Make a Gearhead Motor into a Monster Servo!

This topic is a continuation of the one found in customer feedback. The purpose of this topic is to find a solution to building a monster servo using a gearhead motor geared to about 60rpm, servo guts or H-bridge motor controller and other electronic equipment. Your input is welcome.

my thoughts were to modify the guts of a cheap HS-300 Servo to allow 3-4 amps of operation (as also mentioned by Evolution).

I would also like to see if we could come up with a way to control LARGER motors such as harborfreight winches suggested by zoomkat.

Thanks.

I plan to tinker with the Hitec HS-422 servo (below) as the internal parts are connected by wires instead of being directly soldered to the main board. It also comes apart easily for hacking. Middle below seem to be the typical MOSFET type of H-bridge. Bottom shows a relay setup that might be needed for really large motors. I plan on making a Jameco order soon, so I’ll have to see what they have in the way of PNP and NPN
MOSFETs.

lynxmotion.com/images/html/smodh2.htm

geocities.com/fet_h_bridge/
ayershome.org/~eric/robots/h … ndex.shtml

dprg.org/tutorials/1998-04a/

That sounds like a good idea. I also am going to fiddle with the HS-422’s (goofed up and said 300). The idea of enableing the board to run at higher amperage will ultimately allow it to run a larger gearhead motor somewhere between 3-4 amps I would think.

A candidate as a motor for this kind of project would be a 31 RPM 27.78 oz-inch lynxmotion motor: robotshop.ca/home/products/r … motor.html
with appropriate motor mounts and 4mm hubs, Im sure something like this could be used in an arm. The other question is though, can this motor be used entirely to build an arm? Would it have enough power to lift 3 of these motors? Im not sure about that, but my idea also includes the use of these motors in sumo robots so the motors can be controlled directly from an ABB+Atom or other Micro eliminating the use of a motor controller in between.

We sticky posts that become noteworthy, not in hopes it will become noteworthy. :stuck_out_tongue:

I have ordered 100 of the boards from the HS-5745 digital 1/4 scale servo over a year ago and nothing. I am holding the Open Servo project though. I’m seeing if I can get a prototype made soon.

Jim, would it be possible to purchase servo guts directly from somewhere to experiment with? I’m assuming it would be much cheaper than buying a complete servo.

That’s what I’m trying to do. I’ve ordered servo guts sucessfuly in the past, so I don’t know what the problem is. Rest assured they are looking into the problem.

ahhh, I see now. :slight_smile:

First tinkering with the servo. I output one side of the motor to the input of the previously made MOSFET switch. The pulses being sent to the servo motor also controlled the MOSFET to some extent. I need to find an appropriate pulldown resistor for the switch to keep the 12v motor it is controlling from running on its own. Next thing to do is detach the servo motor and do some testing without it being in the circuit.

http://web.comporium.net/~shb/pix/hb1.jpg

Interesting stuff.

Sounds like some progress is being made. Was the motor running at 12v or servo voltage (6v)? Also, you said the Mosfet is being controlled somewhat by the PWM signals, how is it affecting it? I assume you would need some kind of resistor to counter this.

I opened up an HS-645MG today to take a look at the electronics involved in turning the motor. From the looks of the output, the servo controller goes to an H-bridge the consists of 4 smd components (either mosfet or BJTs). I could not see the numbers on them (I think I’m getting old).
What can be done is to remove these and connect the gates or bases, depending on what they are, to an external H-bridge (like the LM298N). This will permit the hook-up of an oversized 12V gearhead motor. The LM298N will require +5V for its logic circuitry and +12V (the motor voltage) to drive the motor.

There is much useful information from openservo.com on the output stage of servos. Your idea to use a 298 is a dead end. It has bipolar transistors in the output stage. They will not help as their turn on resistance is way too high.

Well depending on the motor used you could always use the L293,L298 or you can make your own H-bridge using discrete mosfet devices. The 298 was just an idea because I used it in a recent automation project which involved an 7" LCD screen and a motorized mount for an in car entertainment and navigation system.

Just did some quick testing with interesting results. I disconnected one servo motor lead, then connected the two leads together thru two diodes to the MOSFET switch. With the servo at position 1448 I had the 12v motor in the servo control dead spot. When the servo control position was increased to 1458, the 12v motor was vibrating, at 1468 it was running at ~30% full speed, 1478 at ~60% full speed, and at 1488 pretty much full speed. Also the same results on the low side of1448. I rotated the servo horn slowly and the 12v motor would slow down, come to a stop, then start speeding up again. This indicates to me that one could drive a large MOSFET h-bridge just using an inexpensive servo (even one of the micro ones). A gear (or the motor) could be removed from the servo allowing the horn/pot to turn easily so it could be placed at the joint for position control.

These are some breakthrough results zoomkat! Can we assume you will experiment with the pot at a joint and see if the motor can be controlled like a servo on a joint?

Edit: it would be easy if you had a pot like this: lynxmotion.com/images/howto/monster.jpg

I don’t know where it came from though, it dosne’t look like it came from a servo, it looks like it came from a tapedrive or somesort.

One could always remove the pot from the case and shrink wrap the board to save space.

I’m curious if it will handle 12V at load though. Or if it would handle 6V at 3 or 4 amps?

No need to get another pot, just use the one that is in the servo. If the motor is removed along with the gears, then the servo housing can be trimmed down so that it is just holding the pot. The housing has mounting tabs on the top surface, and the servo horn makes a very universal way to attach the pot to another surface. The servo board could be located remotely with the h-bridge and connected to the pot at the joint with three wires.

I understand what you mean, I just don’t have a visual picture in my mind yet.

“I’m curious if it will handle 12V at load though. Or if it would handle 6V at 3 or 4 amps?”

The MOSFET I used in the switch is a Radio Shack IRF510 which is rated at 20 watts, which isn’t a lot, maybe 2a at 12v, or 4a at 6v. I looked in the jameco catalog, but didn’t see any PNP MOSFETs that looked useable. The below bgmicro catalog has some inexpensive MOSFETs, one of which is a 60a PNP for $1.49. Maybe somebody that has a purchased h-bridge can look and see what MOSFETs it uses.

docs.bgmicro.com/pdf/lma.pdf

12V at 2amps and 6V at 4 amps is incredible. Much better than the board would handle on it’s own.

I think that in itself is a great accomplishment.