High Power Servo Hack

[Steve Tyng]

Here's an interesting article on modifying a standard servo to drive
an external gearmotor. Basically the geartrain is removed from the
servo which is then used as a position sensor for the gearmotor. The
servo electronics are used to drive an offboard h-bridge.

http://www.fieroaddiction.com/servo.html

An h-bridge that would work well with this setup (the Tecel D200
referenced doesn't appear to be available anymore).

http://www.robotpower.com/products/simple-h_info.html


Steve Tyng

[Paul Hilton]

Steve, you must have been reading my mind. I was just thinking about a
solution for the BT-7's steering. Have you tried this? Seems like a good
choice for large wheeled vehicles!

Paul H.

[pixelFiend]

I've been wanting to do this as well for a project. Every time I look into it however, I keep running up against the Tecel website, which turns up an "access forbidden" message on the D200. I make a note to call or look into it further, then other things come up.

I have several motors I could use for this— does anyone have any info on Tecel? Are they still a going concern, or have you bought from them?

PF

--
PF


astromechbuilder.com
astromechfactory.com

[Steve Tyng]

Paul,

I have not personally tried this yet but I see no reason why it will
not work well. Some possible gearmotors for steering our armoured
cars could be:

http://surpluscenter.com/item.asp?UID=2009022709083981&item=5-1634&catname=electric
http://surpluscenter.com/item.asp?UID=2009022709083981&item=5-1664&catname=electric
http://surpluscenter.com/item.asp?UID=2009022709083981&item=5-1257&catname=electric
http://surpluscenter.com/item.asp?UID=2009022709083981&item=5-1685&catname=electric
http://surpluscenter.com/item.asp?UID=2009022709083981&item=5-1693&catname=electric


PF,

The Tecel board seams to be no longer available that's why I listed
the Robot-Power alternative. I've spoken with the owner of Robot
Power and he stated that his Simple H-Bridge is being used to make
high power servos by others.


Steve Tyng

[pixelFiend]

Ah, thanks Steve. I'm sorry, I hadn't read through the bottom of your message.  I'll check them out.

PF

--
PF


astromechbuilder.com
astromechfactory.com

[Chrysanthos Kanellopoulos]

Lots of voltage but they are inexpensive.  I don't know if they spin indefinitely or if they are "programmed" to rotate only 60 or 90 degrees left and right to suit a steer system, or what does the trick.

Chrys

[Doug Conn]

I think that having a powerful steering servo is only part of the solution. You also need a mechanism to handle the jolts and bumps without transmitting the shock to the servo (or gearmotor). I finally found something that worked pretty well for RocketMan. Check out this out RC Tank Combat post.

 

http://www.rctankcombat.com/archive/2007-08/msg00415.html

 

-        Doug

<BR

[Paul Hilton]

Thanks Steve,

I have a pair of these:

http://www.allelectronics.com/make-a-store/item/DCM-316/12VDC-SUNROOF-MOTOR/1.html

that I may be able to adapt, the shafts are a bit on the short side but I
may be able to stick a slotted arm on. If not, I'll look at the 24V Mabuchi
gearmotor you have listed.

I also wonder if there might be a cheaper alternative to the motor
controller you mention. One might be:

http://www.pololu.com/catalog/product/755

I'm sure that if I look hard enough, I could find others.

Paul H.
----- Original Message -----
From: "Steve Tyng" <Stev...@gmail.com>
To: "R/C Tank Combat" <rctank...@googlegroups.com>

Sent: Friday, February 27, 2009 12:54 PM
Subject: [TANKS] Re: High Power Servo Hack


>

[Steve Tyng]

Doug wrote:

> I think that having a powerful steering servo is only part of the solution.
> You also need a mechanism to handle the jolts and bumps without transmitting
> the shock to the servo (or gearmotor).

