Re: Servos and Controller boards

Caudlet,

A good reply. Let's take up the points in snips and inserts below...

>In CCED, "caudlet" <info@t...> wrote:
>I did not make that post glibly. I would be the first one to
>recommend that building vs buying is an indivdual choice. You
>noticed I said "I would not attempt it". I have done closed loop
>controls with servos and spent a good many years working with

I think this is partly why you don't wish to "do" what
you've "already" done. Sort of like having to "work" at your hobby.

> The newer chips are all SMT and make a hobby level design a
>challenge for anyone.

DIY-CNC is not all about the "newest" chips. In fact many of our
solutions use age-old techniques and circuits... But even when we DO
use the newer SMD's and SMT's, we're just continuing an electronics
tradition, which means we have to change our techniques to match
what's happening now. Mariss recently mentioned on his geckodrive
forum that the "death" of 4000 series cmos seems imminent. But that
thread also reaffirmed that we will have other choices to replace
them. (Perhaps not as good in some ways, but time does march on) I'm
a member of a homebrew pcb and also desktop SMD oven group. So there
are folks exploring how to bring the new chips under a hobbyist's
control. (And it's not as hard as some would have you believe!)

> If someone posted and asked about designing and winding their own
> stepper or servo motors what would you advise?

We both know the answer to this red herring, or straw dog... Making
motors is not ANYWHERE near making circuits. (I'm not saying circuits
are "easy", but a circuit/pcb is DEFINITELY much more accessible to
the average person than making a motor.)

>Each person has to evaluate their skill level and decide if the
>effort (and probable failure) to build something that can be
>purchased for 65 bucks is worth the "learning" experience.

Yes. But as I said, saving money is not the sole reason for people to
join this hobby.

> If someone is looking to dive into the guts of doing a power servo
> design they should slide over to the OSMC (Open Source Motor
> Control). There you can rub shoulders with guys that have already
> tackled the problem of DIY servo controllers, brushless DC (BLDC)
> controllers and AC servo control.

I'm a member of that group. But they don't seem to be CNC directed;
more of a battlebot stance... (And those two "servos" are not very
closely related, IMO.) IMO, they've taken the high road and gotten
stalled, and the low road has product, but it's not suited to CNC
needs. The hoped-for middle ground just hasn't happened yet...

>One of our responsibilities (sounds noble huh!) is to help guide the
>new persons through the maze. Part of that is to take our
>experience and keep people from heading off down a rough road and
>maybe loose them as part of our community.

I think it's pretty noble to actually ANSWER the question ASKED!!!
(Or admit that we don't have an answer; and offer support so the
person doing the asking CAN try to find an answer that may help us
ALL...

When someone asks us, "Where can I learn <more> about servos?"; I
don;t think we are noble (or helpful) to say, "Don't bother. It's a
tough thing to do, and Vlad over here, or JR, has already done it, so
just buy what they've got..."

Certainly we should warn that it's a potentially tough road. But we
need to temper that with encouragement that might lead someone (whose
skill set we really have NO idea about!) to a NEW solution.

As an example, the EMC project has benefitted GREATly from the
efforts of many; but I'd say that Ray H has had a HUGE impact. He
didn't HAVE the necessary skills when he started (according to posts
he made at the time). But he DID have a desire. Because that desire
was encouraged, and "helped", he pursued a somewhat singular goal, we
now have many additions to the EMC code and GUI...

Why don't we give that same encouragement to the poster who started
this thread? I have no idea what Nick already knows about servos, but
He has asked us to help him "know" more. Are we going to answer his
question??? (What can YOU tell us about high current switching
layouts, mosfet drivers, and such so that WE DON'T have to get a T0-
220 in OUR forehead?!?!?)

<snip>I just get the drift that the electronics design skill set is

>not possessed by the majority of the builders.

We don't need "electronics skill set" to be possessed by the
majority. ONE interested and therefore committed individual is
enough. Like Ray H... Or Paul C... Or Mariss... Or Jeff... Or Vlad...
Or Art F... (The list of INDIVIDUALS who have impaced our "hobby" in
a MAJOR way is a long one. In a very polarised discussion on this
list a few years ago, I said that by using the RDTSC pentium
instruction, windows CNC should be possible. And asked for a 1-8K hz
output. ONE person, Art F, took that as a challenge, and with MUCH
effort, and MANY LONG HOURS (and the co-efforts of hundreds, who
followed HIS lead), we now have Mach2...

What if he had listened to the naysayers, who at the time
said, "Windows CNC is just nort possible, or worthwhile. We have good
DOS programs which work fine. Just use them."??? (You can read these
messages in the archive.)

>Maybe we need a "star" rating system like they do on DIY TV.
>Building your own power supply is ** while designing a closed loop
>servo for power control is *****.

You know, that's not a bad idea. But ** to one person is ***** to
another. Use yourself for an example. Nearly any high
current "switching" electronics is gonna be a * for you. But don't
you have areas which are ***** that others would find to be a *?

Let's try and support the INTEREST of the poster, and not just throw
our experience at 'em... And try to answer the question(s) they've
asked.

Food for thought,

Ballendo