#garfield Logs
Mar 16 2015
#garfield Calendar
00:08 katsmeow-afk too many can bea problem
00:10 katsmeow-afk looks like no hurricanes for Rif, weather radar from here says he's got blue skiys for at least a week
00:11 rue_house if australlia isn't drowning, its burning
00:15 tsmeow-afk just had a fight with the browser to prove "abc" != "american broadcasting c
00:17 katsmeow-afk http://www.abc.net.au/news/weather/
00:19 katsmeow-afk rue, glad you talked me into seeing Avatar
00:28 rue_house I did?
00:28 rue_house k
00:28 rue_house what would you use for a stable 0.2V ref?
00:29 katsmeow-afk an opamp, couple resistors, and a stable 2.1v ref
00:29 rue_house hmm
00:29 rue_house what would you use for a stable 2.1V ref?
00:30 rue_house :)
00:30 katsmeow-afk else, a few resistors and a TL431 to push a fixed current thru one of the resistor
00:30 rue_house hmm
00:30 rue_house gnight
00:31 katsmeow-afk The TL431/TL431Aare three-terminal adjustable regulator
00:31 katsmeow-afk series with a guaranteed thermal stability over applicable
00:31 katsmeow-afk temperature ranges. The output voltage may be set to any
00:31 katsmeow-afk value between V REF (approximately 2.5 volts)
00:31 katsmeow-afk so set it for 2.5v, resistor divide it to 0.2 , buffer with an opamp
00:33 katsmeow-afk else, use a 78L05, , resistor divide it to 0.2 , buffer with an opamp
00:33 katsmeow-afk cheap, easy to find, 4 parts total
00:34 katsmeow-afk 20ma output buffer with the opamp, and adjustable, if you like
00:36 katsmeow-afk if you need precision, less than 5mv drift over 30v input or 0c-100c temp variation, see: https://duckduckgo.com/?q=bandgap+reference
00:39 rue_bed oh LM317
00:39 katsmeow-afk ok, that's an older design and may be more drifty
00:39 rue_bed oh
00:40 katsmeow-afk depends on what you need <shrug>
00:41 rue_bed theLM10 is an op-amp with a .2V ref,
00:41 katsmeow-afk nowadays there's tricks to make it more stable, like temperature control, voltage regulation, etc
00:41 rue_bed as they aren't availableanymore
00:41 katsmeow-afk can you find an LM10 ?
00:41 rue_bed no, too old
00:41 rue_bed well, I couldn't before
00:41 katsmeow-afk ditto the LM11 and 12
00:42 katsmeow-afk good ideas, but priced too high
00:42 rue_bed but there are prolly lots of comparable amps, just not with .2V refs
00:43 katsmeow-afk use what you got: to-92 volt regulator are temp compensated, resistor divide down, then opamp buffer
00:44 rue_bed isn't the BE of a bipolar usually .2V?
00:44 katsmeow-afk no, but shotckys may be
00:45 katsmeow-afk the junctions are not temp compensated, and you need a stable current supply
00:51 katsmeow-afk yu can get lower volt refs, but they may not be useful, they need buffering anyhow, and they may be pricey
00:51 rue_bed http://www.play-hookey.com/analog_experiments/preparations/ps_op_amp_control.html
00:51 katsmeow-afk i know of a 1.2v ref, but i cannot think of the name atm, sorry
00:52 rue_bed tl4xx?
00:52 katsmeow-afk you will need to divide down the 1.2 annd bufefr it anyhow, so just use the TL431 and 3 resistors, or a 78L05
00:52 katsmeow-afk maybe
00:53 katsmeow-afk you canget a 78L05 or 78L33 or etc in smd, ditto the opamp and 2 resistors
00:54 rue_bed have you seen that output transistor arrangement before?
00:54 katsmeow-afk you can always tweak it with a avr
00:54 katsmeow-afk looking........
00:56 katsmeow-afk yeas, it's old and common and typical, but the 2.2r is horribly high and willwaste power
00:56 katsmeow-afk the TIP44 isn't going to survive current limiting at 17v acorss it
00:56 rue_bed its hurting my brain looks like a SCR arrangement
00:57 katsmeow-afk and thee's no protection for blowing the 3904's BE junction, and you won't know it's blown till thepsu blows something else
00:57 rue_bed hmmm
00:58 rue_bed oh its a self current limiting output driver?
