[sdiy] uC with mul

[mikko.a.helin at nokia.com]

Seems that the ATMega163-8PI is available at elfa.se for 160 kr = 17.24 EUR.
That's in DIL40 package and is quite similar to AT90S8535 (though I'm not
sure if it can be programmed with the STK200).

-Mikko

> -----Original Message-----
> From: ext Theo [mailto:t.hogers at home.nl]
> Sent: 03. January 2002 8:20
> To: Jim Patchell; synth-diy at dropmix.xs4all.nl
> Subject: Re: [sdiy] uC with mul
> 
> 
> That's a rather large ROM.
> Considering the added costs it might be cheaper to migrate to a
> MegaAVR with mult or move to a ARM part right away.
> 
> Theo
> 
> 
> From: Jim Patchell <patchell at silcom.com>
> 
> >      Here is a thought for doing an 8x8 multiply...use a 
> look up table.
> > If you use a 128Kx8 ROM, you can put a look up table in it 
> to make an
> > 8x8=16bit multiplier.  All you need are two 8 bit 
> registers, and then
> > use one of the address lines to access the MSB and LSB parts of the
> results.
> >
> >    Anyway, just a thought.
> >
> >    -Jim
> >
> > Scott Gravenhorst wrote:
> >
> > > I did quite a bit of assembly programming with a MC6800
> > > MPU (not 68K !) which has no mult instruction at all.
> > > There are cases where you need to multiply a variable
> > > number by a constant.  In these cases, a short sequence
> > > of shifts and adds does the trick and can be faster than
> > > using a generic multiply subroutine.  So if you are
> > > stuck with a no-mult mcu, or if you have 8x8 and need
> > > 8x16 or 16x16, this technique can be useful as well.
> > > Much depends on how fast the builtin 8x8 is.
> > >
> > >>> -----Original Message-----
> > >>> From: ext John Loffink [mailto:jloffink at austin.rr.com]
> > >>> Sent: 30. December 2001 17:08
> > >>> To: Byron G. Jacquot; synth-diy at dropmix.xs4all.nl
> > >>> Subject: RE: [sdiy] uC with mul
> > >>>
> > >>>
> > >>> The multiply cycle on the 68HC16 is 8 cycles.  Other micros
> > >>> do have the
> > >>> multiply instruction, usually multi cycle instructions.
> > >>> These might be
> > >>> good enough depending on your application, particularly for low
> > >>> frequency stuff like note generation, envelopes, LFOs,
> > >>> control sources,
> > >>> etc., but for anything related to analog synthesis it's
> > >>> generally better
> > >>> to have 12-16 bits of resolution to avoid stair-stepping.
> > >>>
> > >>> John Loffink
> > >>> jloffink at austin.rr.com
> > >>>
> > >>> -----Original Message-----
> > >>> From: owner-synth-diy at dropmix.xs4all.nl
> > >>> [mailto:owner-synth-diy at dropmix.xs4all.nl] On Behalf Of Byron
> > >>> G. Jacquot
> > >>> Sent: Saturday, December 29, 2001 10:29 PM
> > >>> To: synth-diy at dropmix.xs4all.nl
> > >>> Subject: RE: [sdiy] uC with mul
> > >>>
> > >>>> Motorola M68HC16 microcontroller series has a 16 bit
> > >>>> multiplier/accumulator.
> > >>>
> > >>> The "little brother" 68HC11 could also do 8*8 multiply, and
> > >>> (IIRC) 16/8
> > >>> division.  They were both kinda slow.  I think a 
> multiply took 12
> > >>> cycles,
> > >>> and the divide something amazing like 42.
> > >>>
> > >>> Byron Jacquot
> > >>>
> > >
> > > =========================================================
> > > - Government: The other religion.
> > > - The media's credibility should always be questioned.
> > > - Lambs who lie down with lions are lunch.
> > >
> > > -- Scott Gravenhorst | LegoManiac / Lego Trains / RIS 1.5
> > > -- Linux Rex         | RedWebMail by RedStarWare
> > > -- FatMan: home1.GTE.NET/res0658s/TekLab_FatMan_WebSite/index.html
> > >
> > >
> > >
> > >
> > >
> >
>