does this seem right?

[Gryd3]

According to the page at http://www.servodatabase.com/servo/towerpro/mg995 that has a motor built in. And I guess the connections to it would be power and control inputs. Not outputs that you can use to control an external H-bridge. What am I missing?

I don't know what you mean.

I believe the Op has taken apart the servo to use a much larger DC motor and is using the control circuitry inside to take care of positioning the larger motor.
Can he confirm?

 

[KrisBlueNZ]

Ah, I see. In that case you need a bit of logic to convert the two motor control outputs to a direction and an enable.

It would be interesting to know whether the +2.5V he measures on each motor control wire is DC or a PWM signal. OP, do you have an oscilloscope? Can you look at the signals on the two motor control wires?

 

[gingerneering 101]

I want to open the servo cut the wires going to the motor and use those two wires to drive a larger motor.this may explain http://letsmakerobots.com/node/34692

 

[KrisBlueNZ]

Right, OK. Assuming the controller drives a motor with PWM, you can convert the two motor control signals to a direction and an enable signal using a bit of logic. Here is the translation you need:

MOTORA    MOTORB    DIRECTION    ENABLE
  L         L          any        OFF
  L         H          H          ON
  H         L          L          ON
  H         H          any        OFF

This translates to some very simple logic:

ENABLE = MOTORA XOR MOTORB
DIRECTION = MOTORB

You can use a CD4070 or a 74HC86 to do the exclusive OR function.

 

[gingerneering 101]

I am just becoming more confused as we go.

 

[Gryd3]

I am just becoming more confused as we go.

Hehe, lets step back for a second then.

When you crack open a servo, the voltage levels sent to the original motor 'should' be processed and then sent to an H-bridge driver for your bigger motor.
What Kris is trying to find out to build a proper circuit is if the wires that go to the original motor are 'pulsed' to reach about 2.5V, or if they are held there constantly.

 

[gingerneering 101]

I don't know I thought all digital servos used a pwm.

 

[Gryd3]

I don't know I thought all digital servos used a pwm.

At the input, absolutely. The tricky part that is not completely understood for the servo you have though, is how it controls the motor directly.

Remember that because your cracking this thing apart, you have your original input signal to the servo control board and the output to the original motor to look at. It's the control board's method of controlling the motor itself that we are looking at.
We could operate on the assumption that in it's:
neutral state the motorA and motorB wires are at 2.5V.
forward state the motorA wire is a higher voltage than motorB wire
reverse state the motorA wire is a lower voltage than motorB wire.
but the scary word assume makes an ass-u-me..

This is the logic we need to convert to send to an H-bridge. The diagram you posted has it's flaws. It may work well enough, but there is concern of the longevity of the mosfets. Instead of looking at the diagram, making assumptions and correction the mistake. It seems as though Kris has put some thought into a better solution which I loosely follow, but not enough to finish up.

 

[gingerneering 101]

Is there a way to test without an oscilloscope. Like disconnect the motor from the geartrain so if I turn the pot slightly so the motor is in low speed and use a voltmeter to see it the voltage is still just 5v or if it is lower. If its slow and at 5v wouldn't that mean its using pwm. I have no access to an oscilloscope.

 

[KrisBlueNZ]

Here's my suggestion using the TC4469. No interface logic is needed.



The TC4469 contains four buffer/inverters with high-current outputs designed to directly drive the gates of power MOSFETs. The MOSFETs provide one stage of inversion, so M1, the high-power motor, must be connected the opposite way from the original.

There may be better alternatives to the TC4469. The following options are all 8-pin devices with two channels each; two would be needed. But they may have better specifications than the TC4469. You can look into these options if you want. http://www.digikey.com/product-sear...nSort=1000007&stock=1&quantity=1&pageSize=250

R1~4 and D1~4 provide simple cross-conduction protection by delaying the turn-ON of the MOSFETs without delaying their turn-OFF. D5~8 clamp the motor drive outputs during the short time that neither MOSFET is ON. The 1N5819 is rated for 1A continuous and should be fine for motors that draw up to about 3A, and the MBR340 is rated for 3A continuous and should be fine for motors that draw up to about 10A.

The MOSFET types given here are taken from the schematic in post #1 and are not necessarily what I would recommend.

CD is a decoupling capacitor and should be connected across the VDD and GND pins of U1 as close as possible to the device.

More decoupling capacitance should be connected between the +12V rail and the 0V rail near to the output MOSFETs.

 

[gingerneering 101]

Thank you soo much. If it wouldn't have the craziest shipping I would send you a beer. When I said my wife I meant fiance as we are getting married in 26 days and this was a side project to keep my sanity during all this planning and prep. I appreciate your patience and trying to teach the way it works. I swear I'm not an idiot Im just more mechanical than electrical, which I'm finding more interest in learning.

 

[KrisBlueNZ]

No problem Congratulations on your impending marriage!

 

[Arouse1973]

This is why I keep coming back. Kris that is supa dupa.
Adam

 

[gingerneering 101]

I work with a guy that does a lot of storm chasing and I guess this peaked his interest. He wants one now so he can mount his big HD camera to it on the roof of his truck.