Alt-BEAM Archive

Message #05248



To: beam@sgiblab.sgi.com
From: Bob Shannon bshannon@tiac.net
Date: Sun, 11 Jul 1999 09:07:06 -0400
Subject: [alt-beam] Re: Something funny with the 1382 voltage triggers?


Steven Bolt wrote:

> On Sat, 10 Jul 1999, Bob Shannon wrote:
>
> > > The first key question is then: Is the performance of the SMT parts
> > > as per datasheets reasonably similar to the performance of the
> > > thru-hole parts? Of course you checked that, and the answer is yes.
> >
> > Check the data sheet again. Several numbers ARE different for the SOT 23
> > package.
>
> So you substituted transistors which per datasheet are *not*
> equivalent to the thru-hole originals, and you are surprised that
> your gadget won't work as specified?
>
> Sorry Bob, but that's plain stupid.

Steven, please drop this attitude. Its totally inappropritate here.

Apparently you have not checked the data sheets either, a SMT 3904/06 is not
quite the same as the TO92 parts. If you think they are, your shooting from
the hip.

I'm getting private responses from people who hare having this problem, and I
can see
why people don't want to discuss it over the list right now.

I'd like to resolve it, and maybe learn something, and let those others know
what
I find. As you seeem more interested in having people see it your way, and
have said
that your not going to debug this uninteresting circuit, why clutter the list
this way?

Your re-posting of the problems with Dave's magbot kit were very helpful, but
scince that you seem to be doing more interested in dedate than debugging.

If one of my engineers showed this type of attitude, we would be having that
dreaded
career-path discussion. Here's a circuit with a problem and some people do not
want to
replace SMT SE's, period. Have you ever heard the expression "the customer is
always
right"? Engineering is a practical service, not a pulpit.

For still those still (suffering) this thread (and problem):

Adding the current limiting diode (or current limiting MPF102 JFET) is not a
pracitcal
solution for SMT SE's.

The SMT parts will remain latched on at well under 3 ma of charge current.
If you limit charge current to values below this, charge and fire times become
very long (and the current limited charge curve looks a bit different too).

Latching up at 2 to 3 ma seems to suggest that this problem is not unique to
the 3733
solar cell. A 2224 can supply this level of current (so much for the large
cell theory).
This also suggests that a series resistor between the solar cell and the rest
of the circuit
is less than practical. I tested several resistors in series with the solar
pannel, but the values keep going up without fixing the problem before they
were effecting the charge times.

But the current limiter trick for more advanced behaviors does seem to work
well
with the thru-hole SE's. I'll persue this later...however uninteresting it may
be...

I had suspected something intermitant with my capacitors, so I totally replaced
that
bank of caps, (due to a sudden increase in leakage current preventing low light
operation).
Those caps were removed, and verified as being leaky. With brandy new caps the

full sun lockup still happens just as before. Its not the caps, for sure, but
my leakage us
back under control.

Also the lockup problem is NOT sensitive to the total capacitance level as it
had seemed
while I had the leaky caps installed.

Adding diodes to the base of the NPN transistor is a very easy way to adjust
the shut down
voltage of the latch, this works perfectly, except that it will change the
lockup voltage
of the latch rather than resolve the lockup problem itself. Raising the shut
down voltage
seems to have a greater effect on total (useful in my application) energy
delivered to the
motors than it does on the cycle time of the SE. More theoretical efficiency?

I'll try to isolate the 1382 with a diode as Terry Newton suggested, now that
the electronics
are no longer attached to a chassis, its fairly easy to try this properly, but
there is only so
much rework a SMT board can withstand.

I'll attack the motor side of the circuit with resistors next, and also
explore
what even larger values of PNP base resistor may do as well.







------------------------------------------------------------------------

eGroups.com home: http://www.egroups.com/group/alt-beam
http://www.egroups.com - Simplifying group communications



Home