#garfield Logs

Nov 11 2015

00:40 katsmeow my design shuld drop to 0v output
00:40 rue_shop3 it was a user fault
00:40 katsmeow errrrr,, <cough> but the opamp may not detect it
00:40 rue_shop3 I was testing the current regulation with an led
00:40 rue_shop3 and a resistor that gave me 0-20ma feedback
00:41 rue_shop3 it only dropped the output to 1.3V when I set the current to "0"
00:41 rue_shop3 which, was right.
00:42 katsmeow you might wanna see if you can return that resistrs from opamp(-) to gnd, to something like +5v , so when the output is at 0v the opamp (-) is still in legal operation specs
00:43 rue_shop3 everythings looking good, I need to put in a digital override so the controller can lock the outputs to 0 untill its started up properly
00:44 katsmeow pull Q2-B up to 36v
00:44 katsmeow that will make Q3 float
00:44 katsmeow hmmm
00:44 katsmeow that opamp gnd really needs to goto -5v
00:44 katsmeow +15, -5
00:44 rue_shop3 I have rail-rail op-amps
00:45 katsmeow ok
00:45 rue_shop3 MCP6002
00:47 katsmeow The MCP6001/2/4 family is available in the industrial
00:47 katsmeow and extended temperature ranges. It also has a power
00:47 katsmeow supply range of 1.8V to 5.5V.
00:48 e_shop3 checks which op-amp he's u
00:49 katsmeow so your input control voltage on (+) input must be in that range, and the feedback resistor cannot aply more than that to the (-) input
00:49 rue_shop3 the other one is the LM6482
00:50 katsmeow hmm, no info
00:50 rue_shop3 ?
00:51 katsmeow getting it
00:51 katsmeow Guaranteed 3V, 5V and 15V Performance
00:51 katsmeow yea, use that one
00:51 e_shop3
00:52 katsmeow oooppps
00:52 katsmeow needs an r and some diodes
00:53 rue_shop3 lets see, what variable resistor will divide 5V to 1.25V and have an impedence of 500ohms...
00:53 rue_shop3 er, 1k
00:58 katsmeow LM6482 has no input pretection, needs diodes and curent limiting
00:59 katsmeow update your http://DesignerThinking.com/images/DSCF6660m.jpg
01:00 katsmeow datasheet says some LM6482 inputs cannot be more than .7v apart max
01:01 rue_shop3 hmm, these are things I didn't check
01:01 katsmeow btw, the kill input can be across the BE of Q1 or Q2
01:01 rue_shop3 ok wolfram alpha has solved it
01:01 katsmeow grounding Q1 is prolly easier
01:01 katsmeow grounding Q1-B is prolly easier
01:01 rue_shop3 the values for the votlage divider are 4k and 1.3k
01:02 katsmeow what's it say abot max power dissipation in Q3 ?
01:03 rue_shop3 its really hard to ask it questions
01:03 rue_shop3 http://www.wolframalpha.com/input/?i=F+%3D+5%2F%28R%2BS%29%2C+R*F%3D1.25%2C++1%2F%28%281%2FR%29%2B%281%2FS%29%29+%3D+1000
01:03 katsmeow supplying 2v at 2amps = 68 watts dissipated in Q3 , which is why i suggested more than one transistor there
01:03 rue_shop3 its not for power
01:03 katsmeow 2amps isn't much power
01:04 rue_shop3 I'm using a pwm driver stage, its input goes from 4.76V @ 0% to 7.43V @100%
01:04 rue_shop3 so even if its just brought up to 12V I'm ok, at like, no current
01:04 katsmeow why did i draw it for 36v then?
