My biped would “jolt” or look like it’s having a seizure.
I have also tried clicking the “All=1500” button in the Lynxterm and my biped still displays the same phenomenon (I’m assuming that button is the same as manually typing in the "#0 P1500 … #31 P1500 "
I can also make each servo move individually by selecting the channel on the top right corner of the Lynxterm and using the slider to position the servo on that particular channel…
I need to have all 12 servos move in conjunction in the “default” state, which is what I’m defining as P1500 for all servos (to make it stand straight up). What am I doing wrong?
Trying to move 12 servos at once would probably need a large power supply. What are you using to power the servos? You may need to move them in smaller groups to limit the amount of current being drawn.
I’m using 4 2500mAH NiMH AA-cells hooked up in series (via a battery caddy). Since they are rechargeable, the resultant voltage across the terminals would be 4.8V.
I have a bunch of 6-cell Sub-C 3300mAH packs (7.2V) from my R/C car days, I wanted to avoid using them since they are quite heavy…
I’m also in the process of changing out the wires, I don’t think the wimpy 24AWG that came with the SSC-32 kit can handle all that current for driving 12 servos…
Assuming you are using windows xp and have the comport in the windows control panel set to the desired baud rate, you might be able to experiment with time delays by running batch files like below. I have a different type of servo controller, but the below code may work with the ssc-32. The “3” in the ping line gives a 2 second delay between sending the bytes to the com port. Copy the below, paste in notepad, and save on the desktop as servo.bat, then double click it to run it.
Basically, your servos are drawing so much current, so quickly that the batteries can’t supply it.
When this happens, the voltage drops quickly.
If you’ve got VS=VL enabled, when it drops under the minimum operating range for the SSC-32 (about 5V or so), the controller will reset.
When it drops under the minimum operating range for the servos (4V or so), they’ll just stop working.
A few choices for ya:
Definitely disable VS=VL and power your logic (VL) off of a seperate 9V battery, if you haven’t already been.
That should stop the SSC-32 from resetting, which is usually the biggest issue.
Try throwing a huge (1000uF or so) reserve capacitor across one of your free VS inputs.
This should help with the servos drawing too much current, since the cap will store free charge when the robot is on but not drawing too much current (i.e at powerup, before you send commands).
Add another cell.
I’m using 5 cells (7.2V) for my biped, and it’s worked rather well, so far.
Of course, I’m using 5645’s, which I’ve found to be quite reliable, even when over-volted a bit.
645’s might not like the extra volties.
If you have provisions on the board to seperate the power supplies for the controller and the servos, then you may be able to use the same batterys for both and seperate the servos from the board chip via a diode. I use a setup like this with a capacitor down stream of the diode to have a reserve power supply for the servo controller which won’t be affected by temporary power dips caused by the servo motors. If the normal battery voltage is low, then a low voltagedrop diode would be needed.
Thanks for the tips Nick. 5-Cells makes the pack a 6.0V not a 7.2V (I’m assuming 1.2V per cell).
The H-645s do pretty well with a 7.2V pack, I"m using a 6-cell 3300mAH NimH matched pack that I use in my R/C car. It powers the servos as if it was a drill seargent yelling “attention” to my servos. However, the knees still do get hot.
I’m still contemplating on converting the biped scout into a biped humanoid. I think the modification would be quite simple, just have to recenter the servos, align all the brackets so that it would be straight, and re-mount the torso so that the head isn’t pointing forward but rather pointing upwards. Maybe buying a humanoid torso to begin the conversion wouldn’t hurt either
But I want to make the scout work with the reverse-knee scheme. I’m going to try some load springs and also more powerful servos (heatsunk) before i go to the humanoid-style bot. Ever since I played Mechwarrior games, I always loved the chicken-leg (reverse knee) scheme of walking, as it tends to keep the COG a little lower as compared to the a humanoid scheme with the same length of legs.