Agreed. A thought I've just had is that since the system we are
discussing involves a position sensor separate from the gearmotor,
there is no overriding reason to attach the sensor directly to the
motor output. The sensor could be connected to the steering assembly
directly instead. The gearmotor could then be attached to the
steering via a slip clutch assembly. Any shocks to the steering would
be absorbed in the slip clutch thus avoiding damage to the gearmotor.
On the sensor mounting, one would need to ensure the travel of the
steering assembly does not exceed the throw of the sensor which would
damage it.

Steve Tyng

[Steve Tyng]

Paul,

The Pololu board would not work to "amplify" the current capability of
a standard r/c servo. The Pololu board is a bit to intelligent in
that it has it's own processing capability and determines motor speed
and direction via separate PWM and direction signals. Though the
Pololu board has an h-bridge the board itself is not controlled like a
typical h-bridge circuit. It is meant to be controlled be a
microprocessor and not as something as crude as a hacked hobby servo.

The hack we are discussing replaces the internal and very basic h-
bridge circuit of the servo with an external higher capacity circuit.
This h-bridge circuit requires a PWM speed signal on one of either two
inputs unlike the Pololu board which has only one input for the PWM
speed signal and a second to determine direction.

Simple h-bridge boards a few and far between. The Tecel was one,
RobotPower simple h-bridge is the only other commercial h-bridge that
I can recall offhand. Most servo hacks off this type I've read about
usually involves making a custom h-bridge board, which actually isn't
very difficult to do. Here's a link for one such project.

http://130.94.182.150/mowers.htm (scroll down to "CONVERTING STANDARD
R/C SERVOS INTO MONSTER SERVOS")


Steve Tyng

[Paul Hilton]

OK Steve, sometimes I'm a bit 'thick" when it comes to electronics. I have
zero experience with H-bridges and MOSFETs in general. Are you saying that
this will do the job?

http://130.94.182.150/servo.jpg

Seems vastly overpowered for my app, but I'll give it a go. I assume that
the ".2" beside the capacitor is .2 micro farads. I also wonder about
substitutions for the MOSFETs. If I used smaller MOSFETs, would it all
still work? What about 2 N-channel and 2 P-channel, rather than 2 and 4?
And what of voltage, is this not an issue? Seems kind of important to me.
And why build a 38 amp H-bridge, then call out "three good gearmotors for
steering servos" none of which draws more than 750 mA, like these guys have
done? Seems like a bit of overkill.

Paul H.
----- Original Message -----
From: "Steve Tyng" <Stev...@gmail.com>
To: "R/C Tank Combat" <rctank...@googlegroups.com>
Sent: Saturday, February 28, 2009 11:15 AM
Subject: [TANKS] Re: High Power Servo Hack


>

[Steve Tyng]

Paul,

Your schematic is a basic h-bridge and would work fine. One note is
that it has parallel 19amp p-channel mosfets on the high side (to
somewhat match the 50amp n-channel mosfets on the low side). You
could remove the extra p-channel mosfets as long as you stay within
the current ratings of the p-channel mosfets. About the current
ratings of the components. At first glance it would seem these
components are vastly overrated for the motors we want to drive. In
fact they may be underrated. The stated current ratings for mosfets
are usually the max rating for a period measured in milliseconds and
the manufacturer is assuming that's with sufficient heat sinking
installed. If you dig into the spec sheets of these componets you
will find the "real world" current ratings are always much less than
the advertised ratings. You will need to build the circuit to handle
the stall current of the motor you will be using. On the voltage, the
components you will use are marked with a maximum voltage. As long as
you stay under your lowest rated component you should be fine.

Mosfets are very sensitive to static and are easy to destroy just by
handling them improperly during assembly so order extra. You can also
assemble an h-bridge using dc solid state relays (SSR's) simulare to
what Joe Summer did with his controler on his Hetzer.

http://rctankcombat.com/articles/speed-control/084Hbridge.jpg

SSR' are nice because they are designed to be abused. Instead of
using the big Crydom SSR's you could use smaller pc mount SSR's like
these:

http://www.alltronics.com/cgi-bin/item/24B003/search/Grayhill-70M-ODC5-Relay

The only issue with the SSR's is that they weren't really designed
with high speed switching operations in mind so may not be able switch
at the frequency that the hacked servo is switching at. We know the
Crydom D1D40's will switch reliably to a certain frequency but that
was proved by buying
and testing them ourselves. If you want to try the Grayhill SSR's
listed above you'll be blazing a new path for the hobby.

