Getting Started

Can I ask why you are manually setting the pot values? the auto tune should get the settings, they may be different for each motor

The DEScribe software has 3 fields it collects from the auto tune in the "position" tab, right now the difference I have from what the image shows is the "soft limits" and "nominal travel" are both manualy set to 2000mV for low and 3200mV for high by me for the travel distance I would like.

If you think I have something off let me know

screen shot of my current settings, my fields show mv not degrees, I have not had to change these settings, when I use the test slider I'm asked if the range is ok.

I'm pretty confident in this statement, if simtools is outputting or can be made to output a carriage return by using <13> in simtools output strings, the code will work with the kangaroo! It's not called "simple serial" for nothing! The kangaroo even ignores if there are spaces simtools adds to the code. Also simple serial with a kangaroo uses text commands instead of raw bytes like the controllers do, so if there are occasional transmission errors, it won't act on them. I'm getting a little more grip on the describe software as I play with it. It's a little tricky to understand at first but it looks great for sims if it works as advertised. People will actually not need to auto tune once we get a chance to understand how to set the figures auto tune normally sets manually. Once we can get a good baseline for a "sim type" from a good auto tune, it should be fairly easy from there.

It's really hard to work on a problem with theory only! So I just need to go pick up a couple of pots so I can hook up my kangaroo. I mean, I bought it the experiment with after all so no better time than the present. I have an electronics store down the street that's open today so I'll go get a couple and start testing! I'll get back too you.

Yes, I would suggest you do what you did for testing at least since you don't have much to work with.

Sorry, I don't know a good free terminal program because I can usually whip up some code of my own for what I need. If I had decided to download simtools during beta I would test it myself but now I will just have to wait. I can test it with the old software once offered here but that's not much good now for the purpose we need .

Looks good to me

Yea mine does show mV now as well that was an old screen when I put "degrees" in the units field.

just used hyper terminal sent the following
1,start<13> 1,units 1000=3145<13> 1,p500 return
and the motor moved to mid position

I'm typing in hyper terminal blind as the roo is not echoing back,
if I then send 1,getp return the roo echos back position value

regards Terry

Great! Can you now test what simtools is outputing?

Nope I dont have sim tools till its released, I'm still sorting out all my hardware before I start a build.

Well then, either simtools is not outputting correctly or StebanJigs has some sort of problem(s) with his setup. However, if eoarobbie is correct when he said you can use <10> in simtools to get a line feed, then <13> should work also to put out a carriage return. At this point I'm thinking StebanJigs has some issues.

BTW telfel, how come your not using the other software with your kangaroo? Doesn't their plugin work?

Edit: Oh yea, you said they didn't home at startup. But have you tired it with a game using output? It doesn't necessary need to home at startup, it will home itself when the game output starts!

I am testing at the mo, yes it does work, I see a couple of posts asking for help, I have some limited experience with the roo.
and had a setup that was working, with the plugin I hadn't got to involved in the serial commands.

I stumbled across the Kangaroo while looking for a 24v high power H bridge, so I'm now using the Kangaroo with the Sabertooth 2x60

Good choice for a controller! Thanks for helping here on this site . Now I don't need to go get those pots to test the code because you have already done it.

The other guys have had a little more time to develop their software, but the nature of the software here sounds a lot better to me. I'm looking forward to the release. And this is just a more helpful and fun atmosphere here. The biggest problem over there is if you accidentally rub certain people the wrong way, your history . Several people here can attest to that! Your entitled to your opinion, as long as you keep it to yourself, otherwise....... You could say I had a short career over there, my first post was my last post. LOL. If you want to stay around over there you better not attempt any "Constructive Criticism" about the software. BTW, it was when this certain person was threatening to shut the development down because he wasn't getting his way, if you get what I mean.

I wish I could be of more help with this, if I actually had any coding experience maybe I could figure this communication issue out. It works as a setup and there is movement, its just figuring out exactly why it responds so slow and doesn't seem to recognize its zero and range of travel properly. Im a little out of my league when it comes to this

Keep the faith my man! It's just a matter of time! If you have a little spare time every now and then, we'll get you going eventually. I need simtools though. I was stupid not to download it when I had the change but I didn't honestly have the time to help with the beta testing.

Just using USBlyzer I can see that when moving axis1 in the test slider there is change in the raw data, with axis2 slider there is no change in output atm.

EDIT: actually again the axis2 output only changes the raw data output in the negative of the slider. I wasn't watching.

EDIT: Ok so I have the units in DEScribe are 1=1 and my outputs in simtools 12bit mode as 1,start<13>1,unit4096=1200<13>2,start<13>2,unit4096=1200<13> I booted up rfactor2 and it works, the only issues seem to be the latency and the range of travel and its recognition of the midpoint and the ability to have an offset if the mechanics are not perfectly set.

EDIT: It seems simtools has the range of motion to control them from -10% to +50% on the sliders this causes max travel in both directions of the motor.

The command "unit" above should be "units". While it might work with "unit", programming languages are usually quite structured and don't like it if it's not perfect. That's why I made the comment in an earlier post that the kangaroo will ignore spaces. An unwanted space can even be an issue executing code. This might be where your range problem stems from though. Good to hear you are getting something out of simtools though!

Here is what simtools is outputing while moving axis1 slider

"
00000000 31 2C 70 32 33 36 35 0D 32 2C 70 32 30 34 37 0D 31 2C 70 32 33 38 30 0D 32 2C 70 32 30 1 , p 2 3 6 5 . 2 , p 2 0 4 7 . 1 , p 2 3 8 0 . 2 , p 2 0
0000001D 34 37 0D 31 2C 70 32 33 38 30 0D 32 2C 70 32 30 34 37 0D 31 2C 70 32 33 39 36 0D 32 2C 4 7 . 1 , p 2 3 8 0 . 2 , p 2 0 4 7 . 1 , p 2 3 9 6 . 2 ,
0000003A 70 32 30 34 37 0D p 2 0 4 7 .


It seems that 1 is indeed getting a command to move on a basis on 4096 steps with axis2 remaining still since I did not move that slider.


It could be viewed as this as well
1 , p 2 3 6 5 . 2 , p 2 0 4 7 . 1 , p 2 3 8 0 . 2 , p 2 0 4 7 . 1 , p 2 3 8 0 . 2 , p 2 0 4 7 . 1 , p 2 3 9 6 . 2 , p 2 0 4 7 .


Movement still has what seems to be a 2-4 second delay from slider change to motor movement.

EDIT: this was the capture of only one sequence "#130" out of 280some different 64byte chunks.

@eaorobbie BTW like I stated above, the resolution is limited to about 12 bits with analog feedback due to a 12 bit ADC. However this limitation is removed if the kangaroo uses encoder feedback.