PDA

View Full Version : Voltage for 6 Port AX/MX Dynamixel Power Hub



guruguhan
02-15-2014, 11:32 PM
Hello,

My question is in regards to using the 6 Port AX/MX Power Hub (http://www.trossenrobotics.com/6-port-ax-mx-power-hub?relatedid=1512) - to date, I've been using this with the AC adapter that came with the Dynamixel Servo Manager Kit, but I want to switch to using a battery.

The project is powered by two LiPos (each 14.8v 4S 60C continuous 6600mAh) in parallel. What voltage does the power hub take? I would like to know what I need to convert the voltage coming from the battery to.

Three AX-12s are connected in a chain to the hub (so in parallel, and I've read elsewhere max draw is up to 900mAh at load).

(If it is relevant, I'd like to achieve similar torque etc that I currently get with the ac adapter).

Thanks

tician
02-16-2014, 12:06 AM
AX series servos are rated for 12V (3S LiPo). The standard power supply from Robotis and Trossen is rated 12V at 5A; that is sufficient to power the 18 AX-12/18 in the Bioloid Humanoids, and the 20 MX-28 in a DARwIn-OP, without any issues. The stall current of each AX-12 is 1.5A, but you should not encounter that much current draw from each servo on a continuous basis. Providing anything less than 12V 2A is likely to cause stability issues; this comes from experience with faulty 12V 5A supplies and underpowered 12V 1.2A supplies unable to keep 2 servos consistently communicating with a controller (and possibly reseting the ID to 1).

guruguhan
02-16-2014, 12:30 AM
Thanks Tician,

So if I'm understanding what you're saying, I should be ok.

This might be a dumb question...but if I'm going from the batteries, to the voltage converter to the power hub terminals (for wire leads) - and then decide to plug in the ac adapter when it doesn't need to be "mobile"...must I disconnect the leads coming from the battery side before connecting the ac adapter to the power hub? Would I blow up my batteries or something if I didn't?

tician
02-16-2014, 01:05 AM
You might damage the voltage converter and/or external power supply if the output voltage of both is not exactly the same. It probably would not damage the batteries, but the easiest way to prevent damage to the voltage converter would be to disconnect it from either the hub or the batteries.

Finding DC-DC voltage converter to reliably step down a 4S battery only to 12V might be a bit difficult and/or expensive. There is a voltage drop across each of the various components within the DC-DC converter, so most inexpensive step down converters require the input be at least 3~8 volts above the desired output voltage. A 4S (12.8V~16.8V) to 12V converter will likely need to raise the input voltage to a higher intermediate voltage before dropping that down to reliably produce 12V at the output. That process of raising then dropping the voltage increases complexity and cost, and decreases efficiency.

guruguhan
02-16-2014, 01:16 AM
Thanks again Tician...

I was worried about the efficiency of all this stepping up and down. I was thinking of using 14.8V batteries because I will be using a NUC and need to step up to 19V, so thought starting at 14.8V was better.

If I used two 3S 11.1V 6600mAh batteries in parallel it sounds like its better correct? The batteries are 11.1 instead of 12V...will that be a problem for the Dynamixels? And I guess using the 11.1V batteries won't be a problem to step up to the 19V for the NUC?

EDIT: If it's possible, I think I'm going to switch to a single 3S battery with as much power as possible for the servos and NUC.

KevinO
02-16-2014, 02:06 AM
I have used an Intel NUC for my last two robot projects and the current one I'm working on now. Three cells are fine.

KevinO
02-16-2014, 02:09 AM
Why do you need two 6,600mAh batteries? What kind of run time are you needing? I run off of one for over 20 min. Can you tell us more about your project?

jwatte
02-16-2014, 01:35 PM
4S may kill the AX servos. It actually killed a few MX servos for me. I'd recommend 3S.

The part you linked to is fully passive -- it just connects wires/connectors. It doesn't have any voltage spec or tolerance; it's the servos you have to worry about.

If you keep the power supply and the batteries both connected at the same time, then make sure you prevent back-feeding from battery into the power supply, and also make sure that you don't over-charge the batteries. LiPos can catch fire (or at least lose a lot of lifetime) by over-charging. Exactly 12V into 3S is pushing it a bit high for a float voltage; you'd be better off (battery lifetime wise) with 11.7V. Even 12.1V may literally push the batteries too hard, depending on the manufacturer and chemistry involved.

Or don't connect more than one at a time.

guruguhan
02-16-2014, 04:31 PM
Thanks for all the help guys - 3S it is then, thanks for saving me some $!

Kevin - in my dreams I'd like to get 2+ hrs of run time, so I know for certain whatever I do I'll be falling short (there are also other sources of battery drain which haven't been mentioned). But I would like as much power as my wallet and chassis can handle without making the system inefficient. You've helped me on a few threads over the past few months (I'm not on here everyday - nor working on my robot everyday so it moves along slowly) - thanks for that (and everyone else of course). I'll certainly share more about my project but its my first, and though I'm quite happy with its beginnings, I'd rather start a thread at a later date (similar to what you've done with your robots). I'm happy to share...its just not ready yet. Once the body is complete (and about to start to focus on the programming side) I'd definitely like to get more of your input.

Jwatte - I'll stick to the 3S and probably just disconnect the battery if I'm using the ac adapter. The thought of a fire freaks me out and I don't know how to prevent back feeding etc (doesn't sound like its worth the hassle).

Noog
02-16-2014, 04:54 PM
Here is the (Bioloid Auxiliary Power Module (http://www.huvrobotics.com/shop/index.php?_a=viewProd&productId=7)) from HUV Robotics. It is a drop in module that will allow switching between wall adaptor and LiPo battery (plus it also gives battery low voltage monitoring/cut-off).

jwatte
02-16-2014, 07:24 PM
Or just put in a Schottky diode with sufficent amperage rating (5A or more.) As a bonus, the power supply will then keep the batteries charged as they deplete. The diode drop is enough to make sure 3S batteries are not overcharged from a 12V supply, even if that supply is slightly high.

guruguhan
02-16-2014, 08:28 PM
How safe is that method of charging jwatte? I would love to have that kind of functionality. Why are LiPo chargers so expensive if such a simple connection works?

The Schottky diode route seems a little above my head at the moment...

I like that HUV Robotics board, think I might go for it but will keep reading for now...

jwatte
02-17-2014, 12:32 PM
Why are LiPo chargers so expensive if such a simple connection works?

It doesn't "charge" the batteries as much as "keep them from falling too low." They will not have a full charge.
Also, LiPo chargers will balance the cells so they're all charged equally; this method will not, so you'll still want to use a real charger once in a while.
A LiPo charger also has other safeties to avoid over/undervoltage, adapt to different numbers of cells, etc.


The Schottky diode route seems a little above my head at the moment...

It's a diode. You can buy it and solder it in series with your connectors. Just make sure to solder it with the ring away from the "+."
Here's one (5A rating): http://www.digikey.com/product-detail/en/SB540E-G/641-1419-1-ND/2075783
Here's another one (10A rating): http://www.digikey.com/product-detail/en/SR1045-TP/SR1045-TPMSCT-ND/2334476

Anyway, if you want a complete board, the HUV board seems fine for your needs.