I'm all about reliability in my combat vehicles. To me it isn't worth
saving a few bucks on a critical component just to have it fail on one
of the two weekends a year I really need it to work. So my preference
if I were to build this hack would be the commercial h-bridge, an SSR
h-bridge, and finally the homebuilt mosfet h-bridge. I'd also acquire/
build a second as a spare.


Steve Tyng






Concerning your voltage

> >http://130.94.182.150/mowers.htm(scroll down to "CONVERTING STANDARD

[Paul Hilton]

Steve,

The SSR option looks quite interesting. How would this be coupled with the
hacked servo? I've only seen schematics for H-bridges, how do they
interconnect?

I may in fact go with the commercial H-bridge. At best I would save $20
building a MOSFET H-bridge, even less building one using SSRs. We'll see.

[Tom Lum]

I was researching better steering and found this youtube video on a
similar idea of using a wiper motor to steer with servo feedback.
http://www.youtube.com/watch?v=eimYnrCMOLI
Tom

[Guy Gregoire]

HI
 
   I recorded short videos of my Recon running in the backyard.  The video are of very
poor quality, I apologize, but I only have my cellphone to record video, so the resolution
is small, and as I was holding the cellphone in my left hand and driving the Recon with
my right hand, it get easily out of frame.   With these 24v 300amp motors the vehicule
have more than enough power.  If only summer was coming faster.
 
   I have one PanzerIII T050, to battle if ever some battler spawn in my area, so was
originally thinking to build a KingTiger, but turn the lower hull as a 6 wheeled Recon, just
for the fun of running it around.  Cant wait for some forest ride in summer.  I use the
same electronics than the PanzerIII, gotta use it at the price it come, and it work
to good to wait an ennemy tank.  I plan to buy a new car soon, it will be choosed to
accomodate a tank in the back to increase my battling range.
 
   I was wondering, as I fine tune this Recon.  As its a 1/6 scale hull of KingTiger, but
on wheels, would it be possible that I battle with it if I install a rotating turret with
paintball gun on top of it?  I build this vehicule with in mind the idea to replace the
turret for anything else ( ex: robot arm ) as a multipurpose RC mobile base.  Simply said,
can we take liberty with our tank design? If not np, I can still use it as Recon and supply
vehicule.
 
Greg.
 

---

Messenger has tons of new features that make chatting more fun. Click here to learn more.

[Thomas Lum]

While I can't give you a ruling, it's looks like a great little machine.  How do you steer it and are all the wheels powered?

Tom

<Video-0002.3gp><Video-0003.3gp>

[Guy Gregoire]

 
HI Tom, yes the six wheels are powered, one motor for the 3 wheels on the right and one
motor for the 3 wheels on the left, it drive just like a tank, and as you seen in the video
it can spin on himself thanks to that "tank" drive.  To image it, suppose the tank track is
just a chain, that bring the power to 3 wheels.  that chain in this case is inside the tank
and the drive is a bit more complicated because I builded an independant suspension for
all the six wheels. 
 
Greg.

 

---

From: tl...@mac.com
To: rctank...@googlegroups.com
Subject: [TANKS] Re: Recon video
Date: Sat, 21 Mar 2009 16:25:09 -0400

[George MacGillis]

Pictures  Pictures?

PATRIOT GUARD RIDERS #106442
08' Harley Ultra Classic

[Odyssey...@aol.com]

sure seems to do well in the snow. how heavy is it?

 

as to having it listed as a tank, i think it requires tracks. what about as one of them 6 wheeled armored cars or armored troop transports? might not give you the same armor as a tank.....

