What kind of servos are you using. If your using 645’s they should be strong enough to move the bot. Are your batteries charged well? is the weight of the robot evenly distributed? These factors can cause all sorts of problems when having to move the bot. The problem shouldn’t be that its too heavy for the servos, but more of how your programming the SEQ’s of the bot. 
maybe i’m just being a bit too paranoid about burning them out or damaging them in some way, i’ve yet to break one so i don’t really know how robust they are i guess.
When i try and program a walk, i’m moving from one position to another, and there’s quite a lot of play between the loaded and unloaded positions which makes me worry the servos are under too much pressure, they get quite loud too, or is that normal?
they say that they are hitec HS-5645MG’s.
Your problem is that it is WAY to top heavy. You need to bring the Wireless device and the SSC-32 down much lower. Try mounting the ssc-32 so that the DB9 connector is faccing down (if nothing gets in the way)rather than up so the wireless device can mount with the antenna pointing down vs. pointing up.
It looks like there is also some kind of gray back plate in the back also? Another improvement is to trash the lexan feet and get non flexing aluminum ones. First you need to lower the electronics and that should be enough and you can probably keep the lexan feet and upgrade to aluminum later.
With these changes, you should have much better results.
In this photo, it shows the SSC-32 mounted midway on the back where you have yours mounted 100% above the body deck!
yes, the grey back plate is there from the previous owner who mounted some pretty fancy electronics on it I believe.
I was hoping that as the weight of the wireless device was small enough that it wouldn’t really make much difference where it went, from what you say i as wrong there, I’ll get rid of that backplate as well.
I think there are some foot brackets in the kit also, so i’ll see about swapping those in, everytime i think i am getting close i realise how much harder this walking buisness really is 
No need to worry, the 645’s can handle the Scout without being damaged, but like SN96 said, it too top heavy (Missed saying this in my last post 
). You need to find an alternative to mounting this a little lower. Also, yes the noise from the 645’s is quite annoying and can become ear-shattering. I don’t know this from experience (though I have many 645’s) but other’s on the forum say the same things. I geuss I don’t put too much load on MY servos 
It might be light but its the leverage it has being so high up in the air. A little sway can become greatly amplified with that configuration.
Are you using a sequencer?
I’m using some software that captures the positions of the servos with a series of sliders, and builds up a move from a series of those positions.
-there’s a link in my sig to the sourceforge download.
I’m not sure if that’s a sequencer or not?
It seemed like a good idea at the time, but i’m left wondering if the real way to go is to be doing mathematical calculations involving the lengths of the parts of the legs and such, kinematics i suppose, and i wouldn’t know where to begin there.
my thoughts exactly were to lower the CG… I found that even with the BRAT having the RF part and antenna way up in the air and especially the antenna off center made it damned hard to do any type of symetrical walking sequence. my scout has the ssc-32 hung on the same bracket but down low as in the old pictures. I put the components and servo connections facing forwardand then stacked a matchboard pcb with the components facing out the back. I hacked up a chunk of .031" g10 to make a small antenna mount that hangs off of the d-sub jackscrews and so the antenna is as low as it can be and on center. I offset the extra weight of the matchboard by registering the battery forward in between the top/bottom plates and it is held in place using a strip of velco I got from a craft store. the last observation I’ll make about the scout is don’t rely on the push pins for servo attachement and rahter use the 3mm screw mounting hardware if possible. I found my scout to be wobbly and not very stable (despite using hs-5645 servos) until I changed these over and even then I had to go back and re-tighten after a couple hours of trying to get offsets aligned in the SEQ. I made a slight error when I asked the person I bought my servos from to open then up to 180 degrees, which seemed like a great idea at the time, because the canned walk program available has the servos configured differently. 
I read in the manual for the ssc32 that it used commands like “#1 P1500 T1000” so I wrote my software to send those, i’m not sure what to do with that code you linked to SN96, is it for the lynxmotion sequencer?
I’ve just tried reversing the mount for the ssc32, and the wireless device’s aerial is catching the knee joint now, this might be a bit tricky, it’s pretty long.
I was planning on getting it walking before spending anymore, but getting a matchport on it would be the next step anyway, and would solve my balance problems now.
As to the plastic push pins, i do indeed have these on the servo case mountings, but not the horns, is that still worth upgrading do you think?
The other thing that i saw was mounting the battery packs on the outside of the legs, i couldn’t find the post to link to right now, but someone here did it already i’m sure, it would lower the CoG, and maybe even counterbalance a bit at the same time.
