What about increasing the PWM voltage on the stock Cube ?

VCD

Joined
Nov 7, 2016
Messages
165
Reaction score
47
Age
44
I guess the stock Solo cube uses the same buffers/level converters as Green Cube / Pixhawk etc, so it should be a matter of cutting some traces and applying 5v to their external side, to make PWM go 5v - thus solving the worst problem.

Does the community have the stock Solo's schematics/PCB layouts ? - it would be a fun micro-soldering exercise to fix them.
 
  • Like
Reactions: Wetstone
What's your basis for that belief? Do you have a schematic somewhere? I was under the impression the problem was that there weren't level converters in the stock Pixhawk cube.
 
There is nothing you can do in/on the stock Pixhawk 2.0 to bring the PWM signal level up to 5 volts. The new Pixhawk 2.1 has an internal jumper to select 3 vs 5 volts. 3v is the norm, but it has a 5v option, mainly for use in a Solo for this reason. The Pixhawk 2.1 green cube comes from the factory with it pre-set at 5 volts.

The only way you're going to get 5v signalling to the motor pods without replacing the stock cube is to install logic level converters between the carrier board the motor pods. Nobody has done this YET, but there are a few people in the mod club and beta group expressing interest in doing it.
 
There is nothing you can do in/on the stock Pixhawk 2.0 to bring the PWM signal level up to 5 volts.
Does this mean you know there are no level converters , (unlikely as I assume they based it very much on Pixhawk) or does it mean you have seen the pcb drawings and k ow there is no way to access and cut traces due to placement or vias ?
Or just assume that?
Using external level converters is simple and not even a challenge, but adds unwanted points of failure as that will require more soldering points.
 
  • Like
Reactions: Wetstone
There is no logic level converter in the stock pixhawk 2.0 that you can modify to make it do 5v per the guy who built it. You would have to do it externally. Plug making it something you plug into the port on the carrier and then plug the motorpod wires into would be best.
 
Plug making it something you plug into the port on the carrier and then plug the motorpod wires into would be best.
There is only so much clearance off the main board to either the tray (rear) or other components (front). Locating off the motor pod would allow for more clearance within the arms, maybe...
 
I guess the question now would be cost, considering that the PHGC2.1 already has this upgraded....$10-20 for a set?
 
To run ArduCopter 3.5 without installing the Pixhawk 2.1 Green Cube. ArduCopter 3.5 does not have the annoying slew rate protection for the motor pod design flaw that the stock firmware has. The green cube has 5v signalling out of the box.
 
This issue has confused me for some time. When I asked [email protected] whether I could use Pixhawk 2.0 in a ProfiCNC carrier board for DIY use they said it was not recommended with this explanation: "3DR Solo Pixhawk 2 has a boosted 5v pwm output, which is different from the pixhawk2.1 which provides industrial standard of 3.3v pwm output." I suspect perhaps the person responding misunderstood my question.

So I bought the carrier anyway, put the Pixhawk 2.0 from a Solo on it and measured the PWM. The logic level? 3 volts. By the way, this thing flys great using APM 3.4.6.

I keep seeing that Solo requires 5 volts. Is it correct to say that the stock Solo has 3 volt logic, but it should have 5 volts and that will allow the slew rate to be increased?

Also, on March 1 Pedals2Paddles said "No you cannot buy a plain old Pixhawk 2.1 cube and put it on the Solo." If the 2.1 has a jumper to switch from 3 to 5 volts, why can't it be used or is that jumper a more recent addition? I don't understand why there is a separate green cube with 5 volt output if the regular 2.1 can output 5 volt PWM.

Please don't think I am challenging anyone. I respect you all. Just hoping to get clarity on this.
 

Attachments

  • 2017-07-22 13.46.04.jpg
    2017-07-22 13.46.04.jpg
    1.7 MB · Views: 62
Yes, hex was confused. Both the 2.0 and all varieties of the 2.1 will function in the 2.1 carrier.

The stock pixhawk 2.0 is 3 volt signaling. The slew rate bullshit in the firmware motor matrix code compensates (mostly) for the electrical flaw in the motor pods.

A standard Pixhawk 2.1 (black) is also 3 volt. It has an internal solder jumper which you can change to make it 5 volt. If you do that, it will be fine in the solo.

A Pixhawk 2.1 Green Cube simply comes from the factory preset for 5 volts and colored green. And with some sorbothane for vibration as an added bonus.

The 5 volt signaling compensates for the electrical flaw in the motor pods so you no longer need the slew rate limiting hack in the firmware. ArduCopter 3.5 doesn't and probably won't have that slew rate crap. So 5 volt signaling is required to not fall out of the sky.
 
Would it be possible to replace the ground wire between the motherboard and motor pod with a thicker wire, reducing the voltage drop during rapid motor speed changes, allowing the stock Cube to safely work without slew limiting?
 
That mod (madhacker) does work, but it is not for the faint of heart. Although pro SMD rework folk can do it in a flash, unless you know one or are one it is not going to be an easy mod. I have done 3 PixHawk 2.0's, the first was hard, the second was a lot easier, the third was a bi$%#tch. Don't try this at home. While it can be done and provides the same (mostly) effect as moving to a Green Cube, I would strongly suggest acknowledging it as cool but GO OUT AND GET A GREEN CUBE!
I may do #4 but only after another month of recovery ;>))
 
I do not know exactly what is this settings reposible for so I can only guess but I think this is not so simple, as fare I understand it is relate to that ESC can miss pwm logical high state at higher frequency (when motor should spin fast). It is caused by small tolerance of voltage, ESC see high logic signal from approx 3V-5V and cube is able to generate max 3.3V. at high freq high logical signal generated by the cube does not reach min border of the logical ESC high signal. So lowering motors courent does not solve this issue. Pls correct me if I'm wrong
 

Members online

No members online now.

Forum statistics

Threads
13,093
Messages
147,741
Members
16,047
Latest member
pvt solo