PDA

View Full Version : [Question(s)] Phoenix Kit Build Questions



Kai
09-01-2009, 09:16 AM
Hi there, I just bought the phoenix hexapod kit (so excited I squeeled when it arrived) and I've been taking stock of the parts and reading the assembly instructions and there is one thing I'm not clear on.

The kit ships with 18 Aluminum Servo horns. I am assuming these are to replace the plastic horns that are on the servos now. However the instructions hint that if you replace the horns you need to make sure the new horns are aligned "correctly".

There are no markers or numbers in the Aluminum horns like there are on the Plastic ones. Is there a hard and fast rule for aligning then correctly? The 4 holes are vertical and horizontal? I just want to make sure I get it right before I assemble then find I have to take the whole thing apart again because I did step 1 wrong. ;)

Cheers!

Kai

tom_chang79
09-01-2009, 12:41 PM
hello Kai, welcome to the world of robotics!

First, to center the servo horns, you need to hook up the SSC-32 and the servos and do a "ALL=1500" command (it's a button) on Lynxterm (available free off of Lynxmotion's .com site).

Once you have the servo centered, you then need to make sure that all the horns are "aligned" for this center point. Make sure that You have two holes straight vertical, two holes straight horizontal with respect to the servo casing...

Don't worry if teeth in the Servo horns doesn't EXACTLY line up the threaded holes in the horn vertically or horizontally, you will most likely make some fine adjustments once your phoenix is assembled. I know I did, I know many other have also, and I don't know if anyone has all 18 servos exactly lined up without fine alignment (called trimming) later on.


What I mean by fine alignment is, after you put your horns on as close to "center" as possible. The true center of the servo horn might be at pulse width of 1525 instead of theoretical 1500...

Kai
09-02-2009, 10:59 PM
Thanks very much or the reply! I've started assembly and I'm having a ball. Fingers are sore from all the small screws... but enjoying it.

I do have a further question tho. I've gotten to the SSC-32 portion of the assembly and I notice a couple of points where the posted instructions differ from what I actually have.

I note for example that I have 2 x cables with a connector for the battery pack and a switch inline.

I also have 2 x battery packs. But the instructions seem to indicate only one of those cables being plugged into the SSC-32 and make no mention of mounting the battery pack at all.

Do I use both battery packs? Do I cable both into the power in point on the SSC-32?

They also show a switch wired into the 9v board power cable that is not present in the 9v cable I have.

Actually now that I think about it I also got 2 x 9v cables as well. I assume I don't need the second as the bot board feeds off the power feed from the SSC-32?

Thanks in advance.

Kai

DresnerRobotics
09-03-2009, 08:46 AM
You've got a couple options there:

You can power the logic (make sure to set the jumpers to isolate Vin and Vlogic) separately using the 2nd 6v nimh battery, or you can use the 9v connector and power it via 9v battery.

For servo power, use the 6v battery pack of course. If you choose to power logic with a 9v battery, you have the option of using the 2nd battery pack as a spare, or wiring it in parallel for double the runtime.

We ship the kit with 2 batteries for these reasons.

The two switches are for each logic and servo power, if you choose to use the 6v nimh battery pack for each. If you want to talk about it further, feel free to give me a call, numbers at the top of the page ;)

Tom- great post, +rep!

Kai
09-03-2009, 12:22 PM
ahhh thanks.. that makes a bit more sense.

If I decide to go 9v for the logic on the SSC-32 and on the Bot Board 2 I assume I can power them each via their own 9v if I wanted as opposed to the passthrough as shown in the online instructions?

Also if I wire up both battery packs to the SSC-32 do I wire them both into VS1 or one into VS1 and the other into VS2?

EDIT : oops, for clarity I mean if I want to power the servos with both Battery packs above.

Thanks again.

Adrenalynn
09-03-2009, 01:43 PM
Powering servos from a 9v battery would be exceptionally not recommended - assuming you don't like the ozone-ish smell of burnt electronics.

You want the 9v going in to VL on the servo controller and to power the BotBoard. You need a 6v battery going to the servos.

That's too many servos to run off the onboard regulator by a loooooong shot. Two, maybe three small servos max.

DresnerRobotics
09-03-2009, 02:10 PM
ahhh thanks.. that makes a bit more sense.

If I decide to go 9v for the logic on the SSC-32 and on the Bot Board 2 I assume I can power them each via their own 9v if I wanted as opposed to the passthrough as shown in the online instructions?