 

Chris,
Odyssey Slipways

---

A Good Credit Score is 700 or Above. See yours in just 2 easy steps!

[Guy Gregoire]

Hi Chris
 
I understand its better to say a tank have tracks, it make thing simple.  Armored car please me.
I dont know how heavy it is, I would guess 80 to 100 pounds, but I will get a more acurate mesure eventualy.
I am still fine tuning it, every few hours I use it I improve something, but as armored car I will have to add
a turret or rocket launcher on it, I will think about it.
 
Greg.

 

---

From: Odyssey...@aol.com
Date: Sun, 22 Mar 2009 17:25:44 -0400

Subject: [TANKS] Re: Recon video

To: rctank...@googlegroups.com

[Chrysanthos Kanellopoulos]

I think it qualifies for designation as a Groundhog armored car.

Chrys

[Sgt.A.Johnson]

That looks pretty cool i reckon

On Mar 23, 5:26 am, "Chrysanthos Kanellopoulos" <xchr...@otenet.gr>
wrote:

> I think it qualifies for designation as a Groundhog armored car.
> Chrys
>
>   ----- Original Message -----
>   From: Guy Gregoire
>   To: rctank...@googlegroups.com
>   Sent: Monday, March 23, 2009 6:59 AM
>   Subject: [TANKS] Re: Recon video
>
>   Hi Chris
>
>   I understand its better to say a tank have tracks, it make thing simple.  Armored car please me.
>   I dont know how heavy it is, I would guess 80 to 100 pounds, but I will get a more acurate mesure eventualy.
>   I am still fine tuning it, every few hours I use it I improve something, but as armored car I will have to add
>   a turret or rocket launcher on it, I will think about it.
>
>   Greg.
>

> --------------------------------------------------------------------------- ---
>   From: OdysseySlipw...@aol.com

>   Date: Sun, 22 Mar 2009 17:25:44 -0400
>   Subject: [TANKS] Re: Recon video
>   To: rctank...@googlegroups.com
>
>   sure seems to do well in the snow. how heavy is it?
>
>   as to having it listed as a tank, i think it requires tracks. what about as one of them 6 wheeled armored cars or armored troop transports? might not give you the same armor as a tank.....
>
>   Chris,
>   Odyssey Slipways
>

> --------------------------------------------------------------------------- ---

[Steve Tyng]

I've renamed the subject of this thread back to "High Power Servo
Hack" so lets not try to hijack it again (do not change the subject on
these emails).


Now back to the topic at hand.

There's been discussion on making custom h-bridges in lieu of the
commercial boards available. I've found the following integrated h-
bridge modules currently listed on eBay.


http://cgi.ebay.com/MPM3002-Motorola-H-BRIDGE-MOSFET-POWER-MODULE_W0QQitemZ200308339465QQcmdZViewItemQQptZLH_DefaultDomain_0?_trksid=p3286.m20.l1116

http://cgi.ebay.com/1ea-ST-L6203-6203-Full-H-Bridge-Motor-DRIVER-IC_W0QQitemZ400020364797QQcmdZViewItemQQptZLH_DefaultDomain_0?_trksid=p3286.m20.l1116

These are obsolescent out-of-production h-bridge modules (include all
four driver FETS) that could be utilized (with a small amount of
additional components) to drive external motors (up to the current
rating of the module) and can be driven by the hacked servo.

An interesting and current (in production) h-bridge device is the
Infineon TLE5206-2 driver. This is a basic h-bridge driver that
includes over amperage and temp sensing built-in and only requires one
external capacitor. It can handle a continuous 5amps which should be
plenty for a steering servo setup. DigiKey has these at under $8US in
single quantities.

http://www.infineon.com/dgdl/tle5206-2_20010619.pdf?folderId=db3a304412b407950112b43820d56a4a&fileId=db3a304412b407950112b438215b6a4b

If you review the datasheets for these devices you will find all the
info needed to utilize them in the hacked servo project under
discussion.


