Re: Rutex R2000 new product! Comments please.

Steve
Good explaination. Thanks for the rundown and the response to his
question. It helped me out, also.
Randy

--- In CAD_CAM_EDM_DRO@yahoogroups.com, Stephen Wille Padnos
<spadnos@s...> wrote:

> Lance Hopper wrote:
>
> >Hello,
> > Been out of the loop for a while. I noticed Rutex has a new
> >product named R2000. It's got SPI, serial peripheral interface,

what

> >does this do? Also it 'appears' to have a pulse generator (up to

7

> >axis!?) on board, ie. "...theoretical speed up to 80Mpps." A
> >realtime 32bit DRO etc... Says to be available Jan 2005, which is
> >well passed, but there is no mention of price.
> >
> > Does anybody well versed in electrical/CNC jargon care to

comment

> >on this products specs/features and it's possible significance to

the

> >hobby/low cost cnc market? It appears to have the potential to be

a

> >hot item, but I don't understand alot of it, and who knows how

much

> >it will cost.
> >
> >
> Well - SPI is meant for communication between chips on a board. It

uses

> 3 wires: data in, data out, and clock. You generally also need a

chip

> enable line, especially if you have multiple devices connected. A
> number of microcontrollers also use SPI (or something similar) for
> programming.
>
> Using SPI to send rate commands, instead of using step pulses,

makes a

> lot of sense. Many chips can run the SPI Bus at 20 Mbits/second,

so you

> can get a lot of information back and forth. Since there's a
> microcontroller on board, there is no reason to use an external

timing

> source - any microcontroller is 100 times better at generating

accurate

> timing than any PC today. (I have made devices with 1 MHz
> microcontrollers that have 1 microsecond jitter on interrupt

response,

> something that's nearly impossible on a 3GHz PC)
>
> The "7 axis" spec looks like if comes from the 7 sockets on their
> carrier board. If they just send out 7 SPI commands at a time (in
> parallel), it takes no longer to speed on 7 axes than it takes on

one.

> If they send commands in series, it can still take only 20-40
> microseconds to send out the full set of velocity commands.
>
> The 80Mpps spec has me a bit confused. I see a PIC microcontroller

and

> a 10MHz crystal on one board (it looks like 11MHz on some of the

other

> boards). A PIC divides the external clock by 4, so you get an

internal

> cycle rate of 2.5MHz. No peripheral on a PIC can run faster than

the

> external clock reference, so there's no part of that chip that can

go 80

> MHz, and there's no PIC faster than 48MHz. Some of the boards

have

> other chips I can't identify, so they may be capable of 80Mpps

operation.

>
> The "Built-In real Time DRO" is just a counter. With bi-

directional

> communication, it's possible to get position feedback from the

drive.

>
> I have no idea how much it will cost :)
> - Steve