00:58 katsmeow-afk yeas
00:58 rue_bed ooh I was trying to see how it was assisting the base current, haha
00:59 katsmeow-afk it begins limiting when the volts across the 2.2r turn on the 2n3904 and it clamps downthe tip44's base drive
00:59 rue_bed yup, I see now
00:59 katsmeow-afk problem is, the tip is slow, and you can short the output and blow the 3904 before the tip drops the current
01:00 katsmeow-afk 2.2 ohms in that cirecuit means the output is 300ma
01:00 rue_bed looking for a ref design for a variable current limited power supply
01:00 rue_bed not seeing much good out there
01:01 katsmeow-afk and with no gain in the 3904 pathway, it's a pretty soft limit, and it's fighting the opamp, which has no clue that it should be limiting
01:01 katsmeow-afk how much current do you need?
01:02 rue_bed 5A
01:02 rue_bed 30V, 36 max available raw
01:05 katsmeow-afk http://www.microsyl.com/projects/powersupply/powersupplysch.pdf
01:07 katsmeow-afk http://www.eleccircuit.com/regulator-0-30v-5a-by-ic-723-2n3055-2part/
01:08 katsmeow-afk i recommend either a current fue circit, or current foldback, because if it current limits into a short, it wil be dissipating 180 watts
01:09 katsmeow-afk dosen ways to do that, prolly small avr is best because it's programmable, you can hook a pot to it's adc for front panel adjustments
01:09 katsmeow-afk also adc a thermistor on the heatsink, and tell how much time has elapsed in the shorted situation
01:16 rue_bed your good
01:16 tsmeow-afk
01:25 rue_bed thats as close to what I was looking for as I ever expect to find
01:26 rue_bed looks like they have a design fault, under 5V current limiting wont work right I dont think, my current sensor solves the issue they have
01:26 rue_bed nono wait, its on the unregulated side
01:26 katsmeow-afk word of warning, the ore complicated it is, generally the slower it is too
01:26 rue_bed which means their drive current must come out on the output
01:27 rue_bed hmm
01:28 katsmeow-afk my most powerful psu for under 100vdc used MJ11016's, and overcurrent shutdown was a 74F74 latch
01:28 rue_bed the shunt resistors might take care of it
01:28 rue_bed 11016,,,,, not familiar
01:28 rue_bed I'm guessing its a darlington
01:28 katsmeow-afk 30amp bipolar
01:28 katsmeow-afk yeas
01:29 rue_bed 5 digit seems to usualyl be darlington :)
01:29 katsmeow-afk i had to use a darlington because it had to go into a existing backplane and use the voltages available
01:29 rue_bed did you see how I used a 555 for the current trip-out on the hexapod?
01:29 katsmeow-afk MJ15003(?) isn't darlington
01:29 rue_bed oh
01:30 rue_bed the 555 was the set/reset AND the relay driver :)
01:31 rue_bed and signalled the cpu that it'd tripped
01:31 katsmeow-afk 15012/23/24/25 are not darlingtons
01:31 rue_bed hmm, I guess I'll throw out the 5 digit notion
01:32 katsmeow-afk feeling the drugs kick in
01:33 katsmeow-afk strangely, because there's several very important variables, almost every psu is custom
01:34 katsmeow-afk input power, output power, operation speed, stability, heatsink size, airflow, time to design, manufacturing cost, etc etc
01:35 katsmeow-afk for instance, if i go over 20amps out, i can expect the banana plugs to overheat
01:36 katsmeow-afk 10 yrs ago, 35v was the cost knee, 5 yrs ago, 80v caps were the knee, now there's cost effective 100v caps
01:38 katsmeow-afk i mean for 10,00ufd caps
01:38 katsmeow-afk 10,000
02:03 katsmeow-afk zzzzzzzzzzzzzz
09:04 rue_house the trouble with the LM723 is how it does the current cutoff, its hard to control with a 0-5V signal
13:52 tsmeow-afk hands rue a transistor tocontrol with a 0-5v si
14:09 tsmeow-afk ponders if time running slower in a gravity well is due to the gravity compressing the stuff the time-measurer is made of, making them slower due to friction between the parts, not to time itself slowing
14:27 katsmeow-afk damn at the bad pronunciation,, spigetti has something to do with spigots?
