APM 2.x Power management and Distribution system

Hi, All
I want to test the water and learn how much support is out there for a power management and distribution system for the APM 2.x. The APM 2.x is now a well understood and reliable FC. Maybe now it is time to design a stable and reliable power system for it?

There are far too many threads and comments in the forums regarding power issues. In most situations the APM is relying upon ESC bec's that were not designed for a UAV FC and there is much trial and error patching together bits and pieces of varying quality from different sources. None of the open source UAV projects has built an intelligent power management system for our UAVs yet.

I spent the last 2 days thinking about what in my opinion our APM 2.x's need. My idea is explained below. But, I want to hear your idea's and thoughts. That's the aim of this post.

My idea is a pcb that sits underneath APM 2.x. The signal pins A0-A11, signal pins for the PWM outputs and one of the UART ports from the APM are extend down into it and attach it to the APM 2.x. Headers on the PCB to attach devices to A0-A11 and Servo/PWM outputs. The pcb has 3 switching power supplies. One for the APM, one to power the PWM outputs and one to power FPV equipment or accessories. A AT2560MEGA processor to control routing of power, and monitor voltage and current. External pcb's that attach to each battery route power to the main pcb, monitor voltage and current, and can isolate the battery.

APM 2.x Power management and Distribution system (apm pmds) pcb

- ATMEGA2560 processor
- Monitor various voltages and currents
- control routing of power between batteries and power supplies
- isolate batteries and servo/PWM outputs
- isolate batteries and power supplies
- Communicate with APM 2.x
- Develop power management strategies

3 switching power supplies

1) apm ,pmds, sensors, rc receiver & rf modem
- 2 amp
- current sensor
- voltage sensor
- able to independently switch between batteries

2) A0-A11 and servo/PWM output
- 5 amp
- current sensor
- voltage sensor
- able to independently switch between batteries
- able to adjust power output for HV servos
- Opto-Isolators on the servo/PWM signal outputs
- Isolate the power via physical switch
- support for optional external capacitor array

3) fpv transmitter and camera or accessories
- 2 amp
- current sensor
- voltage sensor
- able to independently switch between batteries
- able to shutdown fpv transmitter/camera/accessories in a low power situations

Battery controller pcb's

- Isolate battery
- current sensor
- voltage sensor
- Monitor voltage in up to 4 cells

Hi, Guys

Chris - That's great to learn. A guaranteed compatible power source. Is it backwards compatible to APM 1 & 2.0?
Peter Chaffe - mostly it is about having a known good power source.
Chris/Peter Meister - I think a large minority of people are using two or more batteries or would like an option too. To increase flight times or because of space restrictions in their airframes, or to run one of the electronic systems from a separate power source (e.g FPV equipment). I think a system with some form redundancy or isolation would be welcomed. A small backup battery in case of unexpected exhaustion of primary battery, or Isolating batteries in the event of a crash, or isolating and jettisoning a battery.
Matthew - There would be less benefit for a mulitcopter than a plane or rover. The complexity kind of arose once a micro processor was added. It has the abilities, so add the circuits and see what the community can do with it.
John - Thanks for the input. I have very little understanding of circuit design. The atmega2560 was because of needing the analogue lines, but a multiplier is better. I also thought of moving the power supplies to the battery pcb (1 per pcb), now add a 328P and 4067 MUX. You could monitor more cells, current and battery temp. the pcb under the APM would control routing of power and communicate with the battery pcb's via i2C. You purchase the board to sit under the apm and only the number of battery pcb's you want. Or just the battery pcb and connect it directly to an apm 1/2.0/2.5

You need to be a member of diydrones to add comments!

Join diydrones

Comments

John Edmonds November 5, 2012 at 8:23pm

Hi, Guys

Chris - That's great to learn. A guaranteed compatible power source. Is it backwards compatible to APM 1 & 2.0?
Peter Chaffe - mostly it is about having a known good power source.
Chris/Peter Meister - I think a large minority of people are using two or more batteries or would like an option too. To increase flight times or because of space restrictions in their airframes, or to run one of the electronic systems from a separate power source (e.g FPV equipment). I think a system with some form redundancy or isolation would be welcomed. A small backup battery in case of unexpected exhaustion of primary battery, or Isolating batteries in the event of a crash, or isolating and jettisoning a battery.
Matthew - There would be less benefit for a mulitcopter than a plane or rover. The complexity kind of arose once a micro processor was added. It has the abilities, so add the circuits and see what the community can do with it.
John - Thanks for the input. I have very little understanding of circuit design. The atmega2560 was because of needing the analogue lines, but a multiplier is better. I also thought of moving the power supplies to the battery pcb (1 per pcb), now add a 328P and 4067 MUX. You could monitor more cells, current and battery temp. the pcb under the APM would control routing of power and communicate with the battery pcb's via i2C. You purchase the board to sit under the apm and only the number of battery pcb's you want. Or just the battery pcb and connect it directly to an apm 1/2.0/2.5
John Rambo November 5, 2012 at 8:45am
some questions:
- atmega2560 costs 5x the 328P, so wouldn't it make more sense to use 328P + 4051 or 4067 MUX for voltage/current sensing? Or is it the same uC to be used, not the separate-one?
- uC controlled battery switch, where the same uC is powered from that battery sounds a bit irrational. Wouldn't analogue O-ring setup look better here?
Matthew Crocker November 5, 2012 at 7:04am

I plug all 4 BECs from my 4 ESCs into my APM2.0. I haven't had an issue at all with power. Not sure why I would want to add the extra weight and complexity to my quad.
3D Robotics

Chris Anderson November 4, 2012 at 5:02pm

Doug. You're smart ;-)
R. D. Starwalt November 4, 2012 at 4:27pm

And so, perhaps, the revelation of the application for the 'mystery port'?

-=Doug
3D Robotics

Chris Anderson November 4, 2012 at 11:57am

Zach: The case will be released at the same time.
Zach Bayne November 4, 2012 at 11:53am

I know this is off topic, but with that coming to the store soon is there any news for the apm 2.5 case?
3D Robotics

Chris Anderson November 4, 2012 at 11:16am

Peter, it's just designed for a single battery input (to keep it small and inexpensive; almost all users have a single battery). But if there's enough demand for a dual battery version, we'll add that, too.
Peter Meister November 4, 2012 at 10:52am

Chris, glad to hear of the APM power supply. This will solve a lot of problems for those just starting. Hopefully it will include dual battery supply inputs for fault tolerance :)
Peter Chaffe November 4, 2012 at 10:04am

John, I can understand that your motives are good but I can't see the APM2.5 requiring the complication of any advanced power supply unit. The specs say 5v0 -+ 0v5 ie 10% so what is special about that?
All esc becs that I've tested had tolerances better than 3%.

Even a quad will likely carry 4 becs integrated into it's ESC's, the majority of these are usually around 3a or more. The apm takes around 200mA (inc gps,Rx,3DR Radio, Sonar). The bec part will hardly get warm.

I think the reason why there has been so many questions and confusion surrounding the v2.5 power supplies is because it was so badly explained in the wiki.

Peter

of 2

This reply was deleted.