Steve Tyng

[Paul Hilton]

Now that, Steve is good intel! I am VERY interested in the TLE 5206-2
driver. I am ready for the "Paul Harvey", (the rest of the story) whenever
you would like to post it. A simple (very simple) diagram showing what to
connect where and perhaps a socket part number would be useful. I have a
whopping $8 burning a hole in my pocket!

Paul H.

----- Original Message -----
From: "Steve Tyng" <Stev...@gmail.com>
To: "R/C Tank Combat" <rctank...@googlegroups.com>

Sent: Tuesday, March 24, 2009 4:47 PM
Subject: [TANKS] High Power Servo Hack


>

[Derek Engelhaupt]

Has anyone considered using something like a lead screw connected to a motor for steering instead?  I just bought one of these to do my traverse in the Sturmtiger.  I bought an extra one to play with.  I would think that it would work well for steering and be very robust as it's already welded to an aluminum plate.

Derek
T065

[Thomas Lum]

Would response time be a problem for this method?

[Marc]

Thanks for all the info Steve.

Two years ago we discussed the (discontinued)
http://secure.oatleyelectronics.com/product_info.php?cPath=26_57&products_id=204&osCsid=124c5fdeefb5d022e450ff8ef26acbcc

Would that give the same result? And has anyone by change found a
replacement for this discontinued item?

Marc

[Marc]

[Marc]

Steve, thanks for all the info.

Two years ago we discussed the (discontinued)
http://secure.oatleyelectronics.com/product_info.php?cPath=26_57&products_id=204&osCsid=124c5fdeefb5d022e450ff8ef26acbcc

Would this do the same as the setup you mentioned?
I can't believe there wouldn't be a market for the Oatley board.

Marc

[Derek Engelhaupt]

I'm sure that would depend on a few different factors.  First one would be the speed of the motor turning the screw.  The second would be the travel time of the screw which would be determined by the thread count on the screw.

Derek

[Paul Hilton]

I considered and discarded the original idea I had for the BT-7's steering.  I have a rather well made ball screw that I had intended to use for this purpose.  The ball screw, from the motor end, would seem to work well.  Friction is VERY low, good for not overworking the motor.  The problem shows up when something hits the front wheels.  There is a significant mechanical advantage (at least in the BT-7) from the wheel end.  This means that if your motor is trying to go one way and your front wheels hit a rock, driving your wheels the other way, you are going to blow a fuse.  An Acme screw might fare better, don't know how much.  I think the better option would be a worm screw setup like a wiper or window or sunroof motor.  In this setup, the gear can not work upon the worm, thus effecting a sort of stop.  Of course, if a large enough force is applied to the gear, it will destroy the drive.

 

Paul H.

----- Original Message -----

From: Derek Engelhaupt

To: rctank...@googlegroups.com

Sent: Wednesday, March 25, 2009 10:48 AM

Subject: [TANKS] Re: High Power Servo Hack

[Steve Tyng]

Marc,

The Oatley board does exactly what we have been discussing. I don't
have any direct experience with the kit but I know Frank built a
couple and had issues with them. Maybe he might elaborate.


Steve

[Frank Pittelli]

The only issues that I had during my testing was that the controller is
"twitchy" if you have a sloppy pot and/or the motor speed is not
adjusted properly. That is, it doesn't seem to account properly for the
momentum of the motor and load when moving to a new position. The motor
momentum caused it to overshoot the mark, at which point the controller
tries to reverse direction. But, if there's a little sloppiness in the
pot, then it overshoots in the other direction or just hums while it
stays in the same place. So, the lesson is: use a high-quality pot when
trying to build a heavy-duty servo. If possible, use an accurate
multi-turn pot and gear it accordingly to the drive shaft.

Doug, on the other hand, had lots more issues with them when he tested
them ... primarily related to working load, if I recall, but he can
elaborate better.

Frank P.

PS. It took quite a few weeks to get the Oatley boards from down-under
because he sent them via regular snail-mail.

