-dir, etc. Third, the PID, max speed, input type( position, +-10V, +-5V etc) will need to be set.
When these are set, you will need to tune the Linux loop.

Post images and model number.

Hope this helps

Iron-Man

We got the answer from the seller now and it is confirmed that it works like I have decribed and that I have not missad any parameter.

I have never heard about anything like this before and me and my friend that bought the drives feels scewed

If I shall use this drive an external interface that fix the signals feels like the only option.
This is a picture of how the drive is connected. It certanly looks like you should be able to use +-10V, but the fact that there is an input pin that can set direction talks for itself I think

Yes I feel screwed

Been nagging since day 1 when I started with my inquiry that ±10VDC is what we want and seller agreed to this!
Manual clearly states that AI3 is bipolar so that was clear to me that all specs should be OK.

Awaiting answer from seller.

Here is the reply
as for the ±10VDC, in the beginining, I really confirmed and it is no problem, and we also have specially statement on requirements to the manufacturer but we also don't know why they miss when producing, and I will check with them tomorrow and then tell us your further news.

So, is there any timing issues going via HAL?

If the speed/torque isn't continuous through zero or there is a delay switching direction then that is likely to cause poor servo performance or instability.

Cheers
Bengt