Hi, this is a first post and seeing how the great expertise available here is vast, I am hoping that I might receive comments on the feasibility of upgrading the lifting capability of the A4WD1 Gripper. The recently completed A4WD Rover platform performs as expected, current pan video camera and Tx upgrades are in progress from the current ones shown below. The only somewhat disappointing performance is that of the gripper lift force capacity. Using calibrated scale weight it seems as if the maximum weight it is capable of lifting is ~ 100gms or so. Does anyone have experience with this gripper and is this about what one could expect from it. Also are there higher torque servos that would fit the servo bracket and that could improve thie weight lifting capacity?
Thanks
I don’t know your configuration, do you just have jaw and rotate servos (what size?) on the front of the 'bot? I don’t see any “lift” servo associated with the jaws.
'422 servos could be replaced with '475 or even '645, they should be all the same size.
Alan KM6VV
Thanks for the quick reply. The std. A4WD gripper, as supplied by LM, does use the Hitec 422 servos. The unit utilizes 3 of those, 1 for lift=Y axis rotation(closest to the Gripper- assembly body to rover body attachment end) , another for a z axis yaw=twist/rotate, and the 3rd for open/close of the gripper jaws mechanism.
I was curious if the lifting capability (loaded y axis vertical rotation) I observed was typical, I have checked the Y servo for undue friction or binding and haven’t detected any but after all it is lifting the two other servos, brackets and jaw mechanism so the remaining torque left for “payload” is pretty slim. I think LM should maybe supply a beefier servo for this axis in their “kit” as the other two 422’s seem more suitable for their less demanding functions.
For a more basic question, does “Std. Servo” description mean equivalent attachment/dimensions footprints? Even among these the dimensions seem to have slight variations.
The “standard” size R/C servos are mostly interchangeable. Note however that the new Hitec 485 servo that replaces the 475 appears to be slightly bigger, as has been commented on by Jim on this forum. I just got some new '485 servos, and as I have the jaws, older '475 servos, and also the universal bracket, if I get a chance, I’ll see how they all fit. Or maybe Jim will verify/clarify. ;>)
The “pitch” axis (I wasn’t thinking of this joint as a vertical axis) could probably benefit from a larger servo. Roll and yaw are probably fine.
Also, watch out for some confusion on the forum, many of us would rather think of the vertical axis as a “Z”, as opposed to references in early code calling it the ‘Y’. Someone’s convention, I suppose. I’m prone to calling vertical Z from my CAD/CAM, aircraft, ship and engineering experience. ‘Y’ might be from an illustrator’s usage, I could never verify.
Alan KM6VV
I don’t know exactly how you could do this but you could attach a spring to the arm and then to something else like any of the AL5 series arms do to act as a counter weight.
Trouble is, this is a “pitch” axis for the jaw, and we rotate the jaw as well. Hard to attach a spring.
I hate springs! I know they are used a few places in the LM arms. I think I’d rather see a counterbalance, like on the skyscraper cranes. You don’t see any big springs there!
Alan KM6VV
Probably need to check that the servo is getting ~6v for max power. If more lift capablity is needed then there is the possibility of ganging two servos together, or making a custom bracket to hold a big servo, or other heavy lift solutions.
I see how a spring wound’t work,bad idea. 
Thanks all for the suggestions. I must admit I had been thinking along zoomkat’s ideas if a direct replacement servo didn’t do the trick. I have had some success with spring tensioning devices. The view below shows the business end of a drive train for a simple vertical climb robot Iv’e put together to demonstrate the “Space Elevator” concept in space demo presentations to kids. The two wheels, clamp over a vertical rubber belt that allows the climber to simulate an orbital cable stretching up from the ground.
In the case of the gripper though I cant see it as a very practical solution.
Neat! 
(I love the concept of space elavators.) But I don’t see any solution for your robot though. 
Interesting climbing device! Spring loading a pinch roller is a reasonable solution.
If you really want some vertical travel, I’d suggest a short arm, and either another servo or a pantograph linkage.
Alan KM6VV
You say the gripper can only lift 100g. That doesn’t sound right to me. How are you determining this limit. What happens with more weight?
How are you powering these servos? Can you post images of the wiring? I believe you may be powering them from the Bot Boards 5vdc regulator…
This rover is not microprocessor controlled, all components were purchased from LM as the R/C kit version, Invoice ID 27948, so no on-board 5V regulators here. The power supply for the servos is from a 6V 2800 mAH NiMH pack directly connected to rcvr pwr in and is only powering the R/C recvr. A separate 12v pack independent circuit powers the Sabertooth PWM motor controller. All the arm servos are powered from their respective Hitec R/C receiver channel servo outputs. The BEC power leads from the Sabertooth controller BEC circuit to the R/C rcvr. as recommended in(A4WD1 R/C Stick Radio Control Tutorial Rev. 2. step 2.) instructions, were bypassed/disabled. The gripper lifting capacity is being measured with freshly charged packs by gripping calibrated balance/scales weights. There is no problem with gripping the jaws to a 100gm weight and holding it with the available gripper tension. On attempt to significantly raise the gripper with the 100gm weight, servo stutters and stalls but does not raise to full height. A 50 gm weight, last view, is OK.
