Search options (Page 36 of 151)

... and a glue solution wasn't quite strong enough.  I have two other things to try.

I'm working on a set of scenes to show what Melayne is up to.  I'll put them up as a chapter even though they will have to interlard the other story threads.  Right now I'm working on a brief appearance by Pengrit.

Reviewed Randy's A Cartel's Revenge, Ch's 5 & 6.

Happy Ever After, distinct from HFN.

"Perfection is achieved not when there is nothing left to add but when there is nothing left to take away."

K, you got there first!  But Is Anver's distaste for his responsibilities an obstacle or a flaw?  I think it's a flaw.  Or maybe a quirk that helps us empathize with him.

He's goaded into the duel with Alina, maybe for his feelings rather than just cause.  This is why I think letting Anver know about how Alina killed his friend before he commits to the duel is a mistake.  It's too much outer reason and not enough Anver

Anver has flaws.  He doesn't want Kha's job.  He doesn't want more power.  He's stuck with a staff with a mind of it's own.  And he's vulnerable to Katerine.

Kha might not DESERVE Horse's loyalty (which is very rare among horses) and he may not know what to do with it ... but it comes at a critical moment.  If Horse dies, Kha should feel it in spite of himself, if only for helping him save Sil--or, if she will survive death, his own life in which he may enjoy Sil.  (This may provide some of the changes in Kha that spook Anver.  Kat might understand better.)  And Sil might have something to tell Kha about gratitude to Horse.  It might even echo what Airen--and even Marion--tell Kha.

What would Airen be like mimicking her mother?  How would she react afterwords?

If you use it, explain it.

Yes, you can reply to comments in X-line (but not highlight new comments).

Those delays seem to be a function of network connection quality.  I suspect that the protocol stack on Windoze can't handle glitches in the POST operation well, but I haven't looked inside the machinery.  I don't know if there's anything in the web page programming that can help it.  Some internet commerce pages seem trouble-free; some 3rd-party blog engines (discus, esp.) seem to behave the same way.

Imagine that the Edsel was subcontracted to Yugo and built with Pinto and Vega parts, along with some of the glue they used in the Boston tunnel ceiling (the one that collapsed).  Its quality would exceed that of almost all software written today.

And when you're trying to make web programming work reliably, you depend on the reliability of the Edselyugopintovega underneath, glue and all.  Makes you wish for baling wire and chewing gum. And duct tape.

Imagine that the Edsel was subcontracted to Yugo and built with Pinto and Vega parts, along with some of the glue they used in the Boston tunnel ceiling (the one that collapsed).  Its quality would exceed that of almost all software written today.

Congratulations again!

Why is is that when you are thinking, working productively on one things, four other ideas come to you?  And why, when you are roadblocked on one thing, you're roadblocked on everything?

Sigh.

Amy, I'll do two reviews for you.  I have to do a couple more for other people today, and that won't even catch me up.

Meanwhile, while I was working on a 'final' demo box for my flasher circuit--weeks ago, and I still have to cut the opening for the switch--I got the flywheel diode backwards.  It actually managed to flash, but drew so much current it was probably responsible for blowing the output of my little battery-powered power supply.  That has to be fixed, and I'd rather it be resistent to future damage of that sort.

It looks like the bootstrap transistor is blown to a short.  If it's not the bootstrap, it's the main output.  Either way, I'm redesigning that part, using multiple bootstrap transistors in parallel.  That will allow it to carry more current with less current on the 'main' output.  But ... paralleling bipolar junction transistors requires current sharing resistors.  The low values I need are only available in surface-mount, about half the size of a grain of rice.  Since all of that has to be mounted on the heat sink, I have an interesting physical design problem.  I know what I want, but not yet how to get it.

The main output transistor will also get a heatsink.  There's only one kind available for TO-92 package devices, and it's a flat teaspoon-sized blade, so I have to move things around on my little circuit board.  (I'm starting with a new board.)

I want to put a low-pass filter on the voltage-regulating section so that brief current spikes don't let the output drop as the internal resistance of the batteries cuts in--but I want to be able to switch it out.  It looks like it will take less board space with the parts soldered point-to-point on the switch.

The battery circuit switches were a beast last time.  They'll probably be a beast this time.

And now ... as I was trying to work on the physical design of the heat sink-and-bootstrap unit ('power block') I got nagged by the thought that I could still overload my main output transistor.  The solution is a current limited between the output and the current bootstrap.  The current source for the output transistor goes through a resistor that develops a voltage as the output current rises.  When it gets up around the working range for junction voltages, the current bypasses that sensing resistor, feeding the base terminals of the bootstrap transistors.  Their output--up to 200 times the current they receive--is added to the output transistor's current.

So the current limiter has to go between the collector of the output transistor and the current bootstrap.  It will drop probably about 1.4 volts, reducing the output voltage I can get with a given battery voltage.  It will also have interesting effects as the output transistor goes into saturation with its collector drive cut off.

So I need to test and tweak the limiting circuit (3 to 9 hours) and figure out how to squeeze it on the circuit board.  It needs just four nodes, but it involves two transistors tied to each other and the output transistor ... and that makes for a hard fit, especially since the output transistor has to sit in a way that will make the flat heatsink mounting work.  (The heatsink for the power block is a finned chunk of extruded aluminum.  I'd prefer a black finish, but this is what I got cheap.)

AND ... while I was working on that problem I suddenly got A Very Good Idea to increase the tangle Merran and Pausonallie will get themselves into.  I'm struggling to get stuff revised and get to that part.  And I can only remember one of the orders I need to place today ... sigh!

This after four days when I couldn't think long enough about one thing to make progress.

On to reviews!