difference between pololu and "K8055 + Sabertooth x10a"

[ferslash]

I am planning to use one out of this two solutions... 2 polulu cards or "K8055 + Sabertooth + Wiper motors"

talking about performance Is there a big difference that i should know? or they are just the same

I asked motiondave (thanks Dave) for the simplest solution and he told me that by far the simplest one is using pololu, but talking about price the k8055 solution is almost $80 usdlls cheaper. (you can invest your $80 usdlls to improve other parts of your sim)

so... this two solutions at the end gives you the same performance output or no?

best regards

fer from mexico

[tronicgr]

I think safety should be first. Here is a part of discussion I had the other day:

Subject: Blue Sim racing - 2DOF Wiper Motors

Just one question, can you do a test for me? If you have a motor running with velleman and you "accidentally" reset the computer, the motor goes on its rotation?

Thanos

Hello tronicgr

Yes, if you reset the computer accidentally engines run at full speed because x-sim does not control them more and even if you close x-sim by mistake the same thing happens to me

Hello kevin

[ferslash]

thanks tanos, that is all i have to know

if i use pololu i dont have this problem right?

best regards

fer

[Battlestar]

Hi,
Thanos has given a reply as to whether there is danger of runaway motors. There is a solution to that which involves connecting a relay which takes a feed from the digital output 01 on the velleman board. If the program crashes or hangs, the supply to the Sabertooth is cut and cuts off the 12v supply to the motors. They stop immediately. The motors will also stop if the usb cable is pulled out. The instructions for digital output 01 have to be input for each axis, both positive and negative directions.
vicpopo posted the solution a few days ago. I adopted it and it works perfectly. As to which controller is best I'm afraid you'll nedd a reply from someone who has used both and can compare. So far the velleman and Sabretooth seem to be doing the job for me.
Hope this is of some help.
Mike

[ferslash]

thanks battlestar (i will check the solution you mentioned)

[ferslash]

and about performance??? acurancy? life-endurance? any comments?

fer

[tronicgr]

Do you really want to know?

Velleman K8055 positions update rate: 50ms (or just new 20 positions per second) -- (class: arcade)
CKAS T2s 2DOF Motion System positions update rate: 10ms (or 100 new positions per second) -- (class: fairly decent response simulation)
Jrk12v12 positions update rate: 5ms (or 200 new positions per second) -- (class: robotics / cnc / almost realtime simulation)
AMC1.6 positions update rate: 1ms (or 1000 new positions per second) -- (class: realtime simulation)


Thanos

[ferslash]

jajajaj man you rule!!!

so i will completely go for pololu, since i can't construct amc but, now i know that is almost the best option, even if i have to spend a little more $

thanks thanos

best regards

fer

[motiondave]

I'd like to examine something about reaction rates.
Its all good having a card that does 50ms, 10ms, 5ms and 1ms, but what about what the motor does. After all, it generates the motion.
I asked the motor supplier for the big worm gear motors I use, of which are more an industrial design and not the standard thing off ebay.
I asked what sort of reaction times could they get changing back and forth,
He replied with very small movements and under the right conditions, about 50 changes per second which is about 20ms
But that is really stressing the motor and for big movements, the reaction time slows.
I run mine hard, and I am estimating they would get maybe 20-25 changes per second, or 50ms...ish.
However, these motors run at 180rpm at the shaft, so they move fast.
I use to run wipers at 50-60 rpm and 5ms on the jrk.
I now run 140-180 rpm and 25ms on the jrk. I will tell you, the bigger motors provide WAY better motion than the wipers. And quick motion.
1-2ms is good, but we are not running little optical drive motors of hdd's. We are running 20 amps motors at 12v with bigger magnets and armatures.
I think we have to look at motor actual speed in reaction times more as they provide the oomf...
Its all fine and dandy sending 1ms to the motors, but if the mosfets and or the motors cannot actually respond like that, they why the comparison?
Overall, if you have a simulator, ask your self this. Does it move well for you to enjoy it?

[motiondave]

I thought I would ask Pololu if the JRK can run at 1ms , the answer is , yes , as long as we are not trying to take too many analog samples.
Now I am not sure how many analog samples is too many, but basically, the JRK matched Tronics card of 1ms.
But of course, the tronic card can hook up to more than just 12v h-bridges.