14:28 katsmeow-afk they have a tiday and a tuhday, but nothing happens today
14:29 katsmeow-afk i keep tripping over this stuff, haveto error correct, and the speech backs up and is dropped
14:35 katsmeow-afk the wood is on far? Go get it so we can make a camp fire!
14:39 katsmeow-afk a "feint heart" is not a "faint heart" , feint and fait are both words, but they mean different things
14:39 katsmeow-afk faint
14:41 katsmeow-afk "whirled' != "world"
14:41 katsmeow-afk "wore" != "war"
14:47 tsmeow-afk sees whole sentences go by with every word mispronou
14:48 katsmeow-afk news spot the other day, i watched the spot on the 5pm, 6pm, and 10pm news shows, i never figured out what the guy said, and he said it several time, repeating himself in the spot
14:52 katsmeow-afk i suspect the way to "get ahead" is to know just enough to get by, then talk real fast with a fake accent, use prefixes and suffixes to screw up the syntax, point a lot, and look exasperated as if everyone else is too stupid to follow your gibberish
14:53 katsmeow-afk if someone else figures out the correct way to do something, re-state it in your gibberish and say that is just what you said last week
14:55 katsmeow-afk i knew someone who almost did that, but punctuated each statement with a laugh as if she was laughing at those who didn't understand how obvious it was, and she got made head of the dept, but hadn't a clue and had to study up on the buzzwords of the dept every night
14:56 katsmeow-afk i felt dirty, somehow, just knowing her
17:32 katsmeow-afk ok, companied hire people, pimps whore people, but what is it to "har" people?
17:38 katsmeow-afk "if'n you wanna git har'd, go ober dare"
17:38 katsmeow-afk what?
17:38 katsmeow-afk <louder>"if'n you wanna git har'd, go o'er day air"
17:39 katsmeow-afk thanks, that cleared it up
17:40 katsmeow-afk <kid> chewbacca?
17:40 katsmeow-afk <me> what?
17:40 katsmeow-afk <kid> chewbacca?
17:40 katsmeow-afk <me> (looking at mother) what is he saying?
17:40 katsmeow-afk <mom> he is asking if you chew tobacco like his grandmother does
17:41 katsmeow-afk <me> (looking at kid) no
17:41 katsmeow-afk <kid> (looking at mom) what?
20:27 rue_house hired for what?
20:36 katsmeow-afk i dunno
20:38 katsmeow-afk this is in Van?? : http://imgur.com/gallery/SWzGl
20:39 rue_house ...
20:40 katsmeow-afk Technically it's been in the harbor for 30 years. Abandoned for only 25.
20:41 rue_house not that I'm aware of
20:41 rue_house so, lets check google earth
20:41 katsmeow-afk The McBarge, officially named the Friendship 500, was a McDonald's restaurant, built on a barge for Expo '86 in Vancouver, British Columbia. Moored on Expo grounds in Vancouver's False Creek, it was the second floating McDonald's location in the world (first being in St Louis, Mo),
20:42 rue_house well I know where false creek is
20:42 katsmeow-afk https://en.wikipedia.org/wiki/McBarge
20:43 katsmeow-afk well, how bout that
20:45 rue_house cant find it
20:45 rue_house oooh burrard inlet
20:46 rue_house dont know where that is
20:46 katsmeow-afk i wonder why nothing is done with it
20:49 rue_house found it
20:50 rue_house https://maps.google.com/maps?q=burrard+inlet&hl=en&ll=49.293033,-122.979594&spn=0.001482,0.003307&sll=49.280012,-123.160601&sspn=0.03354,0.105829&t=h&hnear=Burrard+Inlet&z=19
20:51 rue_house off the end of penzance Dr
20:51 rue_house way off
20:54 rue_house http://www.microsyl.com/projects/powersupply/powersupplysch.pdf
20:54 rue_house I wonder if an lm324 can take 36V
20:55 rue_house eeek, its the abs max rating..
