A Few Beginners Needed for "Beta Test" Building of Servo Center Hardware.

[Soeren]

Hi,

I have designed a very cheap and simple (but very precise) circuit (hardware) for a 1.5ms reference signal - for servo centering without any programming needed (to rule out eventual programming errors).

Actually, I made two variants of the circuit, of which the least precise will still be within 2..3µs and this tolerance is of no consequence, as it is below the dead-band timing of all servos I know of.


I set out to make a very cheap/simple circuit with adequate timing precision, to help out all the beginners who want to center their servos before programming anything, as having a reliable reference takes out some possible errors in the equation. I imagine that some of the more experienced builders might find some use for it as well.
I believe I succeeded in this first part of the circuit*, but the real test is to see, if a beginner is able to recreate the circuit/signal with a minimum of tools.

*) This is just step #1 in a circuit that can be made into a more general servo tool by adding only a handful of passive components if enough people want it (hardware done, but a bit of PC software remains to be done).


For this first step, I'd like a few volunteers for testing either of the circuits and the profiles I'm looking for...
- Probably beginners up to somewhat experienced.
- Needs to have a servo (or an oscilloscope) and a ~5V supply for testing it against.
- Able to follow a schematic to the letter, without substituting component values (unless verified with me first).
- Able and willing to PM me with a photo of their circuit and their results.
An oscilloscope for checking would be nice, but is not a demand in any way.

If there's just a 1% chance of you not being able to complete it, because your studies/work/family/whatever was a greater demand, please do not apply.


Here's how simple the circuits are...
Circuit #1 has got: 6 resistors, 1 capacitor and 2 transistors.
Circuit #2 has got: 5 resistors, 3 capacitors, 1 diode and one 8 pin IC
(Plus for either circuit, pin headers for the servo and power and a connector of course)

Necessary files (Schematic, PCB layout etc.) will be provided.
You supply the components and finish the building and testing within around a week (assembly should be less than an hour, even on matrix-board, so an afternoon or an evening is all it takes, testing included).

When done, you will  have a precise reference that will be easily made into a lot more (even transcending servos) if demands exists and you will have it before anyone else

When tests are concluded successfully, it will be posted to the public - until such time, I expect the beta testers to not disclose any of it.


If you are interested in becoming a beta builder, you can apply by sending me a PM with a subject like "Beta Test" and a short and simple description of your experience level ad how you fit the profile.
If you plan to apply here (in the open forum), I will assume that you are unable to read and hence are unable to follow the plans, in which case you need not apply

I will start it up, when I have found a suitable range of testers and I assume this to be done around a week after that time.
Beta tester will be mentioned in the finished project (unless they ask not to be of course).


I think I covered everything worth mentioning about the project, leaving nothing unsaid, but if you find anything missing or unclear, just ask here  - despite the sound of this post, it is just the same old me

[Soeren]

Hi,

Just resurfacing this.

Strangely, it seems nobody is interested, which puzzles me, thinking of how regularly people post questions, where this circuit would be a prime solution??

I'll give it a few more days, before moving it to another forums, so if anyone is interested, don't put it off until it's gone.

[rbtying]

My lack of response stems primarily from the fact that I probably don't have the parts for it, and the FIRST Robotics competition season is starting up--otherwise, I'd have jumped at the opportunity.

[Resilient]

Same here, I probably don't have the parts and

If there's just a 1% chance of you not being able to complete it, because your studies/work/family/whatever was a greater demand, please do not apply.

there is probably like a 15% chance I will get busy with classes and end up putting it off. 

[Soeren]

Hi,

If you haven't got the time, you haven't got the time


I always strive to use the most common components (and values), to make sure parts can be found everywhere, but perhaps I should post a parts list here (for the transistor circuit, I think I'll just start with that).

