Lights flicker when controlled by Reef-pi

jakeb

Active Member
View Badges
Joined
Feb 13, 2021
Messages
106
Reaction score
75
Location
United Kingdom
Rating - 0%
0   0   0
I've been setting up a Reef-pi (I bought a Robo-tank as I don't trust my soldering skills). I added a Kessil A160 today, and the light flickers - more so at lower intensities.

The ports are set to 10v PWM and the slider to change the spectrum and intensity in Reef-pi seems to change the voltage correctly. I've tried changing the PWM frequencies, but it doesn't make a difference. The light doesn't flicker at all when it isn't connected to the Reef-pi.

Does anyone know what might be causing this?

 

takitaj

Well-Known Member
View Badges
Joined
Jul 23, 2020
Messages
516
Reaction score
622
Location
Maine
Rating - 0%
0   0   0
0-10v and PWM are different things. 0-10v is 0-10 volts DC steady. PWM is Pulse Width Modulation and it modulates the on time of a 10 volt square wave. Longer on time equals brighter, shorter = dimmer. If you have it set to PWM then it should flicker if you're trying to control something that needs a steady 0-10v DC because it's turning it on and off at the frequency of the PWM wave. Here's a link to Wiki Pulse-width modulation - Wikipedia
 
OP
OP
J

jakeb

Active Member
View Badges
Joined
Feb 13, 2021
Messages
106
Reaction score
75
Location
United Kingdom
Rating - 0%
0   0   0
Thank you, that makes sense. So I think I understand that my multimeter was reading the average voltage of the waves, and not differing voltages.

I've changed the signal type to 10v analogue and the light still flickers.
 

takitaj

Well-Known Member
View Badges
Joined
Jul 23, 2020
Messages
516
Reaction score
622
Location
Maine
Rating - 0%
0   0   0
10v analog is what you want. I'm not familiar with ReefPi but If you measure the output with your meter set on AC it should read zero. If not then you're still getting a PWM output. PWM will give some sort of positive value.
 
OP
OP
J

jakeb

Active Member
View Badges
Joined
Feb 13, 2021
Messages
106
Reaction score
75
Location
United Kingdom
Rating - 0%
0   0   0
Thank you both for your help with this. I've checked and I am getting 0-10v when testing DC, and am getting a reading when testing AC.

The switches are set to 10v analogue (top is off, bottom is on).

IMG_20210424_090844_2 - Copy.jpg
 

robsworld78

Well-Known Member
View Badges
Joined
Feb 14, 2020
Messages
952
Reaction score
1,280
Location
Edmonton, Canada
Rating - 0%
0   0   0
Thank you both for your help with this. I've checked and I am getting 0-10v when testing DC, and am getting a reading when testing AC.

The switches are set to 10v analogue (top is off, bottom is on).

IMG_20210424_090844_2 - Copy.jpg
Ok I think I know the problem. A couple months ago someone told me the resistors for the 0-10v filter should be higher value so I'm thinking that's the problem. I have heard people say they had a flicker but after turning up the PCA9685 frequency it disappeared so until a month ago I've been telling people to do that.

I just changed the resistor to what you have and now I do get 1.5v when I measure AC voltage with slider at 50%. After changing frequency to 1500 it dropped to .1v so that likely explains why higher frequency helps. After you change the frequency go to Configuration tab then Admin tab and press the reload button so the change takes effect and should be gone.
 
OP
OP
J

jakeb

Active Member
View Badges
Joined
Feb 13, 2021
Messages
106
Reaction score
75
Location
United Kingdom
Rating - 0%
0   0   0
Ok I think I know the problem. A couple months ago someone told me the resistors for the 0-10v filter should be higher value so I'm thinking that's the problem. I have heard people say they had a flicker but after turning up the PCA9685 frequency it disappeared so until a month ago I've been telling people to do that.

I just changed the resistor to what you have and now I do get 1.5v when I measure AC voltage with slider at 50%. After changing frequency to 1500 it dropped to .1v so that likely explains why higher frequency helps. After you change the frequency go to Configuration tab then Admin tab and press the reload button so the change takes effect and should be gone.
That looks like it's fixed it. Thank you both for you help with this, I really appreciate it.
 

