Hi Zenta,
I finished the Ripple gait and made it possible to start walking or change directions without initialization. The first thing I want to do next is take a good look at the balance calculations. I really think this makes the movements much more fluent movements. But I’m afraid that big updates will come after my vacation 8)
Xan
Hi,
I was a bit unsure if I shouls start a new thread about this, but since there already are very many “Phoenix threads” on this forum I thought using this thread should be ok 
I’ve started building my red Lynxmotion Phoenix now. One of the major challenges with Phoenix are the small body and where to place the batteries, wires ++. On my original Phoenix I used a very cheap solution with a AA batteryholder for some standard NiMh 2700 mAh AA cells + one 9v battery for electronics. I was never satisfied with this, and it gave me not much effective run time.
Anyway, this time I wanted to use LiPo for the servos. Some of you have probably seen my Felix project, where I’m gonna use some beefy 4000 mAh LiPo’s. And it was a real coincidence that I discovered that these LiPos made a perfect fit between the hex spacers! (actually I had to turn the hex spacers with one of the flat sides against the battery to make it fit).
4000 mAh inside a Phoenix, yeah! 
I just had to make this work! 
I just took this picture for fun (my little robot family ) :
Very soon I discovered that I had to make some custom spacers for the SSC32 card, because I also wanted to make space for the DC-DC converters. My goal was also to make it very easy to replace the LiPo, therefore I made a bracket/plate to be placed between the converters and the LiPo:
When I got the LM Phoenix kit I was wondering what the 3mm hole at the front on the body plates was meant for (since there was only 4 spacers), Jim? 
. As you can see I used this hole to fix the the custom plate and also giving a 5. body spacer.
Then I hooked two Turnigy DC-DC converters together (removed the software switch, just cut the wires):
Here is a picture of the converters mounted and wires soldered to the main switch:
As you can see I’m gonna use two NiMh 200 mAh 9v batteries in parallel for the electronics (this picture only show one):
Ok, some of you may think that the batteries are going to be in conflict for the travel range of the coxa’s. Yes, but not very much… There are always some “give and take”…
I didn’t use the original 3mm screw that came with the kit, I wanted to use black 3mm hex screw for mounting the tibia servos on the tibia part:
I also made some custom spacers (4,5mm(outer) x 3mm(inner) x 2,4mm (height) ), to make it easier to align every servo to the tibia part:
So far I didn’t pay much attention to what was the rear and front part of the Phoenix, and how the SSC32 card should be oriented. But finaly it turned out to be the same way as for my original Phoenix (sorry Jim 
). But for those of you who are interested in the PEP sheet I have a made it possible to define the SSC32 Pin setup easier, I’ve also added a coxa offset for the rear and front coxa.
Here is a picture showing the front of Phoenix. I’ve used 4mm black hex screw for mounting the coxa (only 2 screw) and femur servos:
I’ve used black braided cable sleeving on the servo wires. Another picture of my red Lynxmotion Phoenix so far:
I’m gonna mount Bot Board II on top of the SSC32, but I’m a bit unsure about the servo wires. I’ve been very tempted to just shorten them and remount new servo connectors. What do you think? Or should I just leave them there, you won’t probably see them when the Bot Board are mounted anyway.
A picture showing Phoenix with the LiPo in front of her:
Kare, EVERYTHING about your build is so quality… Do you build/assemble machines for a living? 
Yeah, I wondered what that extra hole towards the front of the upper and lower deck was for… I’m contemplating on putting a post there myself…
I really love your battery solution. So nice and clean and you even have a DC-to-DC step-down for the high voltage LIPO packs!
Did you ever have any issues of running your Phoenix for a long time? I’m assuming you are running your hexapods with 645MGs?
Hey fellas,
Just found this thread again after hunting around for info to get my Phoenix walking.
Kare, I’d go for shortening the servo wires - when building my phoenix i mounted each leg to the chassis, tested the max movement range, snipped off the excess wire and replaced the connectors with new ones. It took a little while, but seems to be well worth it for the reduction in cable mess. Everything will be tidied into braided sleeves once i’ve got her walking around.
Now all the parts have arrived, i’ll be powering her (both the SSC and Servos) with a 7.4v 2170mAh Lipoly. The servos get their power via an MPI HD voltage reg.
I am however flailing around at the moment - trying to understand what i need to do to actually get her walking. (hint, hint )
I needed to make sure I had a couple holes to allow mounting a pan and tilt or similar to the chassis. Nothing developed for it yet though.
Hi Tom,
thanks! No I don’t build/assemble machine for a living, but I do work with making custom solutions for disabled person with electric wheelchairs and more. (BTW, electrical wheelchairs are perfect base for some large robots 
)
Yes, I’m using the HS 645MGs.