01:05 rue_shop3 I'm testing with a linear amp to 'full' voltage
01:05 rue_shop3 the idea is to check the error amps work right
01:05 rue_shop3 its a muddle I'm still trying to get all glued togethor
01:05 katsmeow k
01:06 rue_shop3 right now I have 3 different pwm genorator schematics that all 'work'
01:06 rue_shop3 tho I dont remmeber which one will go to 250Khz
01:06 rue_shop3 I thnk I'v worked out the preamp for the dacs
01:07 rue_shop3 If I was good, I could do all this in digital in time-domain
01:08 rue_shop3 phase shift two 50% clock lines to generate a difference thats pwm...
01:08 rue_shop3 by the error of ...
01:08 rue_shop3 analog to...
01:09 rue_shop3 8-| I thought about it... damnit
01:09 katsmeow there's no lm6482 on ebay
01:09 rue_shop3 ? I wonder if I got them from aliexpress
01:09 rue_shop3 they must have been pricey, I only have like 10 of them
01:09 rue_shop3 normally I would buy more than a lifetime supply
01:10 tsmeow
01:10 rue_shop3 like AOP605
01:10 rue_shop3 lots and lots of those
01:10 rue_shop3 and LM386
01:10 rue_shop3 and 74ALS00
01:20 katsmeow MC34071/2/4 opamps include gnd in their input specs , 13v/us slew, 44v power buss, 28v output swing with 30v psu
01:21 katsmeow i got a reel of them real cheap a few yrs ago
01:22 rue_shop3 hmm, I'v found most of the good op-amps you know hard to find
01:22 katsmeow heh
01:23 katsmeow they are on ebay for $1 - $2 each
01:23 katsmeow it was like no one was bidding on the reel, so i bid 99 cents
01:23 katsmeow grabbed some ...72 and ... 74
01:24 katsmeow 30pcs MC34074D Quad Operational Amplifier SO14 MOTOROLA
01:24 katsmeow $6.00
01:24 katsmeow Buy It Now
01:24 katsmeow +$6.00 shipping
01:27 katsmeow need some 700Mhz shift registers?
01:38 -kornbluth.freenode.net:#garfield- [freenode-info] channel trolls and no channel staff around to help? please check with freenode support: http://freenode.net/faq.shtml#gettinghelp
01:39 rue_shop3 bah
01:39 rue_shop3 nowonder this paper was cheap, every page has 3 holes along the side
01:40 katsmeow i hate it when the pencil gets trapped in them and the lead breaks
01:46 rue_shop3 which button do I use on the pencil to select and area and move it down the page...
01:47 katsmeow the one that extends the exacto blade
01:47 rue_shop3 ARG, FEATURE NOT INSTALLED!
01:47 katsmeow sissors work if you have an older pencil without the blades
01:48 rue_shop3 arg, its leaving a transparent area insead of filling it with blank page...
01:48 rue_shop3 undo! undo!
01:48 katsmeow you haveto supply the appropriate background
01:49 rue_shop3 can I use layers?
01:49 katsmeow yeas
01:50 katsmeow we useto be able to get "pins" that anchored all the layers together on a "light table"
01:50 rue_shop3 oo I made a light table
01:51 katsmeow instead of these new dinly 24inch diagonal monitors, we used 2ft by 3ft layered "tiles"
01:51 rue_shop3 was it pcb design?
01:51 katsmeow yeas
01:51 rue_shop3 tape and donuts?
01:52 katsmeow yeas
01:52 rue_shop3 cool
01:52 rue_shop3 was it 1:1?
01:52 katsmeow i wish i could still get them
01:52 katsmeow 2:1 or 4:1
01:52 rue_shop3 hmm
01:52 rue_shop3 cant you use paint?
01:52 katsmeow hard to lift paint
01:52 rue_shop3 ... paint.exe
01:52 katsmeow it sticks once, and then it dries
01:53 katsmeow my paint.exe is inside my puter, and won't erase the lines i drew on the monitor
01:54 rue_shop3 windex.exe
01:54 rue_shop3 :)
01:54 katsmeow i have some of that
01:54 rue_shop3 at 250Khz the logic gates get little wiggles in the lines from crosstalk
01:55 katsmeow are you sure it's crosstalk and not ground bounce?
