Recently in zotz Category

so, i turned on silvia this morning for the first time in ages, i thought i'd get back into espresso now that i have a fancy new roaster. as soon as i turned it on, there was a spark from the PID and the PID went blank.

oh crap.

so i fiddled with some of the wires -- the wires into the PID never did stay in place -- and got the PID to power on finally. it said: UUUU.

after a whole lot of research, i discovered that this meant that the thermocouple input was out of range, on the high side. i checked the range, and it was getting a reading of 1000F or greater, otherwise it wouldnt say UUUU.

fortunately, i have a thermocouple-probed thermometer. i removed the probe, plugged in the leads from silvia, and got a reading of 68F, quite believable. so the thermocouple was okay.

during a call to spu (the EE) regarding how to test the inputs on the PID, i noticed a burnt out lead on the PCB of the PID. i'd fried it!

so later in the afternoon, i soldered a piece of wire to onto the PCB to fix the blown lead, and... it worked! the PID showed the correct temp reading. so i put everything back together, plugged silvia in, turned it on, and waited for the boiler to come up to temp...

nothing. it didn't budge. argh?

more disassembly, and close inspection of the PID, multimeter readings of all sorts all over the place. silvia's circuitry wasn't exactly fresh in my mind, but schematics and multimeter probes put it back in perspective -- better perspective than when i'd made the mod, since i've been learning electronics since then. well whoopdie do.

the current suspect is my solid state relay. it reads VDC on its input, and i see VAC on its VAC input terminal, but when it's "on" (and it has an LED to note the fact) i see no VAC on the output -- and the boiler doesn't heat up neither. there's continuity between the boiler and the SSR, and the boiler will heat up if i use the steam function (which uses the factory thermostat instead of PID). so the boiler is fine, the heating element is fine, everything seems to be fine but the relay, which i thought isn't a thing to break, but since it's solid state, maybe it could break if it got 120VAC on its VDC input, which could have happened when i fried it this morning.

once i get a new relay and get it all working again, i might make the PID wiring permanent (it's attached by screw terminals now) with hot glue or solder. i really dont want to go through this again, fun though it may be.

i built an adjustable voltage desktop power supply. i felt like a jedi building a lightsaber.

it's a power supply unit... from a more civilized time.

really, i do.

mmm! fluxy.

revision 1 of the Desktop Dingus is finished. it features:

- a green LED that turns red when you flip a switch
- a 4-digit 7-segment LED display that says things like "beer is good for you have some soon"
- 3 other switches that currently do nothing but are detected as on or off by the microcontroller
- a 2 port switchboard, ma bell style. yeah! also detected by the uC but doesn't do much else

as it happens, fabricating the dingus was kind of tough, with many trips to the hardware store for spacers, screws, nuts, etc. but hardest of all was getting the little stuff to work.

examples: discovering that I needed a pull-down resistor for my switches. it was not easy to get a single-color LED to light when the switch was thrown while not altering the detection by the uC of whether the switch was open. it was even less easy wiring up a 2-color LED that changes color at the flip of a switch -- while not altering the uC switch position detection. finally, it was really rather hard to do the above properly. i had to de-solder the LED subcircuit twice, destroying a transistor in the process (impatience).

rev 2 will probably feature a thermistor to tell the user ("the user", heh. me.) the ambient temp of the Desktop Dingus, but the main feature of rev 2 will be cooperative multitasking, to allow switch detection/actions to occur while the LED display is hard at work displaying stuff. this will be a software challenge, which i am better suited to tackle, but i expect electronics challenges to arise.

to assemble rev1, i had to use voltage dividers, a transistor, discover an inadvertent voltage divider in my circuit (whoops! i keep doing that...), pullup resistors, pulldown resistors, lots of crimp terminals, lots of soldering, and neither last nor least, an 8-bit shift register (i spent a couple hours trying to figure out how to control many switches from few uC pins and lit on the idea of the shift register, as suggested by an EE cohort. later, i found an arduino + shift register + many switches tutorial online. heh.).

anyhow, fun.

soldering must be easy, because i can't be good at it.

i've wired up 2 out of 8 segments on my 7-segment (har!) display, plus 2 headers (one for power, one for data/clock/load). everything works. some of the joints are ugly, especially where i soldered 4 wires to a single pin. some of the pads appear fused, but my continuity tester and my arduino attest otherwise. the layout is horrendous -- mostly because i have to wire up 36 pins on a 4-digit 7-segment display -- and there are wires everywhere since this isn't a PCB. but it's my first try and it works.

plus, i figured out how to use my de-soldering pump. i'm becoming familiar with the tools.

and to think i was always awful at "operation".

my arduino arrived yesterday. i dusted off my brain a little (since it's been 2 years since i've done any C, well, not really, i did one small bugfix to my C code a couple months ago and bungled it) and whipped up an LED flasher with teh arduino.

i've been banging my head against the desk trying to flash an LED with 2 transistors and a capacitor. arduino makes it a little easier. okay, a lot easier. flashing LED was too simple. I made mine talk in morse code.

following is the embarrassingly brittle source code (oh, how nice it was to use ?: again!):

naturally.