[Doug Conn]

I tried using the Oatley kits for the second revision RocketMan's steering.
I also noticed the overshoot that Frank mentioned below. The main problem I
had with them was that the position would start to 'drift' after a few
minutes of operation. The device would drive the motor fully left or right
and wouldn't go the other direction. Worse, it would keep the drive current
on. I smoked several sets of driver FETs and burned up two Oatley circuit
boards before I stopped trying. I was on the upper end of the specified load
range (2 amps I think). I suspected maybe the heat was causing the driver
FETs to leak and that's what was causing the drift. I added heat sinks and a
small fan, but it didn't seem to make a difference.

For steering, RocketMan ended up with a ServoCity servo gear box, a beefy
servo, and a home-made mongo servo saver. I thought the steering was
acceptable, but a little slow. We'll see how Tom feels about it :) My
mechanical steering linkages and front suspension were probably the biggest
steering impediments after I changed to the gearbox.

- Doug

-----Original Message-----
From: rctank...@googlegroups.com [mailto:rctank...@googlegroups.com]
On Behalf Of Frank Pittelli
Sent: Thursday, March 26, 2009 10:32 AM
To: rctank...@googlegroups.com
Subject: [TANKS] Re: High Power Servo Hack

[Mike Lyons]

I've just completed a prototype of a custom servo with massive torque
- the gearmotor has a reduction ratio of 1636 (and RPM of 1.5).

The control electronics are the guts of a standard R/C servo driving a
pair of SSRs in a half-bridge configuration ("push-pull") with a
bipolar power supply (+12/-12 V). I had these SSRs lying around,
otherwise I would have used an H-bridge chip as Steve Tyng recommended
elsewhere.

It consistently overshoots twice and then settles nicely (e.g. from
far left, it overshoots a little to the right, then overshoots a
little less to the left, then stops in the center).

I believe there are 2 problems:

1. The momentum of the gear train makes it take longer to slow down
(and start up) than the servo electronics design anticipated. The
damping factor needs to be a little higher so it slows down earlier as
it approaches the required position. I don't have documentation on
the servo controller chip so I don't know how feasible it would be to
change this. I started with the original pot from the servo, then
used a standard pot that is much easier to mount, and the results were
the same.

2. The SSRs have a maximum turn-off time of 1 ms, while the
transistors in the servo are probably under 1 us. The SSR maximum
turn-on time is 100 us, vs under 1 us again. The duty cycle delivered
to my gearmotor may be a lot wider than intended as it approaches the
required position, making the first problem worse. Joe Sommer tested
a couple of my SSRs and found one worked OK with his SSR controller
and one did not, apparently because they were right on the margin and
one was a little slower. In the next few weeks I plan to experiment
with different PWM parameters using a custom controller, and using an
H-bridge chip instead of the SSRs.

3. [Probably trivial] I'm using the existing transistors to drive the
SSRs. This might be compounding the second problem a little, but I
don't think enough to warrant breaking out the PWM signals before the
transistors. If someone can advise me otherwise I'll refine the
design.

I had considered using the guts of a digital servo, but
a) $$$ ($25 vs $9) and
b) the much higher PWM rate could lead to a "shoot-through" condition
where one SSR hasn't finished turning off when the other is turned on,
leading to magic smoke.



On Mar 26, 10:32 am, Frank Pittelli <frank.pitte...@gmail.com> wrote:
...
> ... The motor

> momentum caused it to overshoot the mark, at which point the controller
> tries to reverse direction. But, if there's a little sloppiness in the
> pot, then it overshoots in the other direction or just hums while it
> stays in the same place.

...

[Mike Lyons]

Good news to report: My very slow but very high-torque servo is now
working nicely.

WARNING: ;~} High technology information follows. There are no
relays, microswitches, baling wire, or tobacco spit used. If this
offends you, stop reading. ;~}

I took a cheap analog servo and removed the motor and all surface-
mount devices connected to the motor terminal pads. I connected the
pad to the inputs of a TLE 5206-2 H-bridge (as recommended elsewhere
by Steve Tyng - thanks Steve!!!). I connected the ground of the servo
board to the ground of the bridge. The H-bridge is supplied by a 12
volt SLA battery. The H-bridge output drives a gearmotor, the shaft
of which drives a 5K pot via plastic gears.