Ranjib

7500 Club Member
View Badges
Joined
Apr 16, 2016
Messages
9,825
Reaction score
17,041
Location
Pleasant Hill, Concord
Rating - 0%
0   0   0
10v pwm can directly control kessil, without any alterations (making it look like sine wave). The flickering occurs during low frequency or below certain duty cycle. the minimum threshold for lights were designed to counter that. in my case intensity <13 causes kessil to flicker irrespective of the pwm frequency. I have verified this with other setups, kessil does not allow dimming till 0 (at least the a160,a80,a360 versions)
 

robsworld78

Well-Known Member
View Badges
Joined
Feb 14, 2020
Messages
952
Reaction score
1,280
Location
Edmonton, Canada
Rating - 0%
0   0   0
10v pwm can directly control kessil, without any alterations (making it look like sine wave). The flickering occurs during low frequency or below certain duty cycle. the minimum threshold for lights were designed to counter that. in my case intensity <13 causes kessil to flicker irrespective of the pwm frequency. I have verified this with other setups, kessil does not allow dimming till 0 (at least the a160,a80,a360 versions)
Oops, I've been thinking Kessil was 0-10v analog. :D
 
OP
OP
J

jakeb

Active Member
View Badges
Joined
Feb 13, 2021
Messages
106
Reaction score
75
Location
United Kingdom
Rating - 0%
0   0   0
10v pwm can directly control kessil, without any alterations (making it look like sine wave). The flickering occurs during low frequency or below certain duty cycle. the minimum threshold for lights were designed to counter that. in my case intensity <13 causes kessil to flicker irrespective of the pwm frequency. I have verified this with other setups, kessil does not allow dimming till 0 (at least the a160,a80,a360 versions)
Both PWM and analogue were flickering for me initially (including >13), but now I have changed the frequency to 1500 it doesn't flicker for PWM or analogue. When I originally increased the frequency (before this thread), I was only increasing to around 400.

I'm still waiting for my tank to be delivered, so I ran a 2 hour diurnal profile with no flickering on my box on my equipment...
 

oreo54

5000 Club Member
View Badges
Joined
Sep 18, 2017
Messages
5,531
Reaction score
3,411
Rating - 0%
0   0   0
Oops, I've been thinking Kessil was 0-10v analog. :D
Could sort of be both.
Many meanwell drivers are 3 way.
10v analog, 10v pwm, or "pot".

I've always assumed the 10 pwm was err averaged or smoothed by the circuitry (capacitors or rc circuit ).
Driver design dependent if course
Kessils by definition are 0(1) -10 volt analog.

Changing pwm frequency may just allow the circuity to average voltage more consistently .
Thus the odd behavior.
Using averaged voltage via pwm apparently has it's issues .sometime

Just thinking...
 

robsworld78

Well-Known Member
View Badges
Joined
Feb 14, 2020
Messages
952
Reaction score
1,280
Location
Edmonton, Canada
Rating - 0%
0   0   0
Could sort of be both.
Many meanwell drivers are 3 way.
10v analog, 10v pwm, or "pot".

I've always assumed the 10 pwm was err averaged or smoothed by the circuitry (capacitors or rc circuit ).
Driver design dependent if course
Kessils by definition are 0(1) -10 volt analog.

Changing pwm frequency may just allow the circuity to average voltage more consistently .
Thus the odd behavior.
Using averaged voltage via pwm apparently has it's issues .sometime

Just thinking...
Yeah good point, maybe Kessil is ok with both, I think some people are running analog so must be. I use the LM358 and some caps to filter so yeah PWM signal is smoothed out in my case.
 

oreo54

5000 Club Member
View Badges
Joined
Sep 18, 2017
Messages
5,531
Reaction score
3,411
Rating - 0%
0   0   0
LM
Yeah good point, maybe Kessil is ok with both, I think some people are running analog so must be. I use the LM358 and some caps to filter so yeah PWM signal is smoothed out in my case.
LM385 still outputs a PWM signal or, at best a sawtooth signal only now at a higher voltage. Is that correct?