Yes, I was also going to mention that. If you want, you can make your own packs with some rechargable AA batteries (or take apart the 6V pack) and make two packs that attach to the legs to spread the weight more evenly.
Bob, perhaps for now you can remove the wirless device and just focus on the walking program for now. There are other bluetooth solutions that are shorter for later on.
If I were you, I would mount the ssc-32 and the bot board (if you have one) using the standard mounting configuration like I posted above. If all you have is the ssc-32, then plug the serial cable directly into the ssc-32 connector and program it using this setup.
Here is a video showing Jims scout taking steps. Might want to play this a few times to see how the joints are moving to get you started.
thanks for all the feedback, i do only have the ssc32 for now, it was originally plugged into my serial port via a cable, but the weight and pull of that were so huge that i sorted out the wireless link you see in the first pic, it’s basically just a wireless serial cable using bluetooth, nothing fancy, but i thought it would be usefull for future testing and such as well, for now i am doing all my processing on a pc.
Is there a setup whereby one could use a serial cable and not suffer from the drag it creates? -maybe pointing the ssc32’s serial port downwards & trailing the cable, or possibly suspending it from an elastic, in the style of those mouse bungees, every arrangement i tried with the cable really threw the balance off badly.
That vid is the best one of a scout walk that i have found so far, i’ve watched it many times, and the bit that i’m missing is where it is able to shift it’s weight from balancing on one leg, to moving the other one forward through the air, no matter how many times i re watch it, it still looks like it should fall over to one side once it lifts a leg, no matter how strong that ankle servo is, mine was leaning much much further to one side to be able to balance on one leg. Is that a tether sticking out of the top in the video, possibly taking some of it’s weight?
I tried inverting the ssc32 last night, but the antenna was bashing the back of the knee, so i have just lowered the whole assembly, removing the grey board and reversing all my wiring. It looks a lot lower now.
A lot of the tethered setups are actually suspending the cable from above and are not using a standard serial cable but rather 3 wires bundled/twisted or modular telephone wire soldered to a DB-9M connector. For the brief time I was tethering stuff I did the telephone wire+db-9 thing and used my camera tripod+yardstick+tiewraps to to suspend the cable up over the BRAT.
Be careful about jumping into the matchport solution before thinking it out. “matchboard” is a breakout PCB I did for the matchport a few months ago and the files to make the board are posted on the board somewhere. here is a pic of the matchboard+ssc-32+antenna on the back of my scout before I had it all harnessed up nice and neat like.
http://users.adelphia.net/~eballou/MatchBoard/2337.JPG
while the matchport has excellent speed and range at a midpoint price it is more costly and complicated than several of the bluetooth module implementations also documented in various locations on the board. to the best of my knowledge the cheap bluetooth options do not have 300m range at 230kbps but, uh, is that actually necessary is a valid question to ask as well. Anyway, I am sure even mentioning this subject will bring a few of the regulars in here pitching their solutions as well so hang on and, uh, wait for it. 
8)
In the video Jim has a suspended serial cable. The cable is not holding the scout up, only the weight of the cable.
It sounds like your servos are not working together to create the proper movments. In other words, since you say that your scout leans so much further than Jims, this tells me that perhaps you are only moving the ankle servos when you shold be moving the hip servos as well. The key is to keep the bot with in its stable range by keeping the mass level where the top plane of the bot is near level.
I know this is easier said than done but what may help is studdy how you take your first step taking note of what joints had to move and in what direction. Try not to fall over when doing this. 
I think we should all take a moment to PM Mike and encourage him to open up his creditcard to get some personal hand-on time working with his own BRAT. 8)
In time…
Then you will show you how its done
I got the CoG lower, but the battery died mid-testing, seemed to go pretty quick this time
Here’s a pic of the assembly lowered anyway, do you think that’s far enough?
otherwise i thought i would try and make one of those funky looking U-bend ones like Eddie posted in his pic, it seems like the only way to move the weight down without getting in the way of the knee’s range of motion.
I cringe when I see a bot connected to a computer using a standard serial cable. The first thing one should do when they get a soldering iron would be to make a thin light weight cable for serial communications. The thin light weight three conductor wire used for walkman type headphones could be used if desired.
Bob, I think it looks a lot better. I would try to work with what you have for now and see how it goes. Your last setup was definetly a bad configuration. The two changes you made should be an improvement.