I'll take photos and post them when I build the next one.

The gearmotor arm now moves steadily, slows down as it approaches the
target location, then stops. No overshoot, no hunting, no humming.
(No guarantee this will be true with a faster motor.)

I suspect the problem with the previous approach was NOT the momentum
of the geartrain but rather the slow turn-on and turn-off of the
SSRs. The new H-bridge chip turns on and off in a few us (worst case
30).

I'm very pleased with the result so far. Two minor issues: the travel
is about 135 degrees (vs 80 for the original servo) and the motor is
not running at full speed under servo control.

The travel is no doubt due to different electrical rotation angles
between the original pot and my replacement pot (which is much easier
to mount reliably). I will check this later and report. I can fix it
with a different pot, different gear ratios between the motor shaft
and pot shaft, or use a "servo stretcher" to adjust the endpoints.

I wasn't expecting the speed reduction. I suspect the servo duty
cycle is significantly less than 100% at full speed. (It's possible
this is a design "feature", to avoid possible shoot-through when
suddenly changing directions at full speed.) If anyone has access to
a scope and can measure the duty cycle for any analog servo, I'd
appreciate a report.

One caution: I zapped an H-bridge chip by not paying attention to the
specs. The supply voltage can be up to 40V, but the logic input
cannot exceed 7V. This means you can't pull these inputs up to the H-
bridge supply (e.g. if you had an NPN open-collector output). The
output from the servo board is a little under 6V or a little above
zero, which works just fine.

----

On May 16, 1:10 pm, Mike Lyons <mxly...@cox.net> wrote:
> I've just completed a prototype of a custom servo with massive torque
> - the gearmotor has a reduction ratio of 1636 (and RPM of 1.5).

...

[Mike Lyons]

Correction: I removed all the 2-terminal SMDs (from memory there were
three, presumably capacitors to reduce noise). I did not remove the
two PNP transistors that form the top legs of the H-bridge.

----

On Jun 9, 4:08 pm, Mike Lyons <mxly...@cox.net> wrote:
...

> I took a cheap analog servo and removed the motor and all surface-
> mount devices connected to the motor terminal pads.

...

[Doug Conn]

>> If anyone has access to a scope and can measure the duty cycle for any
analog servo, I'd
appreciate a report.

From http://www.societyofrobots.com/actuators_servos.shtml

The motor of an analog servo receives a signal from the amplifier 30 times a
second or at 30Hz. This signal allows the amplifier to update the motor
position. Digital servos use a high frequency amplifier that updates the
servo motor position 300 times a second or at 300Hz. By updating the motor
position more often, the digital servo can deliver full torque from the
beginning of movement and increases the holding power of the servo. The
quick refresh also allows the digital servo to have a tighter deadband.

- Doug

[HV]

My Hitec servos show 20 millisec pulse repetition period.

On Jun 9, 5:17 pm, "Doug Conn" <dwconn...@comcast.net> wrote:
> >> If anyone has access to a scope and can measure the duty cycle for any
>
> analog servo, I'd
> appreciate a report.
>

> Fromhttp://www.societyofrobots.com/actuators_servos.shtml

[Mike]

>>>>
On Jun 10, 3:04 pm, HV <henryverheg...@earthlink.net> wrote:
> My Hitec servos show 20 millisec pulse repetition period.

...
<<<

At the motor output? The conventional control signal is a 1-2 ms
pulse every 20 ms. It would be simple to convert the control pulse to
an output pulse at the same rate, i.e. one output pulse per input
pulse.

What I'm looking for is the duty cycle (the ratio of pulse "on" time
to period) of the output pulse across the motor terminals (or even
better between ground and the base of one of the output power
transistors). At full speed, it should be 100% or close to it.