max supported input voltage on pca9685 bonnet/shield.

Moderators: adafruit_support_bill, adafruit

Forum rules
Talk about Adafruit Raspberry Pi® accessories! Please do not ask for Linux support, this is for Adafruit products only! For Raspberry Pi help please visit: http://www.raspberrypi.org/phpBB3/
User avatar
knoxvilles_Joker
 
Posts: 183
Joined: Wed Mar 01, 2017 9:15 pm

Re: max supported input voltage on pca9685 bonnet/shield.

Post by knoxvilles_Joker »

I found a better option for power distribution

https://www.revrobotics.com/rev-11-1144/

This is FTC/FRC competition legal.

This would allow for a much cleaner setup for splitting out power injection. My other option is a y cable cleverly wired.

User avatar
adafruit_support_bill
 
Posts: 88093
Joined: Sat Feb 07, 2009 10:11 am

Re: max supported input voltage on pca9685 bonnet/shield.

Post by adafruit_support_bill »

Yes, that looks like a very good solution. I have not used those modules personally, but other Rev products I have used have been well engineered.

User avatar
knoxvilles_Joker
 
Posts: 183
Joined: Wed Mar 01, 2017 9:15 pm

Re: max supported input voltage on pca9685 bonnet/shield.

Post by knoxvilles_Joker »

adafruit_support_bill wrote:Yes, that looks like a very good solution. I have not used those modules personally, but other Rev products I have used have been well engineered.
I am having control issues with that board. I think it is specifically designed to work with their stuff as the control signals are not kosher. Running direct things work.

I have removed the second savox unit and replaced it with a dss32235sg servo (35kg-cm)

By the numbers all I have to do is get the savox, and a couple of other servos on a separate power supply and run the other servos off the main board with a beefier power supply.

At this point I am left with the option of running a y splitter servo pigtail assembly, snipping the positive on the wire running to the micro controller and injecting power that way.

User avatar
adafruit_support_bill
 
Posts: 88093
Joined: Sat Feb 07, 2009 10:11 am

Re: max supported input voltage on pca9685 bonnet/shield.

Post by adafruit_support_bill »

I think it is specifically designed to work with their stuff as the control signals are not kosher.
I've used Rev stuff with quite a few standard servos from HiTec, Futaba, Tower etc. without issues. And the description of the power unit says it does not mess with the signals:
The Servo Power Unit passes through input signals unaltered

User avatar
knoxvilles_Joker
 
Posts: 183
Joined: Wed Mar 01, 2017 9:15 pm

Re: max supported input voltage on pca9685 bonnet/shield.

Post by knoxvilles_Joker »

adafruit_support_bill wrote:
I think it is specifically designed to work with their stuff as the control signals are not kosher.
I've used Rev stuff with quite a few standard servos from HiTec, Futaba, Tower etc. without issues. And the description of the power unit says it does not mess with the signals:
The Servo Power Unit passes through input signals unaltered
I will try again after I get some more parts in.

I need to redo the connections and try to shorten things where I can as there could be parasitic losses from both length and excess connections.

I also need to redo the power connection itself as it sux.

I wil also need to fish out my pwm power meter, but I will need to buy a second one for base line comparisons.

https://www.gobilda.com/8-channel-servo-power-node/
This unit linked above appears to be another option. I just have to disconnect the red/positive power line on the microcontroller input line.

I was discussing things with an old school car restoration guy. According to him cars used to run on 6 volts DC. They moved to 12 Volts for brighter lights more power, and more money mark ups.

User avatar
knoxvilles_Joker
 
Posts: 183
Joined: Wed Mar 01, 2017 9:15 pm

Re: max supported input voltage on pca9685 bonnet/shield.

Post by knoxvilles_Joker »

OK I used a bigger gauge cable and that seemed to fix everything.

I just need to bench test things with the batteries and sort some structural issues. (one tilt gear set needs to be redone to match the other arms setup as the teeth are not fully interlocking. Just not happy I will have to canibalize my first animatronic setup to do so.)

User avatar
knoxvilles_Joker
 
Posts: 183
Joined: Wed Mar 01, 2017 9:15 pm

Re: max supported input voltage on pca9685 bonnet/shield.

Post by knoxvilles_Joker »

OK, most of this stuff comes down to physics. at 8 amps for a 3 foot run you really need 14 gauge wire. This is the same gauge used in solar DC power distribution battery connection cables. Powerwerx and anderson power connectors appear to be the standards for higher amperage power connections.

The other common appears to be xt60. racedayquads.com has a whole host of connection options in that arena.

22 gauge cable will not work with this setup so standard barrel connector setups absolutely will not work with the setup and that is what my issue was and is.

User avatar
knoxvilles_Joker
 
Posts: 183
Joined: Wed Mar 01, 2017 9:15 pm

Re: max supported input voltage on pca9685 bonnet/shield.

Post by knoxvilles_Joker »

OK, I finally got the pwm meters from servocity. voltage is constant on the servos giving me issues. But I am not getting a reliable pwm signal or frequency. I think I have traced it back to the board. So I have a couple of replacement boards on the way from digi-key. I am saving up for a massive bigger order with adafruit for other projects.

Only thing I have left to do is to run straight from the board to the servo and see if the issue is in the long runs of servo cabling. I was concerned more with voltage drop on the longer runs but that does not seem to be at issue but I do appear to have some loose connections.

The other concern I had was the fact that the extension cables do not all make good solid connections and are 22awg and really need 22 or 18 awg for longer runs to ensure no loss of signal. The 14awg cable I am using for power redistribution is too big to work in the servo cable assemblies.