Does anyone have any idea what others that have built the gripper, have experienced or measured for the lift capacity with the HS422 servos as supplied by LM?
To see gripper in action from a camera remote feed lifting a small rock see:
vimeo.com/8202751
Since there seemed some question as to voltages to servos, I decided to check input voltage at the gripper Z axis servo and with a fresh pack charge 6.3V was present. Since I had the cover off to check the voltage at the circuit board, decided to go ahead and look in gear compartment. Here I found several teeth in some gears rounded off. This likely accounts for stutter, lockup and slippage at the “higher” loads over 50gm.
Would be a nice suggestion in the future from the BB for similar problems posted.
I will replace the damaged gear sets and remeasure performance but am surprised that with relatively low weight lifting attempts, unsuccessful phillip screwdriver or cell phone lifting, gear damage seems to have occured.
IMHO, a more robust metal geared servo should be the better choice in this gripper position instead of the supplied HS-422. Very likely we will replace this servo with say the HS-645 for hopefully a more robust performance.
Thanks all for your thoughs
Damaged gears would affect the lifting weight. 
Not sure what you are saying here. Can you reword it?
It’s a cost thing. Most builders are not interested in paying top dollar for the best that is available. The gripper kit was designed as a cost effective manipulator add-on to the rover. Of course you could have purchased SES brackets (ball bearings) and digital servos for a really powerful gripper. At least the problem is easy to fix.
Daneel Olivaw wrote:
Would be a nice suggestion in the future from the BB for similar problems posted.
[Not sure what you are saying here. Can you reword it? ]
As a new forum member and on a learning curve for this very enjoyable “hobby” all responses to our posting have been welcome and have provided food for thought.
I am merely suggesting that in the future, if someone else also posts with a similar set of hardware symptoms as I did, that the bulletin board forum members might also provide the possibility of servo gearing failure as a diagnostic check to weak servo torque such as we have seen.
[Of course you could have purchased SES brackets (ball bearings) and digital servos for a really powerful gripper.]
Overall we are happy with the A4WD1 kit and gripper. We knew it was not a commercial robotic platform but still, we were surprised that with only attempts at using the gripper to lift common everyday light objects, not weighing multiple pounds but such things as a screwdriver(3oz) barely, or cell phone(5oz) not at all, and then determining that it was not able to do so and that servo gears also might have been damaged. We will have a better grasp on this after the current gears are replaced and will post our results.
Perhaps other users have applications were the gripper is more of a movable but very light duty decorative add-on to simulate functionality. At a slight additional cost to the kit ($15-$20) with, as we think, a z-axis metal gear servo such as the HS-645, this would be a worthwhile modification or a kit option likely to significantly enhance gripper robustness and functionality.
Ah, Ok. That’s the way forums normally work eh.
We will discuss this in house. Maybe we should upgrade that servo with the 645. The truth is I haven’t received any other reports of this servo going bad in this kit. It’s possible that the servo was defective from the start, or the servo was damaged not in use, but if the rover was bumped into an object. I’m glad you are generally happy with the kit. Last thing, a non broken HS-422 should be able to lift 65% of it’s stall torque. At 3.5" the math is, (57 ounce inch x .65) / 3.5" = 10.6 ounces. It should easily handle the 3 and 5 ounce items. 8)
Your calculations are pretty much what we had also theoreticaly calculated could be possible from the HS422 gripper. Its a bit more difficult to judge however the actual gripper capacity since subtracted from the 10.6 oz one has two other gripper HS422 servos, the wrist twist and the jaw close servos each 1.6 oz wt each, so + 3.2oz, plus brackets, jaw mechanism, fastner hardware and any frictional losses all being lifted as an inherently present load on that z-axis servo.
OK, I think now that I am fully satisfied with the final word on the Lynxmotion Gripper. The new sets of gears came in and were used to replace damaged gears in the (z-axis) HS-422 Gripper servo. (see below for image of the relevant gears). Before any weight lifting efforts an immediate change was noted. In the beginning, from the very first initial gripper assembly tests, the arm had always had a kind of jerky traverse during servo actuation in the lifting direction even before attempting any object grasping or lifting. This time, with the repaired servo gears it now has a fluid smooth motion from bottom to the top of the z-axis servo motion.
That was the good news.
The so-so news is that even with the very much smoother motion, the Gripper claw lifting capacity has only improved to a max of ~ 120gms, screwdrivers yes. cell phones no. The single 100gm weight is now capable of being lifted but from about halfway up the lift distance, the motion greatly slows down but eventually finally reaches the upper limit. Piling on an additional 20gm weight for a total of 120gms, the arm can’t complete the full lifting range stalling ~1/3 up or so.
In conclusion, I think the gear problem was either a pre-damaged lemon servo, as the arm always had a jerky motion to it from the initial assembly, or less likely, gears were damaged in some way during assembly or lifting attempts (I don’t think so). Since motion now seems significantly improved but weight lifting capacity is nearly the same, what we see is what we’ve got!
A HS-645 used in the z-axis position will likely help and if will likely double Gripper lifting capacity and also provide gears less likely to wind up like those below.