(in truth, one thing i don't need is another thermostat. still, plenty of potential for WATERPROOFED controllers doing stuff in the brewing process).

oh right, i need to brew again, soon.

i went to the "maker faire" over the weekend and saw a couple of cool things. more importantly, though, one day after i'd gotten started learning "the basics" of resistors and batteries and diodes, i got a glimpse of the big picture: microcontrollers!

for some reason, all these years, I have not been doing things with microcontrollers. why not? I'm a lapsed C programmer, and if I may toot my own horn, I'm actually a rather good C programmer. at least, I was, back before I switched to python. now I have my doubts. still, looking at some of the online code samples for microcontrollers, the stuff is rather simplistic. sure, maybe it's harder for more complex stuff, but... maybe not.

my main sack has a microcontroller already, and though he's the man for electronics, he doesn't really know how to program, and for some odd reason, hid the fact that he wanted some help until we were at the faire. so there's that avenue to pursue, and also, i've been looking over the much-hyped "arduino" (they were all over at the faire) and getting a bit jazzed about that.

i already have bits and pieces of a project in mind: a webpage controllable LED matrix to sit on my desk and display inane messages entered by random visitors to my visitorless website. originally, i wanted an LED matrix that would display a snow pattern coalescing into the word BEER. i can combine both, I reckon.

the funny thing is that I've spent several futile hours of frustrated hacking trying to get an LED to flash on my breadboard. it still doesn't work. i've got several more days of reading before it's got a chance of working. but I am pretty confident that were i to order an ethernet-enabled microcontroller and a pre-fab LED matrix, I could have my above project up and running in less than a week.

and then, maybe something useful!

last weekend i took a visit to a warehouse shop that sold all sorts of secondhand junk that local companies liquidated after going out of business. lots of computer chairs, monitors, keyboard, switches, racks, routers, everything.

also, lots of miscellaneous electronic parts.

i was there with my main sack, and he was rather curt with my questions about the purpose and functioning of various electronic parts. that got me thinking: why is my main sack such a sack? also: why don't i know these basic things about electronics?

answers: that's just the kind of sack he is, and there isn't a good reason.

so i ordered a pair of electronics beginner books (the theory being that they would probably both be lousy in one way or another but hopefully they'd complement each other in such a way that together they'd add up to one decent book) and got to reading. yesterday the sack, having been made to feel suitably bad about his rudeness, joined me at the local electronics store to pick out some parts to get me started on circuit building.

a slight detour: it turns out the local electronics store has a big shelf full of NOS (new old stock : original, unsold, 1940s era) vacuum tubes, which is just what i'd need to dink around with my new tube amp to achieve different sounds. a local tube source (assuming they have tubes i can use, i haven't memorized the varieties yet) is a major score. even if i can't use them in my amp, i'd like to bring my electronics know-how up to the level where i can construct a tube amp, and perhaps they'll have tubes for that.

in any case, they had much of what i needed, but not everything. i reluctantly went to radio shack, thinking that they'd suck for that sort of thing just as much as they suck for everything else. i was pleasantly surprised, they're a goldmine for hobbyist electronics, just as they were 20 years ago. i should have gone there first, but then i would not have gotten the warm fuzzies from supporting my local shop (established 1961).

later on we visited Frys and were majorly disappointed. fry's is rubbish, i don't know why i keep thinking otherwise. maybe it's just the PA store. i dunno. maybe i'd go there for an enclosure or something, which looked harder to find at RS, but more likely I'd just order online.

Anyhow, i finally got to tinkering. i got a basic LED circuit working (lesson learned: the D in LED stands for "don't forget it's a diode not a lamp") then moved on to a switched SPST LED circuit, then a SPDT switched dual LED circuit, and then, the biggest challenge so far, I figured out how to design a circuit for my red/green LED, using a DPDT switch (none of the switches were labeled to suggest proper wiring, and by this point, i'd already blown the fuse in my meter, so i had to use reasoning and meterless testing methods to get the layout right).

these circuits are extremely basic, but at the same time, i've only been at it one day. and, though my switched LED circuits are basic, they're quite useful. the last two circuits i made could be used to indicate which checker is ready to ring up your crap at fry's.