jrk12v12 too often connection lost!

Hi all!.
I am using 3 jrk12v12 (plugged via usb) to handle my 3 12v motors using a 10turn pot to each one of my jrk.

Ok the little jrk working good but often I get connection lost!
In 10 tests I get 4 or 5 times that one of my jrk lost the connection with the pc.
And all the 3 jrk have suffered this problem so I think it's impossible that all my 3 jrk are broken!

When this happen I have to unplug and replug the usb cable and the problem is solved. (I tried these tests in 2 different pc getting the same problem)

So, why is that? I think the jrk is not optoisolated. Can be this the problem? or maybe my USB cables?
And if the problem is the optoisolation so I need just to optoisolate all my 3 usb port right?

I would just to can to use my 3 jrk continuously and without problems!
Please help me!! :

How long is your USB Cable?
Did you tried an active USB Cable?

Hi I have 3 usb cables:
first : 140 cm
second : 100 cm
third : 100 cm

Sorry for ignorance but.. how can I know if a USB cable is active? My 3 usb cables are normal usb cables I think..
here the pic:

This happened to me periodically with one of my units up until a week ago. I replaced the internal usb cable(between the jrk and usb hub) with a shorter one and rerouted the power cables so that they were running farther away from the USB cables, and it's been perfect ever since.

I have experienced this too, especially when torturing the motors and (possibly?) power was generated and fed back into the jrks because the motors were pushed back by the weight of the platform. Probably this annoys the jrks? Could there be a relation with your 3DoF heave?

Hi tahustvedt, please tell me if I understand your situation (My english is always a bad crow in my shoulder..):

1) You are using a USB hub. Your USB hub is connected to one of your PC USB ports and provide you 3 or 4 other USB ports, right?

2) You have changed the USB cable that plug your jrk to your usb HUB in one shorten usb cable.

3) I don't understand the topic about the rerouted of the power cables!.


@value1: Hi!, in fact 8 times on 10 I get the problem when I torturing my motor too.. mmm.. This should to mean that max current is overtaken but If the max current is over the limits so the jrk should to show you an error in the error tab and NOT to turn off.. (I solve the problem reinserting the usb cable and my jrk back to life!)

I really hope it's just a optoisolation problem..

That is correct. I should also mention that I did lose contact with it before I installed the hub as well, when the JRK was connected directly to the PC.

By rerouting the power cables I mean that I moved the high power cables for the JRK farther away from the USB cable, so that there is less chance of electrical interference.

Regarding overcurrent. What is your max acceleration set to? I think it needs to be lower than 600, or the JRK won't be able to keep the current from going over max. My JRKs are set to 35 A max and max acceleration around 20-50.

I think he just moved the power cables away from the jrk usb signal cables.

Same here - so far no jrk has died but has just been confused and needed some motivating :yes:

Perfect! now I understand well.
My jrk current configuration is:
MAX CURRENT : 22.5 amps
MOTOR DUTY CYCLE : 600/600
MOTOR ACCELLERATION : 600/600

mmm I should to try to get up the MAX CURRENT to 35Amps as you but mantaining DUTY CYCLE and ACCELLERATION on 600/600 both..!

@value1: haha yeah dude I am trying to do some serious motivational speaking to my 3 little jrk but seems they want just to kick hard my ass!.. I think they start to hate me... haha

Have you tried reducing the max acceleration a little? It might not reduce the performance noticeably. The manual for the JRK recommends that the acceleration is limited to achieve accurate current limiting: For accurate current limiting, acceleration should be limited; otherwise the duty cycle will tend to oscillate when the maximum current is exceeded.

http://www.pololu.com/docs/0J38/3.e

very interesting dude!.
I feel a bit less sad after these advices.
Tomorrow I will do some other tests, I can't see the time about! I want to solve this problem as fast as possible. Without reliability of use all the project shaking like under an earthquake...
Tomorrow I will write here an update of the situation!.
This problem can't continue for a long time again!

i run mine all unlimited.
I use a small dc battery in paralel with my power source to capture feedback amps when motors free spin plus only 1m usb cables as well.
Another problem is lowering the packet speed too much causes card drop out too. between 20-33ms seems to work ok for me longest run was 6 hours before a heat issue poped up.

Hi mate! I don't understand well..
you have your battery in parallel with another power source to capture feedback amps when motor free spin.... mmm... capture feedback amps???
I don't understand well :\
(I am using 2 car batteries in parallel)

for packet speed you mean the X-SIM profiler in the USO page? (the classic with a pause of 33ms write?)
In fact I am using 33ms!

6hours without problem is a great result mate!! Are you using USB hub or just your jrk plugged in the PC usb ports?
And MAX CURRENT, DUTY CYCLE and ACCELLERATION?

i use a hub and a 20amp desktop power supply and a 12v dc battery in parallel.
I will give i run today and try to recreate the issue to find a solution this end.
What pid period do you use and how agressive is the pid. As this can cause issues too but normally ya get a motor error inturns only results in a splt second delay and the card continues without issues.
I had to increase my Pid period to 25ms but with more aggresive games like Dirt3 - 30ms gives less motor errors, and still falls under the 33ms that xsim is sending data through at.

Mmmm please look at this pic!:

As you can see my PID period is a damn little 5ms!.. can be this the main problem here? and my PID propotional/integral and derivative settings are accetable? I never touched it from the start of my project to now..

I have similar problems for time to time, only it seems with the bigger motors.
I set intergral at 10000 up from 1000, and 25ms.
I get the odd error reading comport.... , I click ok and it goes again, It happens it seems on BIG motion spikes, I may have to soften it up a bit.

Thx you too bsft, tomorrow will be a full tests day.. ;D

as bsft recommended, plus ya not really using the pid to its full potential one settings for pid i have used is 7/2,0/3,5/1.
Having not real pid setting might be causing issues and sorry never picked it up in any of your earlier post on the jrks.

Another that work nicely from memory was 2/2, 0/3, 1/1.

Yes Aldoz, you have not touched the PID settings, try the generic of 2/2 0/3 1/1. I found p numbers help speed and twitchiness. Work them both at the same time, leave i ,and d controls overshoot of motors (motor driver error). A tricky one, but work the numbers both at the same time up and down, that and you may have to re-work profiles.
My current PID is 4/5 0/3 8/8. If I lower d to say 6/6 or 4/4, by golly does the motion kick hard! Same with p, if I reduce lower number, it gets real nasty. I am still fiddling with PID on mine, every motor is different, so have a play with the numbers and see how you go.
Cheers, David.

Thank you very much dudes! This evening I will do some tests and after I will tell you the situation..
bsft, so did your jrk working without these connection lost problems?