20:55 katsmeow-afk i gave you that url at 00:44 this morning
20:55 rue_house yea, I'm transfering it to my shop
20:55 katsmeow-afk o
20:56 rue_house I need to find an opamp I can use
20:56 rue_house tl074?
20:56 katsmeow-afk no
20:56 rue_house +-18V thats 36V
20:56 rue_house why not?
20:57 katsmeow-afk because at 36.1v it blows up
20:57 katsmeow-afk some OP series are good to 100v
20:57 rue_house they are running it on 32
20:57 rue_house hmm I'll check what I have
20:58 rue_house know whats got 4 amps?
20:58 katsmeow-afk in an opamp? well, audio power amp modules
21:01 katsmeow-afk look for a 40w power output module , might need to use 2 of them
21:02 katsmeow-afk AD712 , and some other AD opamps, are over 50v power supplies
21:02 katsmeow-afk er, wrong number
21:05 katsmeow-afk TDA7294
21:05 katsmeow-afk $3 on ebay
21:06 katsmeow-afk TDA7292
21:07 rue_shop3 AD712...
21:09 rue_shop3 wonder what XP4151 is
21:09 katsmeow-afk i dunno, but for $3 a 100v opamp that delivers amps, good nuff deal
21:10 katsmeow-afk rated normal +/- 40v , 100w delivered
21:11 katsmeow-afk the TDA7292 is a dual version at 1/2 the power per channel
21:11 rue_shop3 uA3403 ?
21:11 katsmeow-afk short circuit protexction
21:11 katsmeow-afk dunno
21:11 rue_shop3 TA75902
21:12 katsmeow-afk i was looking for cheap and easy to get
21:12 rue_shop3 BA10324 ... hey... 324?
21:12 rue_shop3 these I have...
21:12 katsmeow-afk LM324 is 30v
21:13 rue_shop3 lots of LM stuff
21:13 rue_shop3 TDA or AD
21:14 e_shop3 checks the medical and industrial ba
21:14 rue_shop3 AD595
21:14 katsmeow-afk The TDA7294 is a monolithic MOS power ampli-
21:14 katsmeow-afk fier which can be operated at 80V supply voltage
21:14 katsmeow-afk (100V with no signal applied)while delivering out-
21:14 katsmeow-afk put currents up to ±10 A.
21:14 rue_shop3 CA124
21:15 katsmeow-afk 1PCS TDA7294 Audio Amplifier With Mute IC
21:15 katsmeow-afk $2.83
21:15 rue_shop3 ... 10A?
21:15 rue_shop3 dc coupled?
21:15 katsmeow-afk i copy/pasted
21:16 rue_shop3 I'ma lookin over baords
21:16 rue_shop3 do you know AD595?
21:16 katsmeow-afk spec is +/- 29v 6.75 amps , not sure where they get the "10amps"
21:17 katsmeow-afk no
21:17 katsmeow-afk Supply Voltage (No Signal) ±50 V
21:17 katsmeow-afk I O Output Peak Current 10 A
21:25 rue_shop3 TLC279
21:25 rue_shop3 I think if its not part of a vcr, printer, or monitor, I dont have it
21:25 rue_shop3 ...
