Alt-BEAM Archive

Message #08379



To: beam@sgiblab.sgi.com
From: "Timothy Flytcher" flytch@hotmail.com
Date: Sat, 18 Dec 1999 11:37:40 PST
Subject: [alt-beam] Re: chloroplast


Woe... Wilf, hold up a minuet...
>Interesting behaviours can result, ie the SE may charge
>but not fire until a shadow passes over the LDR (fight or flight response?)

Is this a repeatable response??? Could the be used as a basking response???
A bot that finds the sun then sits and basks like a lizard on a rock???
Timothy...

>Your wish is my command!
>
>Here is a Light Compensated SE (LCSE or should that be lcse?) It should be
>used with a solar cell with 7 to 8V max open circuit. The caps should have
>a
>similar voltage rating. The value of the pot is about equal to the
>brightest
>light resistance of the Light Dependent Resistor (LDR). The LDR points in
>the same general direction as the solar cell for tracking light variations.
>It is usually necessary to desensitize the LDR by partially masking it with
>shrinktube or nail polish etc. Experimentation is required to get the right
>setting of the pot. Interesting behaviours can result, ie the SE may charge
>but not fire until a shadow passes over the LDR (fight or flight response?)
>
>
>
> <>
> > another thing to add on that is a photoresistor. So in more light it
>will
> > run
> > stronger or be more conservative in darker light. Sounds good to me.
> >
> > Steve
><< lcse.gif >>

______________________________________________________



8380 Sat, 18 Dec 1999 11:43:12 -0800 [alt-beam] Re: Solarbotics Miller Engines beam@sgiblab.sgi.com Bruce Robinson > Jean auBois wrote:
>
> ... Andrew doesn't have a patent. Ergo, any attempt to
> enforce your license will certainly fail unless he obtains
> such a patent which is unlikely at this late a date.

> Bob Shannon wrote:
>
> Uhhh, he cannot get a patent in the US now, its already
> been disclosed by Dave's posting.

> JVernonM@aol.com wrote:
> Dave, I hope you didn't pay to much for that license.

Great. A new circuit comes out, and everyone immediately dumps on the
person presenting the information to us.

I'm all in favour in accurate reporting -- I've crossed swords with Mark
Dalton on the Tilden patent issue more than once. But do any of you
really know what you're talking about?

Did anyone check to see if Andrew does have a patent? Or if he has a
pending application?

Even if there is no patent, there are still intellectual property rights
in force. The actual diagram presented (the drawing, not the circuit) is
protected by copyright. You can't legally copy it or sell it or even
give it away free without permission. The term "Miller Engine" is
protected by trademark law, (provided something is actually offered for
sale under this heading).

And as for a license, what has Dave been licensed to use? Did anyone
bother to check? You need permission to publish copyrighted material,
which could be in the form of a license. You definitely need a license
to use someone else's trademark -- the law is particularly strict on
this.

So how do you think Andrew is going to feel about passing on any new
ideas in the future? How are other innovators going to feel if this is a
typical reaction?

Bruce



8381 Sat, 18 Dec 1999 11:53:58 -0800 [alt-beam] Re: Bi ped beam@sgiblab.sgi.com Bruce Robinson Darrell Johnson wrote:
>
> ... Oh, and Bruce, as for you FEDEX example.. what you are
> leaving out is all the countless times students have
> wanted to follow thier dreams, contrary to what thier
> college professors said, and failed miserably.

Yup, you're absolutely right. So if no one tried to do anything for fear
of failing, where would we be? We don't learn a heck of a lot by trying
and succeeding ... we learn by trying and failing. And trying again. And
not giving up.

Not to say that you can get out of a room by running at the wall over
and over and over. You have to try something different each time. And
thinking things through ahead of time helps.

I bet Ian has got all SORTS of new ideas to think about after posting to
this list.

> This will be my last rant on this matter.. It seems my
> practicality makes people upset, and all this typing
> is making my wrists hurt.

Not making me upset :) It's a worthwhile discussion. Pointing out
problems is always helpful -- as long as they don't become reasons to
give up.

Regards,
Bruce



8382 Sat, 18 Dec 1999 12:48:16 -0700 [alt-beam] Miller Engine - licensing... beam@sgiblab.sgi.com Dave Hrynkiw Regardless of the opinions on the list regarding licensing issues, we still
licensed the circuit from Miller because

1) it's a good circuit
2) HE designed it
3) it's quite unethical to make use of it without appropriate permission

Getting something this small patented would be a foolish - waste of money
and time. We're sharing it with everybody because it's simple and cheap. If
it gets utilized in ways that it shouldn't, then I suppose it means it'll
be the last circuit we'll openly share.

Jeez, and I was hoping to have some feedback on the _circuit_, not the
legal issues!



