#garfield Logs

Jul 04 2015

15:40 any27653564 geese, all the storms that went thru this morning are backing up to run me over again
15:40 any27653564 and i just got over the sinuses last nite
15:41 any27653564 this morning, the storms traveled from northwest to southeast, now they are from southwest to northeast
15:41 any27653564 overflowed the outside tank, even tho i had closed the roof diversion to it
15:42 any27653564 if i had been allowed to complete the actual real pool, this would have been a complete year's water supply
15:43 any27653564 why not complete it: just walking out there caused the gods to bark, for which i was threatened with jail time,, for walking 50ft from the property line, on my own property
15:52 any27653564 saw a deer up on the road the other day, no doubt making people think me, a 59 yr old overweight disabled woman, was "wandering around, hiding in the bushes, spying on people"
15:52 any27653564 it was a beautiful animal
15:54 any27653564 it wasn't spray painting obscenities on my gate, shitting or having sex in my driveway, making earsplitting barking noises, or threatening me in any way
15:55 any27653564 i *hate* that i was forced to put up fence that keeps them off my property around the house
15:55 any27653564 the deer and the rabbits were my lawnmowers
15:59 any27653564 if i offset $200/mo from the grocery store by having my own garden and real farmette, i'd have free up $36,000 over the last 15 yrs, which was half the reason for buying the land, it was an investment, that would have paid for itself in groceries, physical exercise, and play space for outside robots
15:59 any27653564 it's why i was looking forland with a small stream, like this
15:59 any27653564 water, for gardening, pool, a/c cooling
16:00 any27653564 fuck
18:53 rue_shop2 hi
19:02 Tom_itx lo
19:47 rue_shop2 irsend SIMULATE "000000000000d827 00 Stop koss.cfg"
22:35 any16403519 engineering in the usa: they put a utility pole in the middle of one lane of traffic : https://www.google.com/maps/place/St.+Johns+Bridge/@45.5873749,-122.7597405,55m/data=!3m1!1e3!4m2!3m1!1s0x549507fd0ac4c1b7:0xad530d182be3b203
22:36 any16403519 here's the street view of it : https://www.google.com/maps/place/St.+Johns+Bridge/@45.5872437,-122.7594196,3a,75y,330.89h,90.32t/data=!3m7!1e1!3m5!1sqdKFVzt1xxUElMQqYjDEdw!2e0!6s%2F%2Fgeo1.ggpht.com%2Fcbk%3Fpanoid%3DqdKFVzt1xxUElMQqYjDEdw%26output%3Dthumbnail%26cb_client%3Dmaps_sv.tactile.gps%26thumb%3D2%26w%3D100%26h%3D80%26yaw%3D240.39249%26pitch%3D0!7i13312!8i6656!4m2!3m1!1s0x549507fd0ac4c1b7:0xad530d182be3b203!6m1!1e1
22:36 any16403519 umm
22:36 any16403519 http://tinyurl.com/pz7jzbv
22:37 any16403519 here's the far-away view of it: https://upload.wikimedia.org/wikipedia/commons/thumb/1/13/St._Johns_Bridge_%28Portland%2C_Oregon%29.jpg/1280px-St._Johns_Bridge_%28Portland%2C_Oregon%29.jpg
22:38 any16403519 question is: knowing they put utility poles in the road, not by accident, and were fine doing so, do you trust the bridge?
22:40 rue_shop3 kat!
22:40 rue_shop3 there you are
22:41 rue_shop3 I need help
22:41 rue_shop3 its prolly just a symantic problem I'm having
22:41 rue_shop3 how far in the road?
22:42 rue_shop3 the bridge dosn't appear to have any glaring errors, how the maintenance?
22:42 rue_shop3 here is my problem
22:42 rue_shop3 I need an analog ciruit, that will output the lesser of its two analog inputs
22:43 rue_shop3 oh , heh, maybe just a few diodes eh?
22:44 any16403519 http://tinyurl.com/p87o9yn
22:44 any16403519 yeas, diode pull down, resistor pull up
22:44 rue_shop3 does the sign say something about not hitting the pole?
