AMC1280USB Motion Simulator Controller

@Gabor Pittner ,

I don’t have any experience with sabertooth config.

Is there any other way in your setup to deal with the motor acceleration than the controller PiD ?

Asking because that’s the case with ac setups and servo setups where you have to find a sweet spot for a proper acceleration

Do you have the Sabetooth setup as Linear or Exponential? I strongly suggest setting it to Exponential so there will be not much drive near the target position. ( on Sabertooth 32A this is setup via dip switch 5 ).

Thanks
Thanos

Or you can take it a step further by setting up a custom throttle curve using Describe software giving you total control of the process.

It has analog HAL sensors

If I remember well I set it to exponential, but I have to check it. Thank you.

No I can set motors acceleration in AMC board PID settings only.
In analog control to sabertooth.

By the way the I integral and D derivative values don't affect anything of analog mode with sabertooth. I tried to change 0-300 both values. Maybe with D changing I get some noise while motors keep position but no else happen.

Hi @Thanos , I noticed when P proportional value is relatively low under 15, the first and second motors start moving some tenths of a seconds lately than the other four. You can see in my video on heave, motors are out of synchron.
Watch from 0:15 seconds when I slowly move heave slider. Is that normal?

I would advise the digital upgrade

I use this type of sensors: Link

You can try the interface with those ones:

https://www.mouser.fr/ProductDetail/Broadcom-Avago/AEAT-6012-A06?qs=9/rGSQZ%2BqojKLSrjWhl6sA==

https://www.mouser.fr/ProductDetail/Molex/15134-0503/?qs=/ha2pyFaduim/QRDJMQYDlBf4%2b4y%2bKd/8OTCBBYsbDHOJ3FyZCs7/g==

Double check that your PID is same for all motors.

Reselect the values and re-apply for all 3, p I and d

It’s easy to make a mistake there with the rotary selector and apply the changes only for one motor

Yes I know that, I've made mistakes several time because of that. You couldn't even think how many times I checked
Morover I tried the 1st and 2nd motors P values to higher than tho others. But it was bad idea
When I set P values up to 50-60 or more, the motors move the same time together. But its too agressive and more jerky.

The hall pots he's using are 12 bit also and I believe the AMC has a 12 bit ADC? If so I don't see a huge advantage in switching to 12 bit encoders when not using more than one revolution. As a matter of fact the halls he's using are 180 degree so the resolution he has now is double what the encoder would be not considering hysteresis.

Increasing the PID's "I" setting should help alleviate motors that are lagging behind if its working properly.

"I" is pretty simple to implement in code so I don't think such an important omission should happen.

Unfortunately PID I setting doesn't change anything in my case. It doesn't matter if its zero or even 300... nothing changes.

Did you try increasing "I" with no "D" when using a low "P" value? You should notice the motor reaching the set point sooner than with no "I".

I will make a try, but If I remember well I've tried several different configuration, and only PID P value modified the behavior of motors motion I and D did nothing...
I will try to make a video of that.

Do the heave loop test with HeXpod software. (Loaded)

Low speed big oscillations. Post it here

Such a good idea, I will do that.

I can only say from experience, having tried them both, I’ve experienced more stability with SSI.

Also, kinematicwise, depending on geometry, sometimes there is an advantage of using 360’s as the curse greater than 180deg can be needed