21:25 rue_shop3 LA3220
21:26 katsmeow-afk you asked for amps in a 36v output, best bet, for cheap, is audio output modules like those TDAs
21:28 katsmeow-afk if there's not enough amps, you at least have enough to drive the bases of a flock of bipolars to get 10x more amps
21:29 katsmeow-afk 4amps from the output module, driving bipolars with a gain of 10 gives you 40amps
21:29 katsmeow-afk driving power darlingtons at a gain over 100 gives you moreamps than you have psu for
21:42 katsmeow-afk driving mosfets may be difficult, i dunno what the gain on the feedback pin is
21:52 rue_shop3 amps that could take 32V power
21:52 rue_shop3 I'm wondering why the amps in that schematic need to be fed with more than say, 9V
21:53 rue_shop3 all the inputs are 0-5V
21:53 rue_shop3 that said, 12V is a handy voltage
21:58 rue_shop3 I wonder what I should use if I dont need a 32V power rail
22:04 rue_shop3 the 6152 goes to 24V
22:04 rue_shop3 er 6482
22:06 katsmeow-afk yanno you can always use any common opamp with a separate hv gain stage of 3 or 4 transistors
22:07 katsmeow-afk opamp pulls up on the base of a npn on the -prw, and it's collector pulls down on a pnp-npn bases pair , a resistor pulls up on their bases, and their emitters are the output
22:08 katsmeow-afk you can easily do 200v and many amps that way
22:08 katsmeow-afk resistor divider back to the opamp (-) input
22:09 katsmeow-afk it's a common power amp schematic
22:09 katsmeow-afk if you don't need the pull-down, then don't install the power pnp pulldown at the output
22:11 katsmeow-afk use bipolar or mosfet power pullups
22:14 rue_shop3 in ym simulation of that circuit, R23 just blew up
22:15 rue_shop3 I'm just doing that npn isolated now
22:15 rue_shop3 from what I can tell, at 3A out, each driver needs 50mA, total 150mA
22:15 rue_shop3 the series 200R is rated for 2W, at 150mA its pushing 4W
22:16 katsmeow-afk that's a gain of only 6 , that's aweful
22:16 rue_shop3 20
22:16 rue_shop3 1A each, hfe of 20
22:16 rue_shop3 x3
22:16 katsmeow-afk oh
22:16 rue_shop3 wait... this gets worse
22:16 tsmeow-afk misremembered the schematic, s
22:17 katsmeow-afk but 20 is still bad
22:17 rue_shop3 1A across .33R
22:17 katsmeow-afk why use 3 transistors for only 3amps out?
22:18 katsmeow-afk at 3amps, the MJ1013,14,15,16 have a gain of 500
22:19 katsmeow-afk the MJ15022,24 has a gain of 60 at 3amps
22:19 rue_shop3 200R, 100mA
22:20 rue_shop3 I ran out of votlage to drive the base
22:20 katsmeow-afk the MJ11015,16 stays above 900 gain from 0 to 20amps
22:20 rue_shop3 they used D44VH10, which has a min gain of 20
22:21 rue_shop3 I'm checking the base drive supply
22:21 rue_shop3 they have 32V to drive a 30V transistor
22:21 rue_shop3 so, from the output trnasistor
22:22 rue_shop3 back thru the base resistor and the darlington
22:22 rue_shop3 we need 2V
22:22 rue_shop3 which seems wrong
22:22 rue_shop3 just the dralington drop is 2.1V
22:23 rue_shop3 they ahve 3 BE junctions
22:23 katsmeow-afk 2
22:23 rue_shop3 THEN they go to a 200R resistor, then they go to a base that needs 50mA
22:23 rue_shop3 3904 + the two in the DBW93
22:24 katsmeow-afk o
22:24 rue_shop3 I see 2.1V drop there,
22:24 rue_shop3 then 20V drop in the resistor
22:24 rue_shop3 leaving them short 20.8V to drive the output transistor
22:25 katsmeow-afk this is why i design my own stuff
22:25 rue_shop3 I wanted this for the control section, but I'm scrutinizing
22:26 rue_shop3 this is the schematic for a commercial supply isn't it?
22:26 katsmeow-afk wanna know what i'd do instead of the ua732 ?
22:26 katsmeow-afk i dunno
22:26 rue_shop3 I had a UA732!