22:45 any16403519 analog 7407
22:45 rue_shop3 it looks like a parking shoulder to me :)
22:45 any16403519 can't really make it out
22:45 any16403519 i saw it in a movie, it looks like it's in the road
22:46 any16403519 if you go fartehr down the road, the lines of the road seem to stay straight, but they moved the other poles off the road
22:46 rue_shop3 :)
22:46 rue_shop3 someone will encourage a relocation at some point
22:47 rue_shop3 heh, google false detected some windows as a cars front bumper and blurred them out
22:47 any16403519 if you put a equal diode in the feedback path of each of the opamps, , they will compensate well for the dc drop in the diode in the output path, if that helps
22:48 rue_shop3 I was thinking of looping it around with a amp to do impedence isolation and drop compensation
22:49 any16403519 search for no-loss peak detector, then reconfigure for negative output and two inputs,, if i remember those circis right
22:49 any16403519 you mean for speakers?
22:49 rue_shop3 I'm trying to work out, logistically, the voltage/current controller for the bench supply I'm making
22:50 any16403519 logistically? "there it is!" , presto, logistics solved :-)
22:50 rue_shop3 I'm haivng gain problems
22:51 rue_shop3 I dont know how to work out how much to knock back the votage to get the current to where it should be
22:51 rue_shop3 because I dont know the load impedence
22:51 any16403519 if you are doing that for speakers, specially with bandpass chokes in the enclosure for woofy and tweety, they are non-linear as they saturate, so you will need to test at diff freqs and pwrs
22:51 rue_shop3 I'm playing this as a voltage limited current source for a bit
22:52 rue_shop3 its a bench power supply
22:52 any16403519 umm, set the circuit to servo the current instead of the voltage
22:52 rue_shop3 low freq to really low freq
22:52 rue_shop3 yea
22:52 any16403519 technically, it's the same circit, but using the current instead of the voltage to servo around
22:52 rue_shop3 so, if I have the main control signal determined by current, and then make circuit select the min. of that or the voltage control signal for the output
22:53 rue_shop3 so, how do I work out what the control votlage should look like with current error
22:53 any16403519 ok, using open collector, you can use both at once
22:54 any16403519 why are *you* working it out? you don't even know the impedance it's driving
22:54 rue_shop3 yea, it could be 0.04 ohms, it could be 1000 ohms
22:55 any16403519 procedure: with nothing connected set the max voltage , then short the output and set max current, then connect your load and let the psu deal with it
22:55 rue_shop3 if you have 300mA error, how much do you cut back the drive signal..
22:55 rue_shop3 I'm designing the psu
22:55 rue_shop3 I'm chasing my tail on symantics I think
22:56 any16403519 oh, the gain you will haveto balance tween overdrive causing ringing and underdrive causing lack of correction, depends tons on your ciruit
22:56 e_shop3 stops to check who's tail this i
22:56 rue_shop3 hahah
22:56 rue_shop3 I'm trying to avoid both those problems
22:56 rue_shop3 one of my ideas is to use a comparitor and a integrator
22:56 any16403519 make the voltage feedback work first, it's easy and conventional, that will prove something works
22:57 rue_shop3 I can get that working, but every design I'v had to throw out all that work for the current limiting
22:57 any16403519 then make the current sense and hook it to the voltage feedback, behind the voltage sense
22:57 rue_shop3 and the best results are hell bent ringing
22:58 rue_shop3 I cant use the LM723 circuit in the old supply because I cant votlage control it
22:59 any16403519 or,, and this is cheating, make conventional adjustable current sense , use that signal to clobber the voltage feedback, just make sure that if the Cfeedback kills the Vfeedback signal that no smoke is released
22:59 rue_shop3 I think part of the problem is that I have a voltage controled amp in the output of a circuit trying to regulate current
23:00 any16403519 no
23:00 rue_shop3 well, if I have the conventional current source, I can take the signal from it, and do that min-voltage-selector with the output voltage control
23:00 rue_shop3 and I'll get the min resulting signal from the current source and the voltage limit
23:01 rue_shop3 if the load is .04 ohms, the gain when adjusting the votlage on the output driver is crazy
23:01 rue_shop3 (just an evil example)
23:01 any16403519 go with your diode approach, whichever sense is needing to reduce the voltage drive (either to limit the voltage or to limit the current), it will happen
23:01 rue_shop3 there is another road
23:01 rue_shop3 pwm, dont pulse if the output current is over limit...
23:02 any16403519 you are maybe overthinking it
23:02 rue_shop3 I know I'm overthinking it
23:02 rue_shop3 but sofar I cant get anything to work
23:02 any16403519 you cannot adjust the voltage even?