Resistors: 4 x 10k and 1  x 22k
Capacitor: 1 x 100nF
Transistors: 1 x BC547 (or 2N33904) and 1 x BC557 (or 2N3906)

Not exactly rarities and probably what most people have in their stash.

A total of 25 solder pads (including connectors/pins), connect and run

[Pogertt]

Actually, I made two variants of the circuit, of which the least precise will still be within 2..3µs and this tolerance is of no consequence, as it is below the dead-band timing of all servos I know of.

If you are specifying 5% Resistors, and you get a 5% low, and a 5% high unit in your project, won't that possibly change the speed of oscillation by up to 10% ?

.0015 X 1.05 =  .001575 or   .0015 X .95 = .001425 or a worst case error of  .00015 between different constructs with worst case high low tolerance component combinations?

This is a maximum variance of 150 uSec.

 
I am curios as to where the 2...3uS precision comes from.

[Soeren]

If you are specifying 5% Resistors, and you get a 5% low, and a 5% high unit in your project, won't that possibly change the speed of oscillation by up to 10% ?

No.

Not that it's relevant for my circuit, but getting the resistors to fall on each side of the nominal value is by far a the ideal situation (as this cancels out a lot of errors). Any batch of resistors will usually be either above or below the nominal value, except when they're all that close that they all "vibrate" around the nominal - and modern days 5% tolerance resistors are way better than 5%, most even within 1%.

.0015 X 1.05 =  .001575 or   .0015 X .95 = .001425 or a worst case error of  .00015 between different constructs with worst case high low tolerance component combinations?

This is a maximum variance of 150 uSec.

No, it's not.

It's just you trying to be negative about a circuit that you haven't seen, but as you have shown before (actually an alarming percentage of your posts here so far), you just want to be negative towards my posts, so you make large and incorrect assumptions, only to criticize something that you have neither seen nor understand.

Had you been putting the same amount of energy into the only circuit I have seen you post so far (and which I have repeated below to give you something real to sharpen your fangs on), you might have found at least a few of its many errors and omissions before posting it to a complete newbie.


It is rare to see a circuit with at least as many errors as there are components and what's really sad is, that it only takes modest changes to make it into a decent working circuit - even an average and half sleepy electronics designer could do that - but that goes well to demonstrate the difference between a Real Designer^TM and a wannabe

Almost sorry to tell you, but that post/circuit shows, that you're still so far from the virtual cookie jar, that you'll have to make do with the crumbs that real designers drop now and then

I am curios as to where the 2...3uS precision comes from.

Of course you are. It must sound like magic to a guy like you, letting component tolerances dictate the precision you get - something like how the enemies of The Black Prince would have felt, had he somehow managed to introduce a Minigun at Crécy in 1346 (well, he almost did, thanks to his archers). And while we're at the subject of Edward, I think that the motto he coined a couple of years later will suit my circuit perfectly, so:
 Honi soit qui mal y pense

And in due time, you can get your curiosity towards Real Circuit Engineering^TM fulfilled - just check regularly in whatever forum that helps me with the initial tests, as I think that they (whoever it turns out to be) deserves to get the traffic that having it publicized on the site will bring - as a return favor.


The real beauty of my circuit and why I made it the way I did is, that anyone can get this high precision, if they're able to read a schematic, or at least make and populate a PCB from given layouts and follow simple written directions


I made this design to help others, mainly beginners, as I have ample instruments to cover my own needs for timing.
What are you trying to achieve by babbling about unrelated issues based in resistor tolerances?
(judging from your antagonistic posts, with very little usable content, I can only come to the conclusion, that you're bent on trying to block beginners from getting good help) - Why is that?

[Pogertt]

I think your Beta test is fast going the way of the Beta video recorder.

Have an Oreo on me.

[Soeren]

I think your Beta test is fast going the way of the Beta video recorder.

Snappy, bright and very related to the subject.

Have an Oreo on me.

I think I'll settle for a couple more of the good laughs, that you haven't yet failed to provide.