22:26 katsmeow-afk i've smoked commercial psu often enough
22:27 rue_shop3 I'm loking at the LMC6482 with an NPN on the outputs
22:27 katsmeow-afk idiot RCA bench supply had "3amp output" with a 2amp transformer, smoked it
22:27 rue_shop3 pff
22:27 rue_shop3 should expect better from RC
22:27 rue_shop3 a
22:28 rue_shop3 this is an emitter follower with a 2.1V drop
22:28 rue_shop3 and 2V overhead
22:29 rue_shop3 I wanted to put an emitter resistor on the NPN to trim the gain a bit, but everything fell apart when I was trying to work out how much current it needed to handle
22:29 katsmeow-afk build a dead simple straight audio amp, leave off the pulldown, put a cuurent sense on the vcc to the npn power-out MJ11016, use that OC signal to limit and trip a avr to wait some calculated seconds before killing power
22:30 katsmeow-afk simple, few things to smoke, cheap, low parts count, and functional
22:30 katsmeow-afk tons of overkill
22:30 rue_shop3 heh, I dont want the avr involved in the safety
22:31 katsmeow-afk you really wann aimpress, use two MJ15024 instead of the 11016 and appropriate driver
22:31 rue_shop3 I made up a nice simple current sensor, works down to 0V, outputs 0-5mA at 1mA/A
22:31 katsmeow-afk ok, an opamp and a cap, and a thermistor to mind the heatsink
22:33 katsmeow-afk add an external input for gating the power, making a ramp, sawtooth, etc, like driving it with a function generator
22:33 rue_shop3 another thing I wrested with on my design, was that their base drive current comes out the output
22:33 rue_shop3 so I used a P channel fet
22:34 rue_shop3 no drive leakage
22:34 rue_shop3 hmmm
22:34 katsmeow-afk ok, design using MJ150xx and a tip147 as a szilaki pair
22:34 rue_shop3 maybe I can still use my P channel fet as an output stage in linear
22:34 rue_shop3 szilaki?
22:34 rue_shop3 is that with the PNP driving the NPN?
22:34 katsmeow-afk checking spekkinh
22:35 katsmeow-afk or vice versa
22:35 rue_shop3 cool
22:35 katsmeow-afk https://en.wikipedia.org/wiki/Sziklai_pair
22:35 katsmeow-afk so all your outputs's's base drive goes to output
22:36 katsmeow-afk terrific gan with the szlaieuir thsala pair and that darlington tip147
22:36 rue_shop3 but I dont want the base drive on the output
22:36 katsmeow-afk why not?
22:37 rue_shop3 at small currents or votages it'll pull it up
22:37 katsmeow-afk ?
22:37 rue_shop3 ths is a 5A/channel supply, but I want to set it to 10mA to drive leds
22:37 rue_shop3 if I so choose
22:38 katsmeow-afk umm, the opamp sets the output voltage, you shold be easily able to goto 0v and 0ma if you like
22:38 rue_shop3 I reasoned that even 1V would be an acceptable min output voltage, but I want zero
22:38 rue_shop3 in one of the prototypes I had drive power pulling the output votage up
22:40 tsmeow-afk s
22:40 rue_shop3 your tellin me!
22:40 rue_shop3 I thought this would be easy
22:41 rue_shop3 see how their current sensor is upsteam of the output driver? that means that its wrong by the base drive current
22:42 rue_shop3 when its pushing 3A out, the output is really 3.1A
22:42 katsmeow-afk if thebase drive is added backto the output, then it's not off
22:42 rue_shop3 maybe its linear error that can be canceled out
22:43 katsmeow-afk if you use darlington or szlaikai withenough gain (2k), then base drive of 3amps is only 1.5ma
22:43 katsmeow-afk that's under the limit of the cheap chinese digital meters
22:44 rue_shop3 the votlage feedback takes off 25mA at 30V
22:44 rue_shop3 ah!
22:44 katsmeow-afk ouch
22:44 rue_shop3 I can also just use the P channel fets I got
22:45 rue_shop3 the 1k takes off another 30mA
22:45 rue_shop3 they aren't making up for the 100-150 they put in
22:46 rue_shop3 ok kat, remind me how to make a linear fet amp :)
22:46 katsmeow-afk i would not use any schematic i find online, i do better myself
22:47 rue_shop3 its ok, I'm mauling it
22:49 rue_shop3 so, why... did they make a 32V supply to just bias a 30V transistor, instead of making a 2V supply and stacking it on the 30V supply?
