>rotation unless I add an encoder
Get 2 slot sensors, one for 110deg east, one for 0deg south
At 6am the motor start from 110deg east, you power the geared-motor for 1sec every 1minute, if you reach 0deg South before noon you make milliSec auto-adjustments and keep track of new value.
You wait (park it) until noon if you reached south before it was time, then you keep moving it west
At 6pm you move the motor back to 110deg east and go to sleep for 12hrs.
Could this just be a matter of the motors and relays pulliing more power than your power supply can comfortably handle, possibly causing a brownout in the Prop circuit?
The prop no longer resets since I have added 2 large capacitors and a diode on the incoming power to the prop. Before that is the connection point where the power will come in.
As stated in past posts, the prop circuit and h-bridges are about 20ft away from the battery bank. The wire will be buried under ground inside PVC pipe. There will be 2 each 4 gauge wire and 2 each 10 gauge wires going through the PVC. The 4 gauge is for the solar power coming back to the battery bank, and the 10 gauge is to power the entire solar tracker system.
I really do not want to rely on time and dates to track the estimated point of the sun. This is why I went simple... LED's used to change frequency signal sent back to the prop. Simple calculations happen, the system moves a little in the direction towards the sun based on the readings from the led light sensor and the average resets and starts over.
The prop no longer resets since I have added 2 large capacitors and a diode on the incoming power to the prop. Before that is the connection point where the power will come in. .
I've always been a BIG believer in OVER stabilizing the controller power. Like you used..large caps & diode protectors.
if you had no protection for the prop from the motors you could've gotten both brown-out and RF. (I know, I'm coming into this thread late and RF was already covered)
I just wanted to say for future projects... give the prop the "cleanest" vcc you can. Even if to the point of having a battery in circuit, and diode switched, to avoid loss by high drain by motors starting up.
I tend to over rate my power systems feeding my brains vs it's requirement. Milspec equipment usually has this and I'm so used to seeing it in schematics..I just do it now.
even if the motors runs slow due to power fade (just a thought, did you voltmeter the supply during operation or better yet oscope it? actually you should proably check it at a point just before the motor controllers..or both.)
[uh...where was I? oh. ] the brain will compensate as long as it's not "confused" too. especially since you are using a "self stabilizing" system.
one other dim memory was just triggered, something about low amp feed to a DC motor causing it to arc too much.
The power wires you ran to motor controllers are you SURE they are rated to power BOTH at the same time?
The main feed to the system, I mean, not the individuals to each one.
I seem to remember doing the same thing to a RC race boat a long long time ago.
(It was 6vdc motor power, I replaced with 7.2 pack, probably china motors,system went nuts, replaced w/heavier 12v dc motors)
there's a thought...since you aren't over worried about speed. perhaps a 10 or 100 ohm wire-wound resistor (or pack depending on wattage) slowing motors and the inductance should help with arcing. Shouldn't slow them down much but will help deal with power issues?
Hi again. As I mentioned in an earlier post, I`m pretty sure about this noise comes from the brushes in the motor, and that it is "air-borne" into your other circuits. My suggestion is to try a "snubber circuit" over the motor terminals. This is simply a 100R resistor in series with a 0,1uF cap. I`ve used this config to get rid of RF noise from contactors and other inductive loads when switching. It helped a lot more than only caps. Take a look at this site: http://coactionos.com/embedded%20design%20tips/2013/10/15/Tips-Motor-Control-using-PWM-and-PID/
Could be worth a try.:thumb:
EDIT * The motors are either on or off. No PWM is used.
He's not using PWM just straight power to and from. So if it's RF it's coming from the brushes like you (EDIT you=we) say . but the RC filter circuit u suggest is more for flyback. WHICH (now that I think of it) COULD be occurring, in a roundabout way. If the motor(s) is/are "staggering" and, the load is not moving smoothly.
Bearing in mind he's got it stable now. I was still wondering if anyone monitoring this thread knows of an RF meter (like maybe a microwave leak sensor) that would be worthwhile for detection?
Comments
Get 2 slot sensors, one for 110deg east, one for 0deg south
At 6am the motor start from 110deg east, you power the geared-motor for 1sec every 1minute, if you reach 0deg South before noon you make milliSec auto-adjustments and keep track of new value.
You wait (park it) until noon if you reached south before it was time, then you keep moving it west
At 6pm you move the motor back to 110deg east and go to sleep for 12hrs.
-Phil
As stated in past posts, the prop circuit and h-bridges are about 20ft away from the battery bank. The wire will be buried under ground inside PVC pipe. There will be 2 each 4 gauge wire and 2 each 10 gauge wires going through the PVC. The 4 gauge is for the solar power coming back to the battery bank, and the 10 gauge is to power the entire solar tracker system.
I really do not want to rely on time and dates to track the estimated point of the sun. This is why I went simple... LED's used to change frequency signal sent back to the prop. Simple calculations happen, the system moves a little in the direction towards the sun based on the readings from the led light sensor and the average resets and starts over.
I've always been a BIG believer in OVER stabilizing the controller power. Like you used..large caps & diode protectors.
if you had no protection for the prop from the motors you could've gotten both brown-out and RF. (I know, I'm coming into this thread late and RF was already covered)
I just wanted to say for future projects... give the prop the "cleanest" vcc you can. Even if to the point of having a battery in circuit, and diode switched, to avoid loss by high drain by motors starting up.
I tend to over rate my power systems feeding my brains vs it's requirement. Milspec equipment usually has this and I'm so used to seeing it in schematics..I just do it now.
even if the motors runs slow due to power fade (just a thought, did you voltmeter the supply during operation or better yet oscope it? actually you should proably check it at a point just before the motor controllers..or both.)
[uh...where was I? oh. ] the brain will compensate as long as it's not "confused" too. especially since you are using a "self stabilizing" system.
one other dim memory was just triggered, something about low amp feed to a DC motor causing it to arc too much.
The power wires you ran to motor controllers are you SURE they are rated to power BOTH at the same time?
The main feed to the system, I mean, not the individuals to each one.
I seem to remember doing the same thing to a RC race boat a long long time ago.
(It was 6vdc motor power, I replaced with 7.2 pack, probably china motors,system went nuts, replaced w/heavier 12v dc motors)
there's a thought...since you aren't over worried about speed. perhaps a 10 or 100 ohm wire-wound resistor (or pack depending on wattage) slowing motors and the inductance should help with arcing. Shouldn't slow them down much but will help deal with power issues?
Just a couple of ramblings (hopefully helpful).
Could be worth a try.:thumb:
He's not using PWM just straight power to and from. So if it's RF it's coming from the brushes like you (EDIT you=we) say . but the RC filter circuit u suggest is more for flyback. WHICH (now that I think of it) COULD be occurring, in a roundabout way. If the motor(s) is/are "staggering" and, the load is not moving smoothly.
Bearing in mind he's got it stable now. I was still wondering if anyone monitoring this thread knows of an RF meter (like maybe a microwave leak sensor) that would be worthwhile for detection?