Hi Zenta,
Cool, you created your robot family, wondering, how much you spent on your robot? I want to created multiple executors, but now still struggling on the first one.
I just released the pre-tested software package. Since you got very good action units, would you like to try add a wifi interface on it? The hardware side you need a few more things:
1. TI DM6437EVM cost around USD$500.
2. RJ45 wirless AP should be not expensive.
3, one mini camear with mic and one small speaker that are not expensive.
These hardware is quite easy to get.
For software side, I can pass you a better version which should be longer enough than the batteries life, and I can do some modification based on your situation.
These days I am working on a document about how to connect and configure the hardware. If you have any interesting about add your robot family with wifi interface and control it from internet, please let me know. I can do something about this.
Please don’t confuse, I am not selling anything, I want my SFIPExe can really running on a robot family.
Hi Zenta,
This is sooo true! You really know how to handle your tools!!
I see that you have found a good solution for the batteries! Since I’ve got my LM phoenix on the way I also need to find a good solution for the battery space. It would be nice if I can also change batteries. If BlackWido needs a new pair I need to disassembly the total body so if the battery dies, I need to hook ‘m up to the charger and I’m out of testing 
I’m still thinking of buying the same NiMH battery pack since lipo batteries are a little bit expensive. (about 4 times the price of a NiMH with the same mAH). What are the specs of the lipo that you are using?
I love to see that!!
Xan
P.S. Don’t forget to mount the botboard
Hi Xan,
I actually found that lipoly’s were not that much more expensive NiMH. I bought 5 2100mAH NiMh tagged cells for about £12 and paid £19 for a flightpower Evo25 7.4v 2170mAh pack. This is one of the better quality, high discharge rate packs - but there are others which are less expensive.
I did however, have to buy a HD voltage reg which bumped up the cost of lipoly use by £35 (reg+intl shipping) 
As far as fit goes - I had to change the chassis posts because the NiMH pack was too long and wide - it now fits but fills the area beneath the SSC and has to be charge in situ.
The Lipoly pack is about the same thickness and lenght, but only half the width - it can be slipped out for charging quite easily. It’s also lighter. Downside is more wire spaggetti - lipoly to ssc/reg/lipoly voltage guard/switches.
Hi Xan,
Thanks! Gongratz for your own Phoenix, cool! What color?
Yeah, I felt it was a real luck that this LiPo fitted so well.
They are not high quality LiPo, but I think they will do the job perfectly. They are a bit over rated too, they “only” manage 65 amps continous draw and some people have tested it to about 3500 mAh. A link: hobbycity.com/hobbycity/store/uh_viewItem.asp?idProduct=6452 (as always you get what you pay for).
No, but I’m thinking to shorten the servo wires first…
Looking forward to see your Phoenix too! 8)
Thanks! I’m very happy with it! I decided to just stick with black. It looks like I’m also working on my little family.
“Only” 8)
I know that you get what you pay for. But at the local hobby shop I pay 119 euro (+/- $175) for a 7,4V 4800mAh. So with this amount of difference, I don’t really care about the quality that much anymore
Offcoure I will keep you all updated on the progress of my phoenix.
Xan
Hey fellas,
I thought i’d skip my usual swim session last night and try to get my phoenix up and running. Unfortunately it turned out to be less straightforward than i’d inagined and i’ve now got a few questions.
I’d be most grateful for any help you can offer - but in return i can only offer a ‘layman’s’ guide to building the phoenix once she’s up and running.
I printed out and read both the SSC and PEP manuals a couple of times over the last few days - to get familiar with what needed to be done.
Fisrt job last night was to connect to the ssc with SEQ, deslect the servos not used, trim the ones i’m using and set the end points.
This is where i encountered my first issue. I’ve built my phoenix to match zenta’s layout - so i could use PEP. This means the body and SSC are 180º (compared to the LM build guide) with the switches at the front and serial at the rear.
With the serial cable connected the rear servo coxa has a VERY limited range of movement - as it is impaired by the connector body. I don’t currently have a bluesmirf so, in order to continue I had to remove the SSC and for the moment have it sitting loose on the top of the chassis.
The next issue i encountered is to do with finding the max/min range of motion for each servo.
Should i be looking for the best case scenario - with the limbs around any particular leg moved so that the maximum range can be used - or should i be looking for the max/min range within the confines of the other legs?
I can read these values in SEQ, but they are given as degrees +/- each side of ‘zero’ the centre point. This doesn’t seem to match the range of values zenta has in PEP.
To be honest i was becoming a little frustrated by now, so i decided to abandon this part of the task until i’d asked a few questions.
I did however have a look at exporting data from PEP and importing it as a project to SEQ. Sorry more questions.