22:50 katsmeow-afk cost
22:50 rue_shop3 they also use a 7805 to generate the 5V from the 32V supply
22:50 rue_shop3 again, instead of using the 30V supply
22:50 katsmeow-afk 2v is a whole new supply, but 30 vs 32 is only a matter of 4 more windings on the transformer
22:51 rue_shop3 in fairness, they did think to use a 4.7V zener to cut down the 32V first
22:51 e_shop3 grumbles 6V and use a tube transfo
22:52 rue_shop3 tho, I doubt there aren't many tube transformers with voltages as low as 30
22:52 katsmeow-afk it's still anotehr transformer, diodes,cap, and regulator
22:53 rue_shop3 they show two inputs, 24VAC 4A and "4 VAC, 250mA" that when rectified becomes 43VDC
22:53 katsmeow-afk 24vac transformers are comon tho, and made 32v firh full wave bridge diodes
22:53 rue_shop3 I have an smps that goes to 36V
22:54 rue_shop3 I ahve a P channel fet that needs 0 to drive it
22:54 rue_shop3 (and does have a good low rdson)
22:54 rue_shop3 so, the plan is to rewrite the output stage from the diode junction forward
22:56 rue_shop3 IRF4905
22:58 tsmeow-afk finds an interesting Suzi Quatro
23:00 rue_shop3 5A * .01R
23:00 rue_shop3 .05V
23:00 rue_shop3 wait, blk brn blk silver is .01 right?
23:00 rue_shop3 er, .1
23:01 rue_shop3 gold would be 1?
23:01 e_shop3 shakes his
23:01 katsmeow-afk sorry, i ate sleepy pills earlier
23:01 rue_shop3 apparenlty my brain stopped
23:01 rue_shop3 1%
23:01 rue_shop3 blk brn blk silver
23:02 katsmeow-afk and i forgot to eat dinner
23:02 katsmeow-afk it's cold now
23:02 rue_shop3 better reburn it
23:03 rue_shop3 oh dog, its almost bedtime
23:03 rue_shop3 I need to burn my supper too
23:06 rue_shop3 my design that failed used a drive current of 2.1mA
23:06 rue_shop3 er 10mA
23:06 rue_shop3 their design uses 2.1mA
23:07 katsmeow-afk rule of thumb: parts change, design for a range of shirt happening
23:11 katsmeow-afk usually, when i design for dc, allis ok, then i look at it from ac pov, and change parts values, which generally makes it overkill for the dc,, and then parts don't fail and all is good
23:12 rue_shop3 I have DC comming in...
23:13 katsmeow-afk any change due to settings, gating, load changes, parameters changing like overcurrent tripping, all that is ac
23:14 katsmeow-afk just the temperature changes is often enough to blow plan dc preliminary values outs the world
23:20 rue_house if I show you what I have, will you check my numbers?
23:20 rue_house its just th output stage really
23:20 katsmeow-afk prolly not till tomorrow
23:20 rue_house thats ok
23:20 katsmeow-afk ok
23:21 katsmeow-afk i'll look, but atm i am so not-here, i am not gonna finish dinner
23:21 rue_house Ill post it
23:25 rue_house http://ruemohr.org/~ircjunk/projects/newpower/p1070122.jpg
23:25 rue_house pls see top right
23:25 rue_house oh I didn't check the wattage on the 2222
23:25 rue_house sigh
23:29 katsmeow-afk won't your IRF4905 drop gobs of voltage if you cannot get it's .... it's a p-channel ?
23:30 rue_house its P channel!
23:30 katsmeow-afk ok
23:30 rue_house 0 is its motivation
23:30 rue_house its got an rdson of like 50mR
23:30 katsmeow-afk your I and V diodes will change as much as 0.1v with temperature
23:30 rue_house they go to the op-amps
23:30 rue_house they are D4 and D5
23:31 rue_house do they generate noise?
23:31 katsmeow-afk ah
23:31 katsmeow-afk no
23:33 katsmeow-afk will the 1k resistor turn off the mosfet fast enough if you short the output?
23:33 rue_house maybe I should put a properly sized inductor in?
23:34 rue_house (series with fet)
23:34 rue_house lets see
23:34 rue_house say, 12V, 1.1k
23:34 katsmeow-afk or a totem-pole driver
23:34 rue_house hmm
23:34 rue_house how many pf for that sucker
23:35 rue_house oh .02R
23:35 katsmeow-afk ??
