Hi David,

> Only problem I can expect from the signal is that they are
> not exactly the same length.
> (e.g. Pulse 1-2 is 20ms + 1,8 ms = 21,8 ms and pulse 1-3 is
> 20 ms + 1,3 ms = 21,3 ms)
> Basically I can try to add a second pause of 19,5 ms for
> pulse 1-2, but is the servo this
> critical.?

The time between the pulses is not very critical.. from memory I think mine
ended up about 15ms.

>(Do you mean this by ramping.?   I can try to change the pulse in
> little steps, but how small should the steps  be?

I start at one point (your 1.3ms pulse for instance) then change it over a
short time into you 1.8ms pulse.. Ie 1.30, 1.35, 1.4, 1.45... etc.
 time..)

Now that I know what is connected, I don't think this is going to be the
problem. Servos are limited in their response by the speed of the motor
anyway..

> > is a PIC based, and each pulse output has a 220 ohm resistor in series.
I
> > have some photos of the setup, but I don't have a handy scanner..
>
> Can you remember why you put the 220 ohm resistor in series..
> ? Just incase? Or was there a reason...?

The board that controls it is actually a general purpose I/O board I
designed that connects to a serial port and you can control it with software
like Macromedia Director for multimedia installations. I just re-wrote the
firmware for this particular board so it didn't need a PC to run it.

I designed this board with ALL the directly accessible I/O lines with 220
resistors for Electro-Static Discharge and over current protection.

> > I also power the unit from a 60W dual output SMPS OEM unit.
> I adjusted the
> > 5V output to 6V for a little extra power to the servos - I
> think 5V is the
> > lower end if their range. And have the 12V output powering the PIC
> > controller (onboard regulator).
> 60W is quite a power supply for the PIC + servo's... :-)
> Normally the servo's are not that critical on voltage supply.
> In RC models they are used
> on 4 NiCD batteries, some people use 5 NiCD batteries for the
> same reason you
> used a higher voltage + they get some extra speed in the servo..

Yep, 60W is a lot, but the giant scale servo's can draw quite large pulses..
And it was the smallest one I could get that didn't cost a fortune.. Nothing
like over-engineering :)

> BTW, they above setup is used to push a mechanical button... :-)
> It is a long story, but basically I am asking help for the
> someone else which I am helping..
> This project is a small black box used to push a button of a
> gambling machine. This are the machines
> which you press a button and three colored (fruit) rollers
> rotate and you win or loose some points.

Aha.. Ok I know what you mean, they are all over Australia as well :) I also
have a little bit of experience with the insides of them myself.. Had a
customer need some custom hardware.. The ones over here are just about all
electronic..

All I can say is that They are a TAX upon the mathematically challenged..
You simply can not win in the long run.

I'd suggest that the problem is that you have the servos trying to move a
certain distance BUT the button won't let them get there because it is
mechanically stopped. This could cause big problems for the servos over
time. Think locked rotor on a motor and large currents..

My suggestion - get rid of the servos and go with small solenoids. These are
simply a big coil with a moving rod. when the coil is powered it either
pushes or pulls the rod.

You should also have some mechanism to stop pushing against the button so
hard - a simple compression spring would do.. Spring is attached to end of
solenoid (or even servo), other end of spring pushes against button. When
activated, the spring pushes the button, and when it can no longer move in,
the spring compresses.

Hope that helps.

Cheers,
Ash.



---
Ashley Roll
Digital Nemesis Pty Ltd
www.digitalnemesis.com
Mobile: +61 (0)417 705 718

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads