Alt-BEAM Archive
Message #07503
To: beam@sgiblab.sgi.com
From: "Richard Caudle" rwcaudle@hpamonon.com
Date: Wed, 10 Nov 1999 11:24:35 -0500
Subject: [alt-beam] Neural network...BEAMish or what?
content-transfer-encoding: quoted-printable
Check out this neural network developed at USC! Could we adopt or adapt th=
is approach to BEAM robots? Anybody out there know these guys so we could =
get some schematics?
Richard
content-transfer-encoding: quoted-printable
>
Wilf Rigter Sounds like that is all you need. One AC240 for the control circuit and the
servo can easily be driven from a driver made with a second AC240.
regards
Wilf Rigter mailto:wilf.rigter@powertech.bc.ca
tel: (604)590-7493
fax: (604)590-3411
> -----Original Message-----
> From: Jason - [SMTP:evenflow88@hotmail.com]
> Sent: Tuesday, November 09, 1999 10:05 PM
> To: beam@sgiblab.sgi.com
> Subject: RE: solar head questions..
>
>
>
>
> >From: Wilf Rigter
> >Reply-To: beam@sgiblab.sgi.com
> >To: "'beam@sgiblab.sgi.com'"
> >Subject: RE: solar head questions..
> >Date: Tue, 9 Nov 1999 16:15:18 -0800
> >
> >Well I would say, "read the manual" but seriously,
> >
> >1.what motor driver?
> >
> >2.Without the solar circuit ie battery power the light stays on. Darrell
> >added the FLED to "attract attention" to the solar PShead when it is
> ready
> >for action.
> >
> >3.That certain component wouldn't be pin 6 and 8 of the HC240? Those
> should
> >be connected to 0V (instead of floating as shown on Darrell's layout
> >drawing. Also make sure you add a small (eg 10uf /10V) electrolytic
> >capacitor from pin 20 to pin 10. The actual cap value does not matter
> much
> >but it should have relatively direct connections with short leads. This
> >stops high frequency oscillations which upset normal circuit operation.
> >
> >Keep em coming guys and do read the manual.
> >
> >regards
> >
> > Wilf Rigter mailto:wilf.rigter@powertech.bc.ca
> > tel: (604)590-7493
> > fax: (604)590-3411
> >
>
>
> Wilf,
>
> I would like to know if i could make a solar head out of these stuff i
> had
>
> 1.)servo
> 2.)a few 74AC240
> 3.)enough caps and resistor for the PSH V.1
> 4.)1381J
> 5.)six small calculator solarcells which rates 3.5V each...
>
> do u think i need a motor driver or something...thanks a lot
>
> ______________________________________________________
>
7505 Wed, 10 Nov 1999 10:25:38 -0800 [alt-beam] Re: solar head questions.. "'beam@sgiblab.sgi.com'" Wilf Rigter Ideally a 0.1 uf in parallel with a 10uf electrolytic cap near the chip
supply pins take care of the switching currents which circulate through the
IC power pins. This combination of local powersupply decoupling is
recommended for normal digital applications as well (check the databooks if
you like). When using these chips to drive motors without the 10uf cap,
there will be spurious oscillations across the motor windings. BTW the
internal resistance of a supercap is not trivial as it can be a couple of
ohms!
Wilf Rigter mailto:wilf.rigter@powertech.bc.ca
tel: (604)590-7493
fax: (604)590-3411
> -----Original Message-----
> From: Victor Snesarev [SMTP:snesarev@ecel.ufl.edu]
> Sent: Tuesday, November 09, 1999 10:16 PM
> To: beam@sgiblab.sgi.com
> Subject: Re: solar head questions..
>
>
>
> SG wrote:
>
> > At 04:15 PM 11/9/99 -0800, you wrote:
> > >make sure you add a small (eg 10uf /10V) electrolytic
> > >capacitor from pin 20 to pin 10.
> >
> > hey wait a sec....
> > since + and - from the storage cap go directly to 20 and 10,
> > i don't see how another 10uf's would change anything.
> >
> > or am i missin sumpthin here?
> > -Sparky
>
> From Wilf's earlier post:
>
> "Also make sure you add a small (eg 10uf /10V) electrolytic
> capacitor from pin 20 to pin 10. The actual cap value does not matter
> much
> but it should have relatively direct connections with short leads. This
> stops high frequency oscillations which upset normal circuit operation."
>
> I will try to add more detail, but as I am not an expert by any measure,
> feel free to jump in and set me right, anyone. It'll only help me
> understand this stuff better.
>
> The cap doesn't add much at all (can consider it's contribution to be 0)
> to the storage of the storage cap. It is there to compensate for noise
> on the power and ground. It has to be a fast switching capacitor to be
> able to compensate for high frequency (tiny duration) spikes and dips in
> power. Those spikes and dips occur when lot's of transistors inside the
> chip switch at the same time, and in general if there's lots of noise in
> the system from whatever source. When a dip in +V (power), for example,
> occurs, the cap discharges and smoothes out the dip making it less deep.
> Same thing goes for the spikes.
>
> The leads of the cap have to be as short as possible because at high
> frequencies (noise for example) a wire no longer has negligible
> inductance. (Inductance adds more fluctuation to the signal.) So the
> shorter your capacitor leads are, the less inductance they have, the
> greater the effect of the capacitor on smoothing the power and ground
> signals.
>
> HTH
>
> Victor
>
> P.S. You know, it is true: you have to teach someone a concept to be
> sure you understand it.
7506 Wed, 10 Nov 1999 13:24:52 -0500 [alt-beam] Re: Neural network...BEAMish or what?...OR WHAT??? "Richard Caudle" Sorry! Got distracted! http://careertech.com/industrynews/10-1-99a.html
Richard
{8}
Home