|
|
Post by riton39 on Jan 7, 2015 8:10:45 GMT
I restarted my PC, the first time it works, but only one time.
I boarded at 254 for P, no oscillation there is a problem, and the app is stopped which is why it works only one time?
|
|
|
|
Post by riton39 on Jan 7, 2015 8:40:20 GMT
I need to disconnect and reconnect the USB AMC
Then I re start the app.
then only one time I can do autotuning.
but there is a problem, increase greatly, the oscillation does not always do this.
I had 2 times violent chutte of the cam.
the cam then resumes its position, and there is no oscillation.
I also had a case where I have a swing, I click stop, but the oscillation stops.
and I can not go back and continue to increase P.
I think we need to continue in case it happens.
edit: I tried starting with P = 45 on my video tests 5.
I also do a test starting with P = 0, the cam does not go all the, I went up to P = 100.
every time I need to reset the AMC USB on/off/ON if it does not validate the position.
edit: |
|
|
|
Post by fabi on Jan 7, 2015 22:28:02 GMT
Can you explain "Tick dashed" "Highlight actor position axis" and "Hide actor position axis"? thanks.
By default, the "lines" on the ocilloscope are one pixel. You can increase it if you like stronger (thicker) lines. If you click "highlight" it makes the "line" from the actor you've selected thicker than the the lines from the other actors. If you wanna watch one specific actor (e.g. you have selected 6DOF, but you only wanna watch actor 1) you can "hide" the lines from the other actors on the ocilloscope. You can take only one ocillation, if you take more, your measure becomes more exact. Approximate is ok. You can also "play" with the value and see how I and D are changing, it is very small. |
|
|
|
Post by riton39 on Jan 7, 2015 23:11:22 GMT
ok thanks.
I tested quickly the 1.0.04, I have the same problem in my earlier video.
I still have the bug position detection.
I have to disconnect and reconnect the USB
when I increment the P, there is no movement up to 254.
there appears to be no order.
it looks like the software is planted ...?
|
|
|
|
Post by tronicgr on Jan 8, 2015 6:18:07 GMT
ok thanks. I tested quickly the 1.0.04, I have the same problem in my earlier video. I still have the bug position detection. I have to disconnect and reconnect the USB when I increment the P, there is no movement up to 254. there appears to be no order. it looks like the software is planted ...? , I think his problem is related to the oscilloscope being freezing if any sensor value that is returned from the AMC1280USB is zero (It stops scrolling on my side here as well) I think you should allow the sensor values to be zero, in case there are no connected sensors on all ports. If some of the ports left unconnected (floating) they may pull-down naturally to the ground, as result of the internal pull-down architecture. |
|
|
|
Post by vicpopo on Jan 8, 2015 6:33:16 GMT
Hi Fabi,
Tested yesterday late in evening .... no ,no today early moring the 1.0.0.4 .I wrote a post but as it was late I probably forgotten to create it because I noticed that it wasn't published.
My notices :
- the evaluate process launched it but I just watched the green bar graph progressing horizontaly 1 or 2 cm and then after windows closed.Before I was able watching the complete bar graph became green.It's maybe just a display " bug " ( or not ) because I was able to make the auto tune process.
-auto tune process : yes , yes and yes for the first time I achieved getting calculation and I , D values.no more XXX !! And I found why !
Because the value Pu returned from the oscilloscope ( after counting oscillations) to the auto-tune window was not in the right format to calculate .It returned 0.065 and I re-type to 0,065 and it works . I changed the dot to a decimal virgule and that.s it.I don't know if that comes from my laptop setup ( Azerty to qwerty ) nut now it's ok ;-).
My originaly PID figured out a couple of weeks ago from tests was 30 12 24 and the one made with auto tune is tatata 30 12 0 !!
You are a genius Fabi ( Thanos also ;-) ) and i'm a genius and a lucky guy!