User avatar
adafruit_support_bill
 
Posts: 88093
Joined: Sat Feb 07, 2009 10:11 am

Re: max supported input voltage on pca9685 bonnet/shield.

Post by adafruit_support_bill »

really need 22 or 18 awg for longer runs to ensure no loss of signal.
Signal runs are low current, so wire gauge is not a factor. What you need the heavier wire gauge for is power and ground.

User avatar
knoxvilles_Joker
 
Posts: 183
Joined: Wed Mar 01, 2017 9:15 pm

Re: max supported input voltage on pca9685 bonnet/shield.

Post by knoxvilles_Joker »

adafruit_support_bill wrote:
really need 22 or 18 awg for longer runs to ensure no loss of signal.
Signal runs are low current, so wire gauge is not a factor. What you need the heavier wire gauge for is power and ground.
This is also true. 3v or 5v logic for the pwm control wire is so low amperage that it is not a factor with gauge and length.

So the poor readings on frequency and pwm signal could just be due to a poor connection. I will need to verify the signal coming out of the board itself is not an issue. This would indicate either a bad solder connection or a bad board connection which could be possible as the servos I am having issues with were drawing way too much current, but the supply max was 2amps, but that should not have exceeded the 8amp draw limit.

unfortunately 18 gauge servo connectors are not really a market item and the ones out there are all personally fabricated. The bigger push with beefier servos and power supplies is pushing this issue so I think it is a matter of time before pre fabricated cables become a market ready item.

I prefer pre-made connectors as they have a more uniform quality. DIY connectors are prone to issues and crimping issues/errors. Plus when I crimp, I prefer to also solder the wire to the connector. Unfortunately these connectors are so small that this is most likely not an option other than tinning the wires. Plus 18 gauge is as big as you can go and even that may require some shaving on things

User avatar
adafruit_support_bill
 
Posts: 88093
Joined: Sat Feb 07, 2009 10:11 am

Re: max supported input voltage on pca9685 bonnet/shield.

Post by adafruit_support_bill »

So the poor readings on frequency and pwm signal could just be due to a poor connection
That is one possibility. It could also be due to electrical interference from the motors in the servos, ground-bounce or even crosstalk with other servo signals.

Most RC servos use brushed DC motors. These generate lots of 'brush noise' from the commutation. Also servo motor current demand tends to be 'peaky' And since the ground wire serves as both a signal reference and a current return path, you can end up with a less than stable signal reference.

Running the signal wire on a separate shielded wire can minimize external interference such as crosstalk or brush noise. Heavier signal and ground wires can minimize the ground bounce.

User avatar
knoxvilles_Joker
 
Posts: 183
Joined: Wed Mar 01, 2017 9:15 pm

Re: max supported input voltage on pca9685 bonnet/shield.

Post by knoxvilles_Joker »

I think this one might be a combination of issues. odd things is that it is only two servos really giving me fits. Other servos are intermittant but i wiggle a cable and it gets better so that I think is a loose cable or connector.

User avatar
knoxvilles_Joker
 
Posts: 183
Joined: Wed Mar 01, 2017 9:15 pm

Re: max supported input voltage on pca9685 bonnet/shield.

Post by knoxvilles_Joker »

https://www.machinedesign.com/mechanica ... ting-noise

Source cited for reference under fair use.

When conductors are too small, cables typically overheat and fail. Normal operation heats cables because wires carrying current always have some resistance. Overloading, however, causes multiple hot spots that age localized areas of the conductors’ insulation. Several small faults degrade performance until the cable leaks too much current to be usable. Conversely, overly large conductors unnecessarily raise cable capacitance.

Part of my issue could be I may have gotten the cables too hot trying to power things.

I could have also caused other connector issues trying to force too much power through single servos...

The issue I am having is not paticularly unique per se. It is a problem with longer conductor transmission loss.

serial communications are a well settled thing.

three servo cables plus two leds equals 13. so a 15 pin db-15 cable could work. The cable could be one of those custom bundled cables then be broken out on each end for quick disconnects. an ethernet cable could not be of sufficient gauge to carry the needed current either so that is out as well.

I had initially discounted serial cables as the bulky connection would make things more painful to deal with. I may have to revisit things.

User avatar
adafruit_support_bill
 
Posts: 88093
Joined: Sat Feb 07, 2009 10:11 am

Re: max supported input voltage on pca9685 bonnet/shield.

Post by adafruit_support_bill »

Source cited for reference under fair use.
With reference to that article, don't confuse the term "cable" with "conductor". The "hybrid" servo cables described in the article still have dedicated conductor pairs for the signals that are separate from the power conductors. They are just all encased in the same jacket to save space. The signal conductor pairs are typically twisted and or shielded to block interference and (most importantly) use either differential signaling or have a signal reference that is separate from the ground-return for the power.
Part of my issue could be I may have gotten the cables too hot trying to power things.
If your cables are getting warm, then you must have a significant voltage drop over the cable length. That is what causes ground-bounce.
I could have also caused other connector issues trying to force too much power through single servos...
Assuming that the supply voltage is within spec for the servos used, the servo will only draw the power it needs. There is no way for connector to 'force' current into a servo.
The issue I am having is not paticularly unique per se. It is a problem with longer conductor transmission loss.
serial communications are a well settled thing.
Unless your runs are hundreds of feet, that is not the problem you are seeing.

Locked
Forum rules
Talk about Adafruit Raspberry Pi® accessories! Please do not ask for Linux support, this is for Adafruit products only! For Raspberry Pi help please visit: http://www.raspberrypi.org/phpBB3/

Return to “Adafruit Raspberry Pi® accessories”