Looking for a DIY Powerbox

[Harpia]

Hi


I searched but I did not find a DIY project of Powerbox. Somebody has a project of DIY Powerbox?

I found this here http://www.powerbox-systems.com/e/powerbox_systeme/powerbox_40_16_expert/start.php but I would like to make one :)

Edson

[Acetronics]

Hi, Harpia

No special secrets about the "PowerBox" and clones ...

just a Schottky diodes coupler ( I didn't read about voltage regul ...) for the two batts. an electronic switch ( P mosfet ). a memo batt indicator . 4 servo signal amplifiers .

You just build that on a THICK copper PCB ( 70 or + µm ...) and that's all !!!

You will find all the sub-systems schemes on the net ... just assemble.

Alain

[Harpia]

Hi, Harpia

No special secrets about the "PowerBox" and clones ...

just a Schottky diodes coupler ( I didn't read about voltage regul ...) for the two batts. an electronic switch ( P mosfet ). a memo batt indicator . 4 servo signal amplifiers .

You just build that on a THICK copper PCB ( 70 or + µm ...) and that's all !!!

You will find all the sub-systems schemes on the net ... just assemble.

Alain

Hi Alain

Thanks for explanation!


Edson

[flashrc]

Hello,

I am starting designing "my" powerbox system (this is for my Discus 4.3m), not to gain money, just for the pleasure of doing it.
The regulation is known for me (I will use the MIC29302), what is unclear is the stage of "amplifier" they are using inside: from my understanding it is just a 5V amplifier for the signal, so may be using a saturated op amp at 5V out is enough, any experience on this area ?
Nobody has opened his box to see the components used ?
Each servo has already an amplifier, so what is the need of such stage ?

I understand it can also avoid feedback voltage from high torque servo.

Thanks for your comments.
Hervé

[Tomapowa]

Here's a project I really never finished...

http://www.diyrc.com/optoisolators.htm

It basically amplifies the R/C servo signal from your Rx (via inverter), passes it to an optoisolator, amplifies again (via another inverter) and then passes it onto the servo. Use highspeed inverters for better accuracy and minimal phase shift. The servo's power also comes off of a built-in regulator (7806 I beleive). I never developed a circuit diagram for this... beleive it or not... but did build a board, tested and still being used in my 1/4 scale Patty Wagstaff.

[flashrc]

Very interesting board. Opto isolation seems very interesting.
Any drawback known so far ?
I bookmarked your website :D

[Tomapowa]

Only drawback (or design difficiency) is that the amplification (inverters) can cause significant pulse reshaping... higher quality and high speed inverters make significant improvements though. If you are only looking to power your servos from another power source (i.e.no optoisolation), then simply connect the +V wires from each servo to another power source, such as a swithcing regulator capable of providing 3+ amps or whatever you require (typically higher when using many large digital servos). I built such a device that plugs into the Futaba 9 channel PCM receiver... provides regulated (+2 amps) from unboard regulator.

[flashrc]

Looks like you have done many nice circuits !
Actually, I am really looking for doing both:

- Vreg for DUAL bat source (2 IC for the regulation)
- amplifier (your proposal with double inverter + opto isolation of the signals) for the servo signals.
- and add a PIC for LED display, LED error, ...

Actually reshaping signal for servo's is not so much a draw back but give more nice waves to them, regulated at 5V, isn't ?

Having a regulator instead of Diode for DUAL bat plays the same role, i.e. if one is in short circuit, the other one doesn't discharge into it: do you confirm ? because in some schematic, I saw a Vreg chip (like MIC29302) + a diode in serial.

Thank for your precious comment.

[flashrc]

I forgot to ask you the value of the Resistors in serial between inverter-opto-inverter.
Thank you.

[Tomapowa]

I forgot to ask you the value of the Resistors in serial between inverter-opto-inverter.
Thank you.
Oh yeh... just noticed it is not listed on either drawing...

R1-R4 and R13-R16 (input amplifiers) are 220 ohm while R5-R12 (output amplifiers) are 1K ohms.

Re: your other questions... when I say the inverters cause "reshaping", I mean they alter the rise/decay of the servo pulses (no longer square in shape). This in fact changes (decreases) the pulse widths a little, causing slight servos positioning issues.

