Coil Winder

[Learner]

[quote author="CJ"]the smaller the wire, the more turns per inch.
the n more turns per inch, the less you want the traverse to move.
so they have to be synced in order to get a nice layer.[/quote]

Does the traverse move from L-> R then R-> L or does it needs to pop back to L again to start a new layer from L->R and always start from L->R?

 

[Learner]

Hey, the API xfrmer looks pretty randomly wound

Whats the advantage of having guided winding besides flat layer?

 

[CJ]

Foldback Winding- one layer is wound, then the wire is brought back to start the second layer, only from the same spot as the first layer was started from.

Back and Forth Winding - Well, self explanatory.

Flat layers decrese leakage and DCR, randome winding can spound good also, depending on the transformer and application.

 

[CJ]

Still on it, agonizing, some solution possibilities:

The PIC genius at work says he can program an F series 40 pin to run the stepper.
He recomends using a serial output to a computer for user interface, with plenty of back up ports for a dip switch backup interface, incase you did not want to boot the comp.

saYS a coupLe of hours work, so figure $200 bucks, UNLESS, there is some debug needed. (you know how that goes, write a program in 1 hour, debug for three years)Maybe I could debug if he just configured the headers and addresses, the I/O, ETC, and wrote the basic code.

The other solution which I am going to try first is a phase locked loop to multiply the winding motor pulses by ten, then I can use Analog Pakrats solution of the divide by 256 chip.
This would make it a two chip, analog solution, easily DIYed by you folks.

Only concearn is freq response at 2 meg hertz! :shock:

 

[jdbakker]

[quote author="CJ"]The other solution which I am going to try first is a phase locked loop to multiply the winding motor pulses by ten, then I can use Analog Pakrats solution of the divide by 256 chip.
This would make it a two chip, analog solution, easily DIYed by you folks.

Only concearn is freq response at 2 meg hertz! :shock:[/quote]
What's the range of pulse rates / motor speeds you need to lock to ? I'd be more worried about that; it's hard to find/build a VCO with more than one octave of tuning range.

Page 4 of this document has a CMOS divider which can be set between 1.000-9.999 . Might be interesting.

EDIT: Thinking it over, a chain of 4527s may be all that you need, with a few BCD thumbwheel switches to set the ratio.

JDB.

 

[McCroskey42]

If you're looking into doing something PIC-based, have a look at the MIDIBox:

http://www.ucapps.de

They've already got a set of modules that provide the basic support circuitry for the PIC, storage for programs, an LCD interface for an on-screen menu, and analog and digital I/O modules for controls, and ready-to-buy PCBs . Plus, since there's already a good-sized group of DIY types on the project, you'll be sure to find help.