Not going to pretend I understand this stuff except from early experience of playing w/ a Typhon controller (had 5v and 10V PWM outputs) and trying to force, without a secondary power supply, a 10V analog signal out of it.
Finally got a-hold of voltage doubler chips to apply to the 5V PWM since the 10V PWM was err not compatible in any way w/ the Meanwell drivers I had. Long story.
That along w/ a tailored RC circuit (PWM frequency I believe was 490Hz-ish or double that one of the 2)

Best I got was a pretty clean 10V analog out except that I now ran out of current. Double the voltage 1/2 the current I believe.
IC used was this:
https://www.renesas.com/in/en/document/dst/icl7660s-icl7660a-datasheet
2 1n5817 Schottkey diodes and one cap.
Out of 5V PWM got 9.5V (or close) but again low current.
.64V was the expected loss 5 to 10-.64

Current was fixed by the design of the Typhon.
Soo time to abandon 10V anything.. :)

Some stuff on filtering the output:

Only posted this because I found the voltage doubler chips ..err..fascinating.
You can "flip" the ground as well.

Hmm got to think about that a bit to "reverse" the output of a TC-420 to run channels on split light fixtures that use negative ground instead of the positive ground for designed dimmable lights.
But that would be more for fw applications.

sorry this is a ramble no more no less..

Oh yea and 10v analog isn't well designed to "dim to zero" so funny things happen. ;)
 

robsworld78

Well-Known Member
View Badges
Joined
Feb 14, 2020
Messages
952
Reaction score
1,280
Location
Edmonton, Canada
Rating - 0%
0   0   0
LM

LM385 still outputs a PWM signal or, at best a sawtooth signal only now at a higher voltage. Is that correct?


Not going to pretend I understand this stuff except from early experience of playing w/ a Typhon controller (had 5v and 10V PWM outputs) and trying to force, without a secondary power supply, a 10V analog signal out of it.
Finally got a-hold of voltage doubler chips to apply to the 5V PWM since the 10V PWM was err not compatible in any way w/ the Meanwell drivers I had. Long story.
That along w/ a tailored RC circuit (PWM frequency I believe was 490Hz-ish or double that one of the 2)

Best I got was a pretty clean 10V analog out except that I now ran out of current. Double the voltage 1/2 the current I believe.
IC used was this:
https://www.renesas.com/in/en/document/dst/icl7660s-icl7660a-datasheet
2 1n5817 Schottkey diodes and one cap.
Out of 5V PWM got 9.5V (or close) but again low current.
.64V was the expected loss 5 to 10-.64

Current was fixed by the design of the Typhon.
Soo time to abandon 10V anything.. :)

Some stuff on filtering the output:

Only posted this because I found the voltage doubler chips ..err..fascinating.
You can "flip" the ground as well.

Hmm got to think about that a bit to "reverse" the output of a TC-420 to run channels on split light fixtures that use negative ground instead of the positive ground for designed dimmable lights.
But that would be more for fw applications.

sorry this is a ramble no more no less..

Oh yea and 10v analog isn't well designed to "dim to zero" so funny things happen. ;)
Yeah the LM358 is to increase the voltage to 10v, a PWM signal can come out. Then it's just a cap and resistor to average it. This is where I got the idea.


I added a couple jumpers, one for the cap so analog can be turned on/off and another so it can be 5v or 10v. The resistor inline on input makes a big difference along with the PWM frequency, with 4.7k and higher frequency the output is smoother, if the frequency is low the signal has lots of spikes, basically a PWM signal pattern but triangles. With a 10k resistor it is smoother but the frequency has much less effect so my thinking is 4.7k is more adjustable.

A lot of this I don't completely understand either, it's all quite amazing. Thankfully there's a lot of helpful people out there willing to help nobs like me. :)
 

Attachments

  • DAC.PNG
    DAC.PNG
    20.4 KB · Views: 50
Back
Top