multiple 555 servo control question

[jtrout13]

Hi, I'm new to this site.

I am building a robotics controller for part of a junior design project at university. I am studying mechanical engineering, and would appreciate any input to this question:

Anyway, my robot has 4 servos. I am using potentiometers from a model airplane controller to adjust pulse width for speed and direction control of modified servos. Anyway, I plan to use one 555 timer in astable mode setting a clock frequency of 50hz, triggering monostable 555's which set the pulse width. I'm sure you are familiar with this type of application. On to the question:

Will it be okay to use a single astable 555 to trigger all 4 monostables, or do I need to set up an astable/monostable combo using a 556 chip for each servo? To reduce parts count and space in the controller, i'd like to use a single astable 555 and combine the 4 monstables on 2x 556 ICs. opinions will be greatly valued.

Thanks!

 

[(*steve*)]

The trigger input has a far higher impedance that the output of the 555. You should be able to drive many 555 trigger inputs from the output of a 555.

The important considerations are the polarity of the trigger signal and the length of the trigger pulse (but I'm sure you're aware of that). I would be thinking about something like this.

 

[jtrout13]

Thank you! The astable trigger is normally high, with short low triggers around 0.4ms, which then triggers the monostables. I have all that modeled in Multisim and working well, but the provided link confirmed what I'm doing.

Since I'm running all 4 outputs as monostable, 2x 556 should be fine, but the 558 really intrigues me. What do you think? I am also having trouble locating the 558 for sale.

 

[(*steve*)]

I think that it doesn't matter how wonderful a chip is, if you can't get them they may as well not exist.

 

[jtrout13]

You have been so helpful. Anyway, I upgraded my online electronics shopping experience from hobby stores to 'superstores' and found much greater variety of parts and cheaper prices as well. I have located the 558 at good prices, so am still considering it.

Since an astable trigger is a square wave output, the fact that the 558 is edge triggered and not level triggered should not make a difference, should it? I am just making sure before I order them.

I have to fit all this inside a model airplane controller, so the 558 will hugely simplify this project. Any concerns? I really appreciate your help.

 

[(*steve*)]

Look at the circuit you're using to trigger. I think there was even some traces of the waveform it generates on the link I gave you.

If it is truly edge triggered, then you won't actually need that circuit.

I would breadboard something and ensure that the trigger pulse can indeed be longer than the monostable time delay (which is what edge triggering would imply).

If holding the trigger low extends the monostable's period, then it's not actually edge triggered.

Can you point me to where you found the 558 and to a datasheet for it?

 

[jtrout13]

Steve,

I got the 558 here: http://www.futurlec.com/ICLinearOthers.shtml If you click on the price, the datasheet is there too (Fairchild Semiconductor).

I'd really like to make the 558 work because of simplicity. Here is a screenshot of my astable trigger in Multisim, along with an oscilloscope showing its output: I'm confused as to whether the edge trigger will trigger each time the voltage rises/falls, or just on falling edges?

 

[(*steve*)]

Yeah, I'm familiar with Futurlec

Thanks. It seems like a thinking man's quad 555. less pins dedicated to each timer but not much reduced functionality (and enhanced in some areas).

I've never really found the need myself for lots of 555s in a single circuit, but this does seem well suited to times when you might.

From the datasheet: "The trigger functions only on the falling edge of the trigger pulse only after previously being high."

 

[jtrout13]

I didn't catch that little footnote about the falling edge. Sounds like it's all figured out then. The 'timing' pin which is combined threshold/discharge, and the common control voltage for all timers really makes this chip appeal to me, so I'm glad I can use it. Thanks again!

 

[kflynn11]

If the 555 and it's like are too difficult to use, Linear Technology has a line of simpler "555 replacements" called TimerBlox (http://www.linear.com/products/timerblox). Programming is done with a single resistor (no capacitors). They even have a dedicated PWM chip LTC6992 that could be a good fit.

 

[(*steve*)]

Interesting device. But I don't think that "difficulty" is the issue. And the 6990 is $US3.81 each! Not a cheap 555 replacement :-o

c.f. $US0.65 for the 558

 

[jtrout13]

Grounding a circuit like this?

Steve, I have another question about this that maybe you can clear up for me.

Here's the electrical overview:
--4x servos with a main power of 4x AA batteries onboard my robot
--Signals generated in a controller from a 9V battery, then sent through tether wire (CAT5) to the servos on the robot. Servos have a three-wire connector, for Power, Signal, and Ground.

Now here's the question:
My professors keep telling me I need a wire from my controller going to a chassis ground. I don't understand this. Since I am just doing signal generation in the controller, why can't I connect all my common grounds in the signal circuit back to the negative side of the 9V battery? If I used a common chassis ground, then where would I hook the negative side of the battery?

They are also saying the same thing about the servos onboard the robot, I don't understand why I can't just hook the grounds coming out of the servo back to the negative side of my AA batteries?

I don't mean to say my professors are wrong, but nearly every battery-powered DC device I've taken apart (remotes, RC cars, etc) just had common/grounds connected to negative side of the batteries.

 

[(*steve*)]

The resistance/impedance of your ground lead can cause noise to appear on your other supply rails.

 

[jtrout13]

The resistance/impedance of your ground lead can cause noise to appear on your other supply rails.

So are you recommending that I put a common ground wire hooked to the chassis of this thing (aluminum plate)? Lets just discuss the 558 circuit you helped me with. Obviously there are some capacitors connected to ground which oscillate with a resistor network to produce pulse widths. I don't understand where to 'ground' these capacitors and my timers, since I'm just running the whole thing off a 9V battery. My intuition tells me to hook all these 'grounds' to the negative side of the battery. I just simply don't understand why I'd need another wire connected to a real 'ground'.

So you can better understand the dilemma, here is a picture of what's going on:

A salvaged RC controller (http://www.futaba-rc.com/radios/futj40.html) will be used to control this robot, however instead of radio waves, signal wires will be connected to the robot instead. I will be installing the 558 circuit and its 9V battery in this controller. Part of the grading includes how many signal wires are used. I still don't understand why I can't accomplish this with only 4 signal wires, one for each servo controlled. I'd really appreciate any help, as time is running out to put the finishing touches on this thing before starting to machine parts.

 

[(*steve*)]

Yeah, the ground is often the -ve rail of your battery.

If you are using a different power supply for the servos, you may want them to share the same ground. i.e. the -ve rail of both sets of batteries is common. That may be what has been suggested to you.

Depending on how you connect the 2 negative rails, however, you may introduce noise into your system. Generally you want to keep power and signal separate.

This may mean linking the grounds near the servo rather than near the batteries so that IR effects (I x R, voltage drop caused by current through the non-zero resistance of your tether) along the common ground are as low as possible.