23:02 rue_shop3 I can
23:03 rue_shop3 I suppose I didnt' mean anything literally
23:03 any16403519 <rue_shop3> but sofar I cant get anything to work
23:03 any16403519 o
23:03 rue_shop3 I meant any of the supposed complete systems dont work
23:03 any16403519 well, show someone the schematic
23:03 any16403519 maybe that way they'd would have a clue what you are trying and failing
23:03 rue_shop3 I have a 1" pile of pages with schematics here
23:04 any16403519 oh
23:04 rue_shop3 switching should make everything nice, if something is over, dont pulse
23:04 rue_shop3 but I cant get any of the switching controlers to regulate with less than 200mV ripple
23:05 rue_shop3 and none of them like the idea of going down to 0V
23:05 any16403519 throw more capacitance on the output, parallel caps, not one big cap
23:05 rue_shop3 they still schmitt down 200mv before deciding to turn on again
23:06 any16403519 ok, screw the switchers, tell them to regulate to 1v more than you need, then use a mosfet to analog regulate the 1v surplus
23:06 any16403519 or 1/2 volt
23:06 any16403519 <shrug>
23:06 any16403519 switchers be noisey, it's th nature of the beasts
23:06 rue_shop3 yes, or (even with some analog that wont behave) move the 0V ref down a few volts synthetically
23:07 rue_shop3 it thinks its outputting 2V, but its 2V from -2V
23:07 any16403519 a big arese diode, but those can be way pricier than a mosfet, surprisingly
23:07 rue_shop3 you just bump the amps reference down -2V
23:07 rue_shop3 a few mA at worst
23:08 rue_shop3 it thinks 0V is 2V
23:08 any16403519 or, it thinks 0v is .7v
23:09 rue_shop3 a cap between an amps - and output makes it integrate right?
23:09 any16403519 yeas, slows dow the edges
23:09 any16403519 low-pass
23:09 any16403519 dampening
23:10 any16403519 series R would help lots
23:11 any16403519 bringng noisey power to an audio pcb, feed pwr to 100 ohm R, then to the 1000u cap
23:11 any16403519 all the noise volts get dropped across the R
23:15 rue_shop3 Id like to look at the lm723 circuit in my existing bench supply again, but its current limiting has some odd quirks
23:17 any16403519 k
23:21 any16403519 https://duckduckgo.com/?q=%22adjustable+current+source%22
23:22 any16403519 https://duckduckgo.com/?q=%22adjustable+current+source%22+schematic
23:22 any16403519 hard to get any simpler than this one : http://www.merghart.com/p/28/opamp-adjustable-current-source
23:22 any16403519 and, it's what i wold do
23:24 any16403519 now, to merge the voltage regulator, put a diode on the output of that opamp, so it can pull down, then glue in adjuable voltage regulator that the current opamp can clobber
23:25 any16403519 they have that CA3130 operating way close to the negative rail, i dunno that's a good idea, but basically, the circuit is good
23:26 any16403519 the two clamp diodes acting as a zener is real temperature sensitive, you might improve on that
23:30 any16403519 classic way to make a 3-terminal regulator (LM338, 7805, etc) be a current regulator : http://html.alldatasheet.com/html-pdf/22757/STMICROELECTRONICS/LM338/12987/8/LM338.html
23:31 any16403519 your task, if you choose to accept it, is to use an opamp to compare that connection at the top of LOAD to the wiper on your adjustment pot, and servo the ADJ pin on the 3-terminal accordingly
23:31 rue_shop3 the current sources level will be set by a microcontroller, with a 1.2 to 3.7V analog output
23:32 rue_shop3 I should build that just to see if its stable
23:33 any16403519 k
23:42 any16403519 water contact probe: 1/4 R glued to a to-92 LM36, in air the temp will read high, soon as it its water, you will know because the temp readout will drop
23:42 any16403519 chose R to warm the LM36 a few degrees
23:42 rue_shop3 cool
23:43 any16403519 this works best if the water is colder than the air
23:43 rue_shop3 :)
23:43 rue_shop3 I told you about my capacitive sensor?
23:43 rue_shop3 I get a resolution of about 500 counts/inch
23:43 y16403519
23:43 rue_shop3 about 7 counts of instability in the signal
23:48 rue_shop3 I dont get this, on the bench supply I have, they use a current source to set a votlage on the current setpoint pin, instead of just using a pot to put a votage on it
23:54 any16403519 the barrels inside thwe plugs on the capacitance meter can push in and bottom out in the case, below the pin of the jack
23:56 any16403519 each time i put in the leads, the barrel is pushed a lil more down, and it doesn't pull back up all the way when removng the lead
23:57 any16403519 eventually, the meter stops working because the barrel is too far down for the banana jack to make contact
23:59 any16403519 or the mechanical movement breaks the solder connection to the pcb