When I place my Adapt-a-damper into manual fan mode and set it at 0% the servo puts out a high frequency noise (whine). When the damper is fully open and the fan is at 100% there is no such noise. I do get whining when the damper is partially closed/open. Because the damper would not be fully closed at 0% nor fully open at 100%, I modified the servo control on the web page to be: Servo pulse duration closed 930 us, 1800 us, fully open at 100%. Any ideas on how to eliminate the servo whining noise
High Frequency Whine From Servo When Damper Fully and Partially Closed
- Thread starter Mache
- Start date
Bryan Mayland
TVWBB Hall of Fame
Usually that means that there is some binding happening. You can tell if this is the case if you can give the dial a little manual jiggle to to get it to stop. If that's happening, then you can loosen up the dial a bit, or polish the mating surfaces with a soft cloth for a minute or so to remove any sticky residue from the 3D printing process.
Earlier whining servo burned out. Luckily had a replacement lying around. Finally got everything working. New servo seems to click/readjust every 5 seconds or so regardless of whether open or closed. Given my experience with the thermocouple, I think the Raspberry Pi 3B+ generates a lot of RF noise.
Bryan Mayland
TVWBB Hall of Fame
Well that sucks that the servo died so quickly, considering I'm using the SG92R from Adafruit which should be authentic and cost me $6 each after shipping, compared to the <$2 SG90 I was using before. I had a few of the MG90S die on me personally pretty quickly and found the SG90 to be more reliable, but the DOA percentage to be somewhat high so I started testing them all with a servo sweep test before they went out. I figured the Adafruits would be the most reliable of the bunch.
As far as the twitching, have you tried putting a capacitor on the 5V/GND line on the servo end? That doesn't exactly sound like it would fix it, since that's the fix for where the servo just goes nuts jumping to random positions and back to where it should be, but might be worth a try? I found that even a 0.1uF capacitor fixed the test setup I had that exhibited that problem, but I'd recommend maybe 100uF if you've got it.
As far as the twitching, have you tried putting a capacitor on the 5V/GND line on the servo end? That doesn't exactly sound like it would fix it, since that's the fix for where the servo just goes nuts jumping to random positions and back to where it should be, but might be worth a try? I found that even a 0.1uF capacitor fixed the test setup I had that exhibited that problem, but I'd recommend maybe 100uF if you've got it.
Cooked some chicken on an 18.5" WSM with the new HeaterMeter 4.3 with a RPi 3B+ for a 4th of July BBQ. The four hour cook was able to keep the pit on its setpoint but the MG90S servo was constantly twitching and did not just adjust and stay quiet. I used 5Ghz WiFi.
I plan to purchase a 100uF electrolytic capacitor on Tuesday at a local electronics store and insert across the 5V/GND servo line in the keystone jack of the Adapt-a-damper. What voltage should I get? Will 10 volt work? Is there a better type of capacitor other than electrolytic I should purchase?
I plan to purchase a 100uF electrolytic capacitor on Tuesday at a local electronics store and insert across the 5V/GND servo line in the keystone jack of the Adapt-a-damper. What voltage should I get? Will 10 volt work? Is there a better type of capacitor other than electrolytic I should purchase?
I added in a 100uF 10V capacitor bridging 5V Servo Power and GND. Still have the problem. I have recorded a video that shows what is going on and placed on Google Drive - https://drive.google.com/file/d/1bZ7wmDGr-AYcVNDEDc-1zm_DXlEX93hX/view?usp=sharing. Anyone have an idea on how to resolve this?
MG90S servo's are notorious for being pirated/faked. You probably do not have the genuine article. I would suggest the MG90 servo sold by Vorpal Robotics. It's made by Tower Pro, and I have never had any issues with them. Same form factor as the MG90S. Great servo and reliable.
Vorpal Brand MG90 Micro Servo (Made by Tower Pro)
We worked directly with Tower Pro to create the Vorpal MG90 servo. You get genuine Tower Pro quality in a servo optimized for our robotics projects. We also test every servo ourselves and reject defective servos (about 2%), so what you're getting is higher quality than even what comes from the...
vorpal-robotics-store.myshopify.com
Bryan Mayland
TVWBB Hall of Fame
I don't think any size capacitor is going to make that any better. It looks more like a function of the tiny servo unable to make precise movements of such a small distance.
The servos should be authentic. They come from Adafruit and have the Tower Pro / TP markings on the case. If they're not authentic then I'm going to be pretty upset about them charging $6/ea for them.
The servos should be authentic. They come from Adafruit and have the Tower Pro / TP markings on the case. If they're not authentic then I'm going to be pretty upset about them charging $6/ea for them.
I think you are right. I went to a friend's house this morning and we put the 5V and control servo lines on an oscilloscope. While there are minor noise issues on those lines, the signals to the servo appear well defined and clean. We also added in an additional capacitor bridged across the 5V and GND with no change in servo twitching.
This removes the Heatermeter PCB and Raspberry PI 3B+ as sources of servo twitching and leaves only the servo itself as the root cause of the problem. Per WBegg's suggestion above, I purchased two Vorpal MG90 servos. They should arrive in 7-10 days. They are pricey but I am hoping they will eliminate this last issue.
This removes the Heatermeter PCB and Raspberry PI 3B+ as sources of servo twitching and leaves only the servo itself as the root cause of the problem. Per WBegg's suggestion above, I purchased two Vorpal MG90 servos. They should arrive in 7-10 days. They are pricey but I am hoping they will eliminate this last issue.