Alt-BEAM Archive
Message #05062
To: "'beam@sgiblab.sgi.com'" beam@sgiblab.sgi.com
From: Wilf Rigter Wilf.Rigter@powertech.bc.ca
Date: Tue, 6 Jul 1999 21:52:52 -0700
Subject: [alt-beam] Re: Something funny with the 1382 voltage triggers?
Hello Bob,
you wrote:
>If you read the list archives, there is a body of thought (supported by
some evidence)
> that lowering the 2.2K resistor to 1.8K often solves the whining motor
oscillation problem.
> I agree that this trick works in practice, but the circuit theory might
lead you to believe
>otherwise, that it narrows the operating range of the circuit.
I assume the whining problem refers to the START of the discharge cycle and
I can understand that lowering the 2.2K resistor value will solve the
problem by increasing the PNP and NPN base currents and cause the NPN to
saturate for that specific motor current thereby avoiding the "whining
motor" syndrome which is caused by instability of operating the SE on the
edge of saturation.
However you earlier described a different problem which was resetting the
latch at the END of the discharge curve during which in fact you really want
the NPN to come out of saturation and turn off. I suggested that increasing
the resistor could solve that problem but could cause the problem of motor
oscillation at the beginning of the discharge. So far everything fits my
theoretical model of the SE but requires confirmation by experiment (just to
keep me honest).
Did you actually try increasing the 2.2K resistor to 4.7K?
You wrote:
>Personally, I think teaching debugging is as important as electronics
theory.
Yes, it is useful to learn the theory and operation of these very simple yet
quirky SE circuits.
Nevertheless it is good engineering practice to design circuits which are
insensitive to "normal" variations in passive component values or to
transistor gain variations or to changing parameters caused by temperature,
humidity etc. If you are a beginner and want to play it safe, build a well
designed circuit and use it as recommended. It will work as advertised but
you may not learn much about the underlying electronics.
However if willing to take a little risk (no pain: no gain =(=) and you
want to know how and why circuits work (or don't work), then build an SE and
play around with it until you understand it and get it to work. In the
process you build up a useful mental model of component and circuit
behaviours which with some practice will allow you to design, build and test
your own ideas. All this requires performing real experiments, asking
questions, learning theory to explain your observations and making
modifications to progress to a better circuit design. The bottom line is
that, aside from the reusable motor, solar cell and super cap, the cost of
the other SE components is virtually zero. You can learn more about
electronics from solving SE problems (if you survive the frustration) than
putting together a "build by number" albeit reliable, tried and proven
circuit design.
On the other hand, quick results can also be encouraging and can make the
assembly of a reliable kit a satisfying experience. If you are willing to
hack into your (possibly expensive) kit to experiment and push the design
envelope, you may also learn a lot about the operation of electronic
circuits and parts.
If all goes well either process will lead to the same end result of learning
why and how things work!
You wrote:
>Now that we are all brushed up on the circuit theory, and failings of BEAM
as a teaching tool,
>would anyone care to look at the actual problem?
Oops. I guess maybe we disagree on the teaching tool bit but I will look a
little more at the actual problem.
As Steven Bolt suggested a timed discharge cycle avoids many of these
problems but he suggested using ICs to do the job. My first SE, the FLASH
(not FLED) SE can probably be adapted to the 1381 SE. It uses just one
additional capacitor to reset the "latch" after the "flasher" times out.
enjoy
Wilf Rigter mailto:wilf.rigter@powertech.bc.ca
tel: (604)590-7493
fax: (604)590-3411
> -----Original Message-----
> From: Bob Shannon [SMTP:bshannon@tiac.net]
> Sent: Tuesday, July 06, 1999 10:40 AM
> To: beam@sgiblab.sgi.com
> Subject: Re: Something funny with the 1382 voltage triggers?
>
> Steven Bolt wrote:
>
> > On Tue, 6 Jul 1999, Sean Rigter wrote:
> >
> > > And hopefully learning something about electronics in the process.
> >
> > I'm sure there is a lesson to be learned here. But can we be so
> > optimistic as to say that this is significantly advancing anyone's
> > knowledge of electronics?
> >
> > Methinks BEAM should become a little more efficient as a teaching
> > aid...
>
Personally, I think teaching debugging is as important as
electronics
theory .
> I'll admit to being chronically rusty at analog transistor circuitry, but
> I
> think we are
> missing something here.
>
> This SMT SE lockup problem is not being reported, and cannot be duplicated
> with identical component values and conventional parts. This problem is
> being
> reported on several SMT SE's built by different people.
>
> One helpful email suggested that lead lenght and EMI might enter the
> picture if any
> real difference existed, but I've looked at this carefully, that this is
> not a factor.
> When using the SufBoard for SMT prototyping, you actually have much longer
> lead
> lenghts than with a free-formed SE.
>
> Yet the SMT SE's do behave a bit differently, as a matter of objective
> fact. I'm testing
> this right now, and I see a difference. This difference is harder to see
> with less efficient
> motors however, but this makes some sense after some study.
>
> If you read the list archives, there is a body of thought (supported by
> some evidence)
> that lowering the 2.2K resistor to 1.8K often solves the whining motor
> oscillation
> problem. I agree that this trick works in practice, but the circuit
> theory
> might lead you
> to beleive otherwise, that it narrows the operating range of the circuit.
>
> So all the detailed circuit theory is very helpful, but when we seem to
> see
> conditions where
> the theory does not match the evidence we need to find out why, and debug
> the actual problem, not the theory. In my experiance, the 1.8K resistor
> 'trick' is a sure fire fix for the
> whining motor full sun lockup condition, with walkman style motors. The
> motor type
> makes a big difference here.
>
> The fact remains that the same measured component values produce different
> behavior
> with the SMT parts. The 'flaws' in the SE are an issue here, but they
> alone are not the cause of the problem.
>
> I've swapped the actual solar cells motors and, and the capacitors as
> well,
> emperically.
> I've measured the resistor values, and tested 3 full sets of SMT
> transistors and triggers.
>
> There is a difference.
>
> Now that we are all brushed up on the circuit theory, and failings of BEAM
> as a teaching
> tool, would anyone care to look at the actual problem?
------------------------------------------------------------------------
eGroups.com home: http://www.egroups.com/group/alt-beam
http://www.egroups.com
- Simplifying group communications
Home