-Dave
---------------------------------------------------------------
"Um, no - that's H,R,Y,N,K,I,W. No, not K,I,U,U, K,I,_W_. Yes,
that's right. Yes, I know it looks like "HOCKYRINK." Yup, only
2 vowels. Pronounciation? _SMITH_".
http://www.solarbotics.com



8383 Sat, 18 Dec 1999 11:58:55 -0800 [alt-beam] Re: PCB iron on transfers beam@sgiblab.sgi.com Bruce Robinson Richard Caudle wrote:
>
>> Good points... I'd also like to ad that "Board houses"
>> don't really want your business... I mean, just set up
>> for a one time batch of sixty boards is a lot of work... So
>> it's not really profitable for them...
>
> Oh yeah?
>
> www.expresspcb.com
>
> Software is free! Quantities from 2 to 200! Two sided
> boards! Good pricing and you can order them off the
> internet! Usually under 5 days to your door!
>
> Richard

Exactly. They have automated the process to the point where very little
human intervention is involved. And they charge you separately for the
various steps in the process. So if you want one board, that's fine --
they make money on the setup as well as the board. And you get a better
price because they've done the automation so well.

Bruce



8384 Sat, 18 Dec 1999 11:56:07 PST [alt-beam] Re: PCB iron on transfers beam@sgiblab.sgi.com "Timothy Flytcher" Yeah!!!! The boards we bought were about $.35 each.... $100.00 for TWO???
You are just paying for there set up!!! My point is NOT that it can't be
done but that it is simply not coast efficient for them to do just 60
boards... So YOU pay for it!!!
Timothy...




> > Good points... I'd also like to ad that "Board houses" don't really want
> > your business... I mean, just set up for a one time batch of sixty
>boards
>is
> > a lot of work... So it's not really profitable for them...
>
>
>Oh yeah?
>
>www.expresspcb.com
>
>Software is free! Quantities from 2 to 200! Two sided boards! Good
>pricing and you can order them off the internet! Usually under 5 days to
>your door!
>
>Richard
>
>
>
>
>

______________________________________________________



8385 Sat, 18 Dec 1999 12:10:13 -0800 [alt-beam] Re: Solarbotics Miller Engines "'beam@sgiblab.sgi.com'" Wilf Rigter Greetings Dave,

Nice job presenting the MSE design

Some comments:

These designs are simple and effective in the best beAM Innovation
tradition.

The MSE circuit, using the 2N7000 is of course identical to the PM1 (that
Miller's too?) but using a bipolar transistor output with a large C2 is a
nice twist.


In the MSE, if a higher trigger voltage is desired, why not use a 1381 with
a different suffix letter rather than adding diodes. BTW the diode voltage
drops at low current should be much lower than .7V ie 400mV or less and for
a 1N914 @ <1mA is more like 50mV.

I really like the MLVSE circuit and see no reason why it should not be used
for all voltages up to the 4.8V - 1381U. The MLVSE is intriguingly simple
in design and application but quite complicated in it's internal operation:

It will trigger at the rated 1381 trigger voltage and at that point the OUT
pin 1 will be at the Vdd pin 2 level which is now connected directly to C1.
The voltage on the C1 starts to fall from load current and the OUT voltage
also drops tracking the main cap. The 1381 does not reset immediately (after
100mV hysteresis) as would be the case without the Vss resistor, because the
voltage at Vss pin 3 is now driven negative through C2 to below the 0V
level! The R2/C2 time constant in conjunction with the 1381 reset voltage
determine the duration of the SE on time.

I think that the MLVSE circuit should work very well with bipolar
transistors. However for the 2N7000, the gate voltage should be 2V or the
fet's Rds(on) will rise too high. Since the fet gate voltage is equal to the
C1 (Vdd) voltage) which is dropping while the 1381 is still on, the fet
Rds(on) is also increasing, robbing voltage from the motor load but still
discharging the main cap. Very careful selection of C1/motor load and R2/C2
time constants is required to avoid wasting energy.

>C2 sets the hysterisis value (the on time) by keeping the voltage the 1381
sees above it's shut-off voltage.

the term "hysteresis value" is generally reserved for the difference voltage
between the trigger and reset voltage thresholds and for the 1381 that value
is between 50mV and 300mV. This is distinct from "the on time" set by the
time constant of R2/C2

Note the clear difference in the operation of the MSE and the 1381/34164
Chloroplast circuits.