Also if I wire up both battery packs to the SSC-32 do I wire them both into VS1 or one into VS1 and the other into VS2?

EDIT : oops, for clarity I mean if I want to power the servos with both Battery packs above.

Thanks again.

You can power the SSC-32 and Bot Board from separate 9v batteries if you choose so, or power them both from the same, depending how you wire it. Make sure the VL=VS jumper is disconnected though.

As long as VS1 and VS2 are jumpered to be connected, you can connect a battery to each terminal, which will take care of your parallel connection. Likewise, you could wire the two batteries in parallel via a splitter and run them into VS1.

jes1510
09-03-2009, 02:31 PM
One more thing about the separate power supplies; servos are extremely power hungry. Servos can need a pretty good amount of current when they are loaded and if your battery pack isn't up to the task (high internal resistance) then it can cause power fluctuations on the power bus. That can cause all kinds of funkiness such as microcontrollers resetting etc. It's almost always a better idea to run motors off of one power supply and the electronics off of another.

Kai
09-03-2009, 07:28 PM
Once again.. thanks for all the really helpful replies. So I've taken a photo of my understanding of a pretty ok setup. See the attached file. Is this ok or am I going to melt something?

I'm going to power the logic off 1 x 9v and the servos off 2 x 6v.

I have 9v plug connected to VL on the SSC then a pair of wires carrying that up to VL on the bot board as well. I have VL-VS unjumpered on the SSC.

I then have one of the 6v batter pack plugs wired into VS1 and the other into VS2.

VS1=VS2 are jumpered on the SSC.

On the Bot board I also have a jumper on VS-VL which I'm unsure about. In the online schematics it's shown unjumpered but I'm unsure how voltage get to the PS2 controller without it?

Also my board shipped with the jumpers on VS-5v for each of the pinouts in the same row as the PS2 controller on the BB. Also not on the schematic... these need to come off?

As well as the jumpers on DAT CND and ATT? (also not shown in the schematic but on my board as shipped)

And last thing.. in the online instructions there are two schematics for the SSC-32 and one shows both jumpers in the "BAUD" area connected and the second only show the first one (left to right). Which is correct or are they different stages in the setup?

*phew* I think that should about do it. ;) Thanks everyone for your help! I'm a really cautious type and while I think I have it sussed I want to be 100% before applying power. ;)

tom_chang79
09-03-2009, 10:46 PM
well, the VS=VL jumper works the same way as it does on the SSC-32... It's basically the jumper to see if you want to power the on-board regulator for the Logic circuit (microchips, microcontrollers, etc.) off of the VS (servo power) rail...

If you're tapping into the BB2's VL directly from SSC-32's VL (which is what I do too), you can take that jumper on the BB2, VS=VL off...

As for the DAT CND and ATT, unless you're planning to use the PS2 controller for it, you don't need to touch those... As for the PS2 controller, I have no experience with it...

You're on the right path, you're almost there!

Kai
09-05-2009, 01:49 AM
Once again.. thanks for all the help everyone...

I am finally at the end and what do you know. I'm getting the dreaded "continuous beeping of death" from the bot board when I power it all up. :/

I've gone back and double checked everything and it all matches this image

http://www.lynxmotion.com/images/assembly/phoenix/phoesch1.gif

Exactly. All the jumpers are the same and all cables connected as shown.

When I power one the PS2 controller the left red LED on the controller flashes at regular intervals (which is I assume it trying to establish a connection) and with each beep of the bot board the left (as shown in the attached image) red LED flashes.

The programming of the bot board went fine and nothing else seems awry.

Suggestions?

EDIT : the only difference I can see is that on the wires coming out of the PS2 receivers cable there is an additional black wire not shown on the diagram above that is feeding into the same connector as the red wire.

Also the colors on the diagram above do not match the ones in this diagram from the same instructions. http://www.lynxmotion.com/images/jpg/wire08s.gif

Do I have the older controller?

Kai

Kai
09-05-2009, 02:03 AM
Never mind.. I am an idiot.

Apparently it's possible to plug the PS2 receiver into the reciver cable upside-down.

Who knew? :tongue:

Sorted and the bot is working! Huzzah!! Off to play! Thanks so much to all the people who helped out!!

DresnerRobotics
09-05-2009, 09:31 AM
Good to hear! Was about to post about that beep error, which indicates the botboard isn't detecting the ps2 receiver. I've had a few calls about that. =)

LinuxGuy
09-05-2009, 02:00 PM
Just a small note on using two battery packs to power servos.