I followed the instructions in the PEP manual. For the purposes of experimentation i used PEP as it comes - ie i did not create any new motions. I opened “Ind.Sekvenser”, clicked eport and then saved this worksheet as a CSV file and closed it.
I moved to SEQ, clicked projects, selected an empty project to export and DESELECTED the PIN and T fields not used with my phoenix. An “export success” message appeared and i had both a .csv and .shp saved. I renamed the .shp file to match the exported .csv file from PEP.
Back in SEQ i went to projects, clicked import and tried to load this csv file. Instead of loading i received an ERROR message telling me that i needed values for “project”, “Sequence” and “Step”. These ‘seem’ to exist in my .csv file so with my limited knowledge/familiarity of what ‘should’ be happening i ground to another halt.
If anyone could offer some guidance to help get her walking I’d be a very happy man.
I’ve attached a few pics of my phoenix this morning.
Hi,
Nice pics of your Phoenix!
Are you just talking about the coxa values or for all (femur and tibia)?
I defined the coxa in PEP the same way as for an inline hexapod. I’m soon gonna send a new version to Jim that support the Lynxmotion SSC32 pin setup and an option for setting the coxa offset (ex +/- 60 deg) for the front and rear leg/coxa. Tha last part for making the SEQ SSC32 setup combatible with Xan’s and hopefully Powerpod setup.
But I’m very soon going for a two weeks vacation so I can’t promise, maybe I send a mail tomorrow.
This sounds correct. But are you sure your .csv file are semicolon delimited? (not comma delimited).
BTW: Have you tested the .csv and .shp files at the project page? :lynxmotion.com/images/html/proj098.htm
Hey Zenta,
Thanks for posting the PEP-LM. My phoenix now has meaningful movement. Thought she’s not quite walking yet.
First job this weekend was to rebuild her to the LM layout.
While doing this i also took time to tidy the cables and shorten the last couple of servo leads. She looks very tidy now.
While testing the fit of my Lipoly and HD reg i found that the SSC-32 mounting posts needed shortening by 5mm to allow the reg to slide under the SSC-32. The board is now as close to the upper chassis plate as i can go without making it impossible to plug ther servos in.
This took around 4 hours of general building and rebuilding.
While the lipoly was on charge I started testing with my original NiMH pack - I had it wired up to upply both VL and VS power. I had hoped to be going wireless with a bluesmirf, but it just didn’t turn up as expected. The initial set-up tests with SEQ turned out to be frustrating. The servos just wouldn’t hold power.
They jerked into position but quickly went limp again - no matter what i tried.
Eventually I decided to try the NiMH pack on VS and a 9V PP3 on VL. This seemed to cure everything.
After all this planking around, I had just enough juice left in the VS pack to try a very simple file exported from PEP.
One thing i’ve found is that my copy of Excel (2003) only has a comma separated export option - so i have to import the files into a text editor and find/change to semicolons.
One other thing i’ve noticed is that nearly every horn and femur bolt seems to be on a mission to unscrew itself at any opportunity. One minute everything is tight, next the legs are flopping around. They’re all going to get a dab of red thread lock.
I’ll post some pics of her tomorrow.
Hi plingboot,
Great to hear she’s walking/moving now!
I’m on vacation now but I’ve been thinking about the SSC32 card orientation so I’m gonna rebuild my Phoenix again to the correct LM layout. Or else it would be harder to share PEP projects.
Yes that is a bit strange that your 2003 does not support semicolon delimited .csv files. All mine (both home and at work) support it. But I remember having that trouble when I started making PEP. At first I used a program called Killink CSV, it’s a simple spreadsheet editor. And you can easily copy and paste from MS Excel to Killink and then just save the file. The Killink “save as” function support comma , tab, pipe and semicolon !! delimited cells.
Maybe running some different office updates may solve this for you?
Looking forward to see them!
Thanks to you both!
Im not 100 % sure this will help you but in my version of Excel (2007) you can open the Control Panel, select Regional and Language options. Under the tab Regional option click on the button Customize and in the new dialogue that is opened you can select the separator in the textfield “List separator”. I hope that helps you!
Thanks Jonas
That sorted it out and will save a lot of faffing around - i’m a mac user with my day job and find the windows experience (associated with owning robots) awkward enough without having to mess around changing commas to semicolons.
Just thought i’d upload a pic and teaser movie link of my phoenix.
The movie is a 360º view of the phoenix, but it’s in apple’s .mov format, so you may well need quicktime to view it.
Hi plingboot!
Wow 
Excellent build! I liked the hi-res teaser movie, showing your build from all directions. Very tidy wire work.
Your yellow/black theme - awesome! 8)
