Re: rE: re:RE: Pendant

Terry,

(snips, inserts below)

>I would point out that on this base model I am trying to beat the
>price down as much as possible and don't see uses other than a power-
>feed usable on four axis without swapping cables.

Which would be a very useful product for some. Add a couple features,
and it may become a useful product for MANY. Now, I've rallied
against feature-itis before and I know of the pitfalls :-) But...

You can sometimes take advantage of the effort/money spent vs.
results gotten curve! (i.e., sometimes a LITTLE more IN, gives a LOT
more OUT!

>Yes I might attempt to cut some semblance of an angle on none
>critical parts using the hand-wheels manually and so could use this
>version of the pendant to do the same, pressing two of the axis
>buttons at the same time, but I wouldn't expect to attain a near
>perfect result as one would expect from computer control.

Why not? The PIC IS a computer! If you output steps to axes with
equal SPI (Steps Per Inch) at the same rate, it WILL be a perfect 45
degree cut. Other angles, or uneven axis SPI values, will/could be
accomodated with a simple linear interpolation.

>The speeds available are not yet set in stone as the programming is
>not completed. These have come about as a result of "No comment"
>from anyone as yet, and me having a Bridgeport with a lead-screw
>lead of .2". I am more than willing to change the list of speeds
>should anyone wish to suggest them.

I HAVE been suggesting that you NOT set this in stone, but provide
SOME way of setting an 'overall' range, which 'contains' the 12
speeds... For 4 axis s/d, you need 8 I/O pins. 2 additional pins
(with jumpers or DIP switches) would give 4 'ranges'; 3 pins would
give 8 ranges; 4 pins would give 16, etc.

>Should I believe there is a market for it I could write a whole bunch
>of programmes for this device and burn the chip to nearer a customers
>requirements. The PIC is in-circuit programmable. The end user would
>still only get the 12 speeds but they could be nearer his choice.

My paragraph above is intended to prevent you from NEEDING to do it
this way. In product design, it is easy to overlook the long-term
results of choices made early. Sometimes they don't matter, sometimes
they do...

Using a slightly more expensive PIC (to get the pin count to
implement ideas I've been expressing) may raise the cost a dollar or
two. BUT, the 'Busy-ness' business costs: writing and
maintaining/developing near duplicate code, the multiple descriptive
paperwork, shipping error possibilities, misunderstandings, product
perception!! will ALL be impacted by the choice you make.

>The problem is I have not data as to what is required as no-one has
>yet offered an opinion to this end.

I have used machines from 500 to 100000 steps per inch.
IMO, COMMON SPI would be;
500 (on woodworking routers operating at high feedrates, also pick
and place, again high feedrates);
2000 (probably the MOST common, as it is the result of a 5tpi
ballscrew and a 1/2 step 1.8 degree stepper);
4000 (same as above, but 1/4 step drive);
8000 (sherlines' and others with 20 tpi leadscrews, some encoder
types);
around 16000+ (microstepped or encoder derived servos').

Feed rates from .1 IPM (Inch Per Minute) to 900IPM. As a 'longtime'
machinist, you probably know what feeds/speeds are most used. To
accomodate the wood router people, think of high speed aluminum
machining, the two are very similar. So it's just a matter of
applying the math to the numbers above, combined with your knowledge
of your market.

Really, 1IPM to 150IPM will prob'ly cover MOST apps...

Hope this helps.

Ballendo

P.S. A well designed electronic product can often be sold in 'dumbed
down' OR 'smarted up' versions more easily(leave components off/on)
than many other products... But you have to have at least some idea
of the top-level functions needed, and provide for them in the
beginning.