With one battery connected to the VS1 terminals and another battery connected to the VS2 terminals of the SSC-32 and the VS1=VS2 jumper removed, servos on one side of the SSC-32 gets power from one battery and servos on the other side gets power from the other battery.

This is the preferred way to power servos on both sides of the SSC-32.

8-Dale

DresnerRobotics
09-05-2009, 04:58 PM
Just a small note on using two battery packs to power servos.

With one battery connected to the VS1 terminals and another battery connected to the VS2 terminals of the SSC-32 and the VS1=VS2 jumper removed, servos on one side of the SSC-32 gets power from one battery and servos on the other side gets power from the other battery.

This is the preferred way to power servos on both sides of the SSC-32.

8-Dale

Yeah, but then if one battery discharges more than the other, you have half of a hexapod responding quicker than the other. It's in my opinion, to leave VS1=VS2 jumpered so you have equal voltage across all servos.

Adrenalynn
09-05-2009, 06:09 PM
The only time you don't want to do that is when your current draw (amperage) is too high for the traces and you risk lifting them.

In that case, it should be fed from an offboard switching regulator. I don't know what they power bus traces are rated for, but I wouldn't push it over 10A continuous myself. I think even that's probably pushing it.

LinuxGuy
09-05-2009, 11:24 PM
Yeah, but then if one battery discharges more than the other, you have half of a hexapod responding quicker than the other. It's in my opinion, to leave VS1=VS2 jumpered so you have equal voltage across all servos.
For this method, I'd recommend at least a 6V @ 2800mAH pack to run both VS1 and VS2.

8-Dale

DresnerRobotics
09-06-2009, 12:22 AM
For this method, I'd recommend at least a 6V @ 2800mAH pack to run both VS1 and VS2.

8-Dale

I don't think we're quite on the same page here; my suggestion earlier was to run both 2300mAh packs in parallel (one plugged into each VS1 & VS2) with VS1=VS2 jumpered. Running two battery packs on VS1 and VS2 without the VS1=VS2 jumper doesn't really have any advantage, and you run the risk of getting the packs at different voltages which could effect performance.

That said, I have my own Phoenix running off a single 2300mAh pack and it works just fine, a 2800mAh battery would just give it a bit more runtime, certainly not a necessity. The Phoenix simply doesn't have enough peak current draw to necessitate a large capacity battery, unless you simply want more runtime. I get about 20-25 minutes on a single 2300mAh battery freshly charged.

Adrenalynn
09-06-2009, 01:00 AM
Agreed with Tybs - if you have two identical packs, you don't have two identical current draws. For lithium-*, it's going to hold its voltage, but not its ability to deliver current. For nimh or nicad, you're going to have unbalanced voltage. Unless both sides are delivering exactly the same current all the time AND the batteries really are perfectly identical - which they aren't.

Kai
09-06-2009, 02:28 AM
Thanks for all the really good info guys.

May I introduce you all to... "Toby".

Don't ask me why.. I ran a vote with my friends and Toby won. :P

I'm currently having to run him off 1 battery pack at the moment anyway as I didn't really have room under the SSC-32 for both packs. I'm having to mount the 9v for logic on the underside as it is (means I have to walk him higher than the usual walking height as it catches sometimes on my carpet.

I'm looking at building an external shell of some kind for him and hopefully then I'll have space on top to integrate the second battery pack.

I must admit I'm very impressed with the strength of the digital ultra-torque servos. It's quite powerful. My only real problem is that as it walks the servo horn screws come loose no matter how hard I tighten them. This results in more than a little "wobble" in his walk after a while and I have to stop him to tighten them up. Any pro-tips on getting round this? I don't want to go so tight I strip the bolt.

Other than that, a _very_ rewarding project/kit that I recommend to anyone. Challenging without being incomprehensible for a beginner like me.

I must admit when he stood/walked for the first time I cackled out "It's alive, IT'S ALIVEEEE!! AAHAHAHA"

I'm sure my neighbors think I'm crazy now. :tongue:

DresnerRobotics
09-06-2009, 02:32 AM
Use some Loctite my friend, it works wonders on those screws. I had the same problems with mine until I remedied it with some standard Loctite.

You're using 5645 servos? I've got 18 spare and have been thinking about upgrading my own Phoenix, hrm..

Kai
09-06-2009, 11:12 AM
Use some Loctite my friend, it works wonders on those screws. I had the same problems with mine until I remedied it with some standard Loctite.