23:35 rue_house 200W
23:35 katsmeow-afk lol
23:36 rue_house 3400pF input capacitance
23:36 rue_house off by, 2V? maybe?
23:36 tsmeow-afk
23:37 katsmeow-afk also, with no much pf-age, there'll be some coupling of that into the load
23:37 katsmeow-afk arg
23:39 rue_house I think its less than 1us
23:39 rue_house I'm haivng a problem finding anything with the resolution to calc it for me
23:39 katsmeow-afk yo can dump a lot of ampsfrom the main cap into a short in 1us
23:40 rue_house lead wire, inductance
23:40 rue_house the internal wire is only 18guage
23:40 katsmeow-afk 1us = 1000ns , i shut down my psu under 50ns, saved dumpping 47,000ufd at 48vdc into shorts
23:40 rue_house oo I should bump up the grounds to remove the error
23:41 katsmeow-afk IRF4905 is only to-220, good enough for the power?
23:42 katsmeow-afk 75 is a lot for a to-220
23:42 rue_house hmm less than 3.7us
23:42 katsmeow-afk 15v @ 5a
23:42 rue_house its rated 200W
23:42 rue_house ...
23:42 rue_house I can give it a really mean heatsink
23:43 katsmeow-afk no 20-220 is good for 200w for over 2 sec
23:43 katsmeow-afk use two, spread them at least an inch aprt
23:43 rue_house ok
23:43 rue_house 6800pF
23:43 tsmeow-afk calls up a data s
23:44 katsmeow-afk recalc your Rx for two turning off "fast enough"
23:44 rue_house I cant keep the smoke in by just adding a large clamp to the plastic case?
23:44 rue_house I'm trying to calc it at all
23:44 rue_house oh, where is my black book
23:45 katsmeow-afk The TO-220 package is universally preferred for all
23:45 katsmeow-afk commercial-industrial applications at power dissipation
23:45 katsmeow-afk levels to approximately 50 watts
23:46 rue_house ah
23:46 rue_house hmm
23:46 rue_house 30V 5A
23:46 rue_house peak should be at 5A 16V
23:46 katsmeow-afk heck, i'd use 3 or 4 in parallel if IRF is saying they are good for only 50w and you caneasily make them dissipate 70 watts at half voltage output
23:46 rue_house hmm 80W
23:46 e_house thinks about
23:47 rue_house what if my output is at 4V 5A
23:47 katsmeow-afk smoking them is contraindicated
23:47 katsmeow-afk then it's worse
23:47 katsmeow-afk much worse
23:47 katsmeow-afk told ya i cannot do this now, i am just averaging case, not worse-case
23:47 rue_house maybe I should just assume 150w and use 4
23:48 katsmeow-afk you see how things like heatsinks canget totaly blow outa the ballpark?
23:48 rue_house :)
23:48 rue_house I have a few cpu heatsinks that did more than that in their sleep
23:48 katsmeow-afk 4v out at 5amps is dissipating 140 watts
23:49 katsmeow-afk if the to-220 is good for only 50, the 4 is recommended
23:49 rue_house gives me 50W margin
23:49 katsmeow-afk if you use fan-on cpu sinks good
23:49 rue_house yep
23:49 rue_house I'm gonan switch it to switching converters later
23:49 rue_house I'm trying to get this thing working
23:50 rue_house the output stage will just be a module
23:50 katsmeow-afk case is drain, if you electrically isolate the sinks, you don't need to use insulators or such on the transistors
23:52 katsmeow-afk i forget where/when atm, but i got a case of ~2in square sinks with fans on ebay for cheap last yrs
23:53 katsmeow-afk if the load has a capacitor in it, and you lose your Vcc , will the load's cap back drive the fet and pop it?
23:57 rue_house yup
23:57 katsmeow-afk 200w is more than most to-3, no way
23:57 rue_house I have a large bin of cpu heatsinks
23:57 katsmeow-afk erg,, nm
23:57 katsmeow-afk i scrolled
23:58 katsmeow-afk 50 cents for those irf4905
23:58 rue_house :/ I paid more than that
23:59 katsmeow-afk 25 for $12, or 50 for $19
23:59 rue_house found it
23:59 katsmeow-afk 50@$19 = 38 cents each