Ex. Using slow cheesy 4069 inverters, I think I was seeing 1420ms pulses where in fact the input was 1500ms... or something like that. :confused:

As for the battery redundancy feature, as long as both regulators have built-in "diodes" as you mentioned, you should be safe.

p.s. I also dug up a circuit diagram for one channel... instead, I used inverters vice "buffers" and used 2 DIP ICs that house 4 opto-couplers each (saved board space)

[Tomapowa]

Here's a simpler optoisolator I dug up... don't know how well it works though as I never built/tested. No amps... just optocouplers.

http://www.toshiba.com/taec/components2/Datasheet_Sync//207/4346.pdf

[flashrc]

Thank you Tomapowa.
The solution with inverter-opto-inverter looks nice and should really cut all noise which can come from long wires (in sailplanes wings for instance). I intend to do a SMD version.
I still do not understand why you are loosing pulse width because when I look at the datasheet, the response time is far less than what you are loosing. It is stated to put a Cl o 50pF, have you tried ?
I saw an old review of powerbox system 40/24:
http://www.modellbau-deutsch.com/e/testberichte/pdf/powerpox_engl.pdf
it clearly shows the IC on each servo signal which is a SOI8 plus some resistor. I don't have any idea of what it is. On their latest design, they are speaking of push pull technology, I guess this is for fast response in bothe edge of signal.

[Tomapowa]

Thank you Tomapowa.
The solution with inverter-opto-inverter looks nice and should really cut all noise which can come from long wires (in sailplanes wings for instance). I intend to do a SMD version.
I still do not understand why you are loosing pulse width because when I look at the datasheet, the response time is far less than what you are loosing. It is stated to put a Cl o 50pF, have you tried ?
I saw an old review of powerbox system 40/24:
http://www.modellbau-deutsch.com/e/testberichte/pdf/powerpox_engl.pdf
it clearly shows the IC on each servo signal which is a SOI8 plus some resistor. I don't have any idea of what it is. On their latest design, they are speaking of push pull technology, I guess this is for fast response in bothe edge of signal.

Hi flashrc,
After looking over the PowerBox and PCB, it appears they are not using optoisolators, even though they say their products "Isolates the receiver from possible RF noise". Without optos, you are not isolating signals from input to output (i.e. most likely still using common ground). Looks like they are only using pulse amplifiers (i.e. glorified non-inverting buffers in SMD form) for amplifying and reshaping signals. Other circuitry shown (2 other DIPs) are for the monitoring of each regulator output (could do with 2 cheap Volt Watch units). Not all too complex to be honest. Some other manufacturers do in fact sell "opto" versions of these... such as the Jomar (EMS) Glitch Buster (http://www2.towerhobbies.com/cgi-bin/wti0001p?&P=SM&I=LXPNU7)

[flashrc]

I share your comment, so there is two approach: opto but drawback is reshaping, amplifier but drawback is noise immunity.
However even using Powerbox system, I suppose the ground and VCC is common, only the signal is "decoupled".
They are using SOI8 chip for each line (servo), I think I have found what they might use, or anyway, I might be interested in this: this is digital isolator
Si8440 from Silicon Labs (they do a quad version)
ISO721 from TI in SOi8
What do you think ?

[Tomapowa]

I share your comment, so there is two approach: opto but drawback is reshaping, amplifier but drawback is noise immunity.
However even using Powerbox system, I suppose the ground and VCC is common, only the signal is "decoupled".
They are using SOI8 chip for each line (servo), I think I have found what they might use, or anyway, I might be interested in this: this is digital isolator
Si8440 from Silicon Labs (they do a quad version)
ISO721 from TI in SOi8
What do you think ?


Cool...Pretty decent specs on those isolators too...
http://focus.ti.com/lit/ds/symlink/iso721.pdf
Just ordered a few samples to mess with...

[flashrc]

So you will get them before I do :-)
Let me know your results (oscilloscope: response time, etc ...) once available.
Do you think Si8440 is the same than 4xIso721 ?

[Jag01]

I am closely watching this thread as I too would love to build myself a power system including a regulated power supply, dual battery supply and noise filtering.

Has anyone come up with the regulator solution yet?

cheers,

Jeff