The 1381 Choroplast effectively increases the 1381 trigger and hysteresis
voltage by using positive feedback from pin 3 to pin 2 which causes the
voltage at pin 2 to rise to the main cap voltage (up to double the trigger
voltage) after it triggers. The 34164 uses the difference in ON/OFF state
quiescent current producing a difference in forward voltage drop on a series
resistor to do the same thing. This permits this type of 1381/34164 SE to
be used at higher than rated trigger voltage with the corresponding bigger
motor kick.

regards

wilf








> -----Original Message-----
> From: Dave Hrynkiw [SMTP:dave@solarbotics.com]
> Sent: Saturday, December 18, 1999 3:13 AM
> To: beam@sgiblab.sgi.com
> Subject: Solarbotics Miller Engines
>
> Hi all.
>
> In the rapid explosion of SE growth going on lately, here's our little
> contribution. We recently licensed this solarengine design from the Mr.
> Andrew Miller of AM Innovations. Simple and effective, we hope you enjoy
> it!
>
> This circuit is available for private use only, and any commercial use
> must
> be approved by us, the license holder.
>
> Regards,
> Dave Hrynkiw
>
> Instructions:
> The Solarbotics MILLER ENGINE (under license from AM Innovations)
>
> The Miller Solarengine (MSE) is a simple, effective Type-1 solarengine
> with
> a configurable discharge time that can drive both inductive and
> non-inductive loads. The advantage of this solarengine is that it can
> allow
> for frequent very high level bursts of energy that last for only a set
> period of time, rather than letting stored power run completely out as
> with
> traditional 3904/3906 type solarengines. Originally designed by Andrew
> Miller in 1995, this Solarengine design has been distilled to this most
> optimum layout.
>
> The 1381 drives a power transistor or FET via it's CMOS output for a
> period
> determined by the value of the time discharge capacitor, C2.
>
> The diode D1 is any standard signal diode (i.e.:1N914), which adds
> approximately 0.7V to the trigger value of the 1381 selected (ie: C
> trigger
> is 2.2V trigger plus 0.7 = 2.9V trigger voltage). More diodes can be added
>
> in series to further increase the trigger voltage in 0.7V steps (for
> standard signal diodes).
>
> C1 should be a high value capacitor, along the lines of a 0.33F 2.5V
> Panasonic Gold AL capacitor. If voltages required are higher than the cap
> value, placing two in series is advisable (2 x 0.33F caps in series = 1
> cap
> 0.167F at 5V maximum rating).
>
>
C2 sets the hysterisis value (the on time) by keeping the voltage
the 1381
sees above it's shut-off voltage. The value of C2 is highly
dependant on
> the drive component selected. For FETs, values of 0.1µF to 10µF give
> pulses
> 0.4 to 4 seconds long. When using BJT (standard) transistors, the
> capacitor
> value should range anywhere from 47µF to 1000µF, giving pulses durations
> between 0.4 to 2 seconds long. Experimentation is required based on your
> device load.
>
> Current limiting resistor R1 is only really necessary for designs using
> transistors, and are optional for FET variations. 2.2k being standard,
> other values of 470 ohm to 10k are quite suitable depending on your
> application.
>
> If low-voltage (lower than the 1381C set at 2.9V) operation is required, a
>
> germanium diode can be used instead of the standard 1N914, which will
> lower
> the voltage step from 0.7V to 0.4V. This means a 1381C setup will activate
>
> at approximately 2.6V rather than the 2.9V using a standard silicon diode.
>
> For even lower voltage use, substitute a wire for the diode, and place a
> 20k resistor (R2) between the ground leg of the 1381 after C2 and the
> ground rail. C2 will have to be raised appropriately, as it will now
> discharge through the 1381 and the resistor. << File: SB-MSE.gif >> <<
> File: ATT02048.txt >>



8386 Sat, 18 Dec 1999 12:14:34 -0800 [alt-beam] silicon VS SunCeram "adam-m"
content-transfer-encoding: 7bit

Yes Silicon or monocrystalline pv panels (blue) have significant amperage
over the SunCerams... but there good reasons you don't see them in many
beam applications...

1. They suck like a tornado in anything other than bright sunlight. Those
wonderful SunCerams produce an amazing amount of light in 'less than really
bright' conditions. To get the numbers you read for the Blue panels, you
need to be in noon Texas sun.

2. Their voltage output is poor, necessitating hooking a bunch in series.
Now your robots get big and ugly.

If you want to make big bots that only run outside in direct sunlight, then
go for the blue PV's.. They will kick ass, and really go fast. BUT take
them inside and prepare to be underwhelmed.

I bought a bunch of Monocrystalline cells a while back and after a few tests
i was completely filled with anticipointment.

I like bots that wander around on my desk at work under halogen lamps and in
bright room lighting. For that, SunCerams are the best thing.


content-transfer-encoding: quoted-printable



pe>
big =

and ugly.

class=3D010365819-18121999> 

If you want to m=
ake big =

bots that only run outside in direct sunlight, then go for the blue PV's..&=
nbsp; =

They will kick ass, and really go fast.  BUT take them inside and prep=
are =

to be underwhelmed.


class=3D010365819-18121999> 

I bought a bunch=
of =

Monocrystalline cells a while back and after a few tests i was completely f=
illed =

with anticipointment.

ONT =

face=3DVerdana> 


I like bots that=
wander =

around on my desk at work under halogen lamps and in bright room lighting.&=
nbsp; =

For that, SunCerams are the best =

thing.




Home