Cheers. I'll try to find some. Sounds like the solution. :P



You're using 5645 servos? I've got 18 spare and have been thinking about upgrading my own Phoenix, hrm..

Yeap.. they are surprisingly strong. When he's standing I can push a significant portion of my arm weight onto his back. To be quite honest I stopped applying weight before I felt the servos start to give for fear of breaking them, I was leaning that hard on them.

I might put together some kind of weight test rig and test him from sitting on his belly to see "unable to stand up" point.

darkback2
09-06-2009, 12:55 PM
One thing to be careful of, I'm not sure the 5645s have overload protection. Seams like you could end up burning things out before you know you had gone too far.

Oh...and now that its done...how about some pics and a video!

LinuxGuy
09-06-2009, 01:38 PM
I don't think we're quite on the same page here; my suggestion earlier was to run both 2300mAh packs in parallel (one plugged into each VS1 & VS2) with VS1=VS2 jumpered. Running two battery packs on VS1 and VS2 without the VS1=VS2 jumper doesn't really have any advantage, and you run the risk of getting the packs at different voltages which could effect performance.
I do understand what you are saying and proposing. I just don't agree. :)

8-Dale

Adrenalynn
09-06-2009, 02:55 PM
So you propose that connecting two different current sinks to two different battery packs will drain both battery packs at the same rate. I haven't observed that, and the electrical theory I was taught seems to differ.

Can you propose a mechanism by which this can be observed, or at least some hypothesis on why it might be the case? Some form of quantum communication whereby quanta are exchanged between the battery packs, or maybe the circuit itself? What battery chemistries have you observed this behavior in?

Any pointers you can provide to assist in duplicating these results would be much appreciated!

Thanks!

LinuxGuy
09-06-2009, 04:17 PM
So you propose that connecting two different current sinks to two different battery packs will drain both battery packs at the same rate. I haven't observed that, and the electrical theory I was taught seems to differ.
I did not propose such a thing.

8-Dale

Adrenalynn
09-06-2009, 05:24 PM
[shrug] It doesn't really matter. There's a proper way and an improper way, and I think that's all that's important to the OP. Tybs' post is factually correct and accurate.

LinuxGuy
09-06-2009, 06:30 PM
[shrug] It doesn't really matter. There's a proper way and an improper way, and I think that's all that's important to the OP. Tybs' post is factually correct and accurate.
My information is just as correct as what Tyberius posted. What I proposed is just a different way of doing things - no less correct, accurate or useful.

If this is how you are going to be, I just won't bother trying to help and contribute.

8-Dale

DresnerRobotics
09-06-2009, 10:59 PM
My information is just as correct as what Tyberius posted. What I proposed is just a different way of doing things - no less correct, accurate or useful.

If this is how you are going to be, I just won't bother trying to help and contribute.

8-Dale

I just don't see the value in using 1 battery for the 9 servos on the left side, and 1 for the 9 on the right. It's certainly a different way of doing it, but I don't see why you would. I mean if you can explain to me your logic behind this I'm all ears, I just haven't heard you explain why it's a better or even comparable solution. :)

If your hex were to say, turn to the right more than the left, or decide to roll the body more on one side than the other, than you're creating more current draw on one battery pack over the other. This can lead to unbalanced voltages and thus create a difference in performance between the left and right side of your hexapod. While this might not be a major difference, it's sort of an unnecessary risk given there isn't any advantage over your proposed power solution.

The Phoenix doesn't overdraw even a single 2300mAh battery; any modern NiMh chemistry can provide more than enough amperage for the robot without even overheating the batteries in the slightest.

Adrenalynn
09-06-2009, 11:32 PM
>> While this might not be a major difference

Even if they were identical, the current draw of the servos will always be different. No perfect motors in an imperfect world. And over time, that adds up tremendously.

Kai
09-07-2009, 12:36 AM
One thing to be careful of, I'm not sure the 5645s have overload protection. Seams like you could end up burning things out before you know you had gone too far.


Good tip.. I'll leave off stress testing then. ;) Seems to be able to easily carry more than I could physically fit on it anyway. :)



Oh...and now that its done...how about some pics and a video!

Coming soon. Gonna put a build website together. Not that it's terribly different from anyone elses phoenix hexapod at the monent. I have some ideas for body kit I want to add to it tho.

I'm thinking something that ends up looking vaguely like this would be hawt.

YouTube - Spider robot - walkcycle test

Just need to get some time to get some plastic moulding sorted.

Kai

jes1510
09-07-2009, 12:51 AM
Dude that shell is just plain awesome!