Thanks !
|
|
|
|
Post by tronicgr on Jan 8, 2015 6:49:51 GMT
fabi, Good point about the Decimal character... I wasn't sure when I was doing the calculations correctly because not sure if that was a "." or a "," resulting in reading 1.627 as 1627 ...?!! Perhaps the windows decimal system is confused by this as well! Thanks vicpopo
|
|
|
|
Post by fabi on Jan 8, 2015 6:52:45 GMT
ok thanks. I tested quickly the 1.0.04, I have the same problem in my earlier video. I still have the bug position detection. I have to disconnect and reconnect the USB when I increment the P, there is no movement up to 254. there appears to be no order. it looks like the software is planted ...? , I think his problem is related to the oscilloscope being freezing if any sensor value that is returned from the AMC1280USB is zero (It stops scrolling on my side here as well) I think you should allow the sensor values to be zero, in case there are no connected sensors on all ports. If some of the ports left unconnected (floating) they may pull-down naturally to the ground, as result of the internal pull-down architecture. Hello Thanos so is my understanding correct: Allowed values to the Thanos board are "1 - 65535" Valid values from the Thanos board are "0 - 65535" Thx Fabi |
|
|
|
Post by fabi on Jan 8, 2015 6:54:53 GMT
fabi, Good point about the Decimal character... I wasn't sure when I was doing the calculations correctly because not sure if that was a "." or a "," resulting in reading 1.627 as 1627 ...?!! Perhaps the windows decimal system is confused by this as well! Thanks vicpopo Thanks both! I will look into it later today. |
|
|
|
Post by tronicgr on Jan 8, 2015 7:00:50 GMT
fabi, I think his problem is related to the oscilloscope being freezing if any sensor value that is returned from the AMC1280USB is zero (It stops scrolling on my side here as well) I think you should allow the sensor values to be zero, in case there are no connected sensors on all ports. If some of the ports left unconnected (floating) they may pull-down naturally to the ground, as result of the internal pull-down architecture. Hello Thanos so is my understanding correct: Allowed values to the Thanos board are "1 - 65535" Valid values from the Thanos board are "0 - 65535" Thx Fabi Correct!! I limit the input not to be zero to as a fix to an old bug with Ian's software some months ago. Perhaps its fixed now, but still good to have in place. So yes, allowed values to the Thanos board are "1 - 65535" The outputs values from the Thanos board can be "0 - 65535" no restriction there. Thanks!! Thanos |
|
|
|
Post by fabi on Jan 8, 2015 8:41:58 GMT
Hello General info: - Once I have a new release, I'm updating the first entry from this thread (below the video I'm adding the changes I've did) - Also sending a message in Shoutbox I've removed now the restriction from Zero values (have ignored them previous to have send and receive consistent)... but yet  it can break heave evaluation because when zero arrives, nothing arrives for the heave evaluation I've changes the values from millisecond to second (it was display only). I've entered in the text box both: "," and "." and it works both times same.... let me know if you encounter a different behavior. Thx |
|
|
|
Post by vicpopo on Jan 8, 2015 11:56:59 GMT
Thx Fabi for your incredible reacitvity and quick new release.
Will test this evening normally.
|
|
|
|
Post by fabi on Jan 8, 2015 15:17:45 GMT
Checkout 1.0.0.7, added logging capability
|
|
|
|
Post by riton39 on Jan 8, 2015 16:07:35 GMT
Fabi !! you are warm !! Wow, I can not follow !! I'll test the 1.0.07 ... thanks |
|
|
|
Post by riton39 on Jan 8, 2015 16:24:36 GMT
home nothing works ...
click on the validate, valid, then I said 65535 and 1
if I test low and high, engine cam does not move at all ...
it seems that there is no control.
if you need, I can film?
oops, sorry I made a mistake in the test was P 0, then any movement.
I indicated P = 10 for manual testing.
but ...
I click "Evaluate heave position", it works. with 1 .... 65535
Then I tested up, down, Works.
then I go to Auto Tuning.
I have to redo the position, and there, Idem the bargrapher increased by 1 cm and stop ...
|
|
it can break heave evaluation because when zero arrives, nothing arrives for the heave evaluation