01:55 rue_shop3 I suppose if I use a schmitt nand gate I dont have to use 2 gates of a 7414 and 2 gates of a 00
01:55 rue_shop3 I'll check the ground
01:56 katsmeow plaid works too
01:58 katsmeow ttl can make for some really dirty grounds
01:59 katsmeow once has to run differential ecl across a board because of the noise induced in a 12" trace by the *ground*
02:01 rue_shop3 *blink*
02:01 rue_shop3 my... scope... has crosstalk issues...
02:01 rue_shop3 5V on one channel causes .4V noise on the other
02:01 rue_shop3 wtf?
02:01 rue_shop3 k
02:01 katsmeow even with the other channel input gnded?
02:01 rue_shop3 yea
02:02 rue_shop3 if I unplug it, goes away
02:02 katsmeow unplug which?
02:02 rue_shop3 maybe...
02:02 rue_shop3 oh, its lead capacitance
02:02 rue_shop3 if I properly ground the probe clips it goes away
02:03 katsmeow if you are grounding the probes in 2 different places, there's your problem, there's prolly .4v ofnoise between the two gnd points
02:03 rue_shop3 I was just using chassis ground
02:04 katsmeow some high performance probes have 7 wires int he leads, to auto-calibrate the tip
02:05 rue_shop3 cool, I always thoguht the little extra pin on the tek for knowing if its x1 or x10 was cool
02:05 katsmeow or x100 or x1000
02:05 katsmeow i have some x100 probes
02:05 rue_shop3 *blink*
02:05 rue_shop3 hv probes?
02:05 katsmeow seen x1000 hv probes
02:06 katsmeow mine are 2kv
02:06 rue_shop3 huh
02:06 e_shop3 thinks about dv/dt @
02:07 katsmeow transmitter "signals" , they are only 350Mhz probes tho
02:07 katsmeow they derate on freq
02:08 rue_shop3 ok so, option 1) 555 (+2R and 2C) ramp/clk gen + lm393 for comparitor
02:08 katsmeow 25v@350Mhz, 750v@10Mhz, 2kv from 2Mhz down
02:08 rue_shop3 option 2) lm393 + 5R and 1c, 1 led, and 1 transistor to do it all
02:09 rue_shop3 I think the 393 only option is better...
02:09 rue_shop3 and I can adjust the ramp span
02:09 rue_shop3 did tom just implode?
02:10 katsmeow i guess
02:11 rue_shop3 ok I'm down to 4 pages of schematics
02:11 katsmeow 200PCS (100Each) MPSA92 & MPSA42 PNP & NPN High Voltage Transistor 500mA 300V
02:11 katsmeow ( 331210484993 )
02:11 rue_shop3 the dac and preamp, the error amps, the pwm gen, and the fet drivers
02:11 katsmeow $5.69 + $2.49 s&h
02:11 rue_shop3 oo
02:12 rue_shop3 whats pn2222 500mA or 700?
02:12 katsmeow 700
02:12 rue_shop3 hmmm
02:12 katsmeow err,, mybe 2222A is 700
02:13 rue_shop3 http://www.thingiverse.com/thing:1095858
02:13 rue_shop3 here kitty kitty kitty...
02:15 katsmeow is it flexable?
02:15 rue_shop3 no idea
02:15 rue_shop3 but I took it good for a laugh
02:18 rue_shop3 there must be a way to make a votlage controlled pulse delay for a 250Khz square wave
02:19 katsmeow synch a pll to it and phase shift that vco with your control signal
02:19 rue_shop3 oooh, brilliant
02:20 rue_shop3 either z's match heads are working or not eating a sub sandwich for lunch yesterday is working
02:20 katsmeow good
02:21 rue_shop3 intersting, my fet driver only puts a 5V drive to the fet base
02:22 rue_shop3 it should be 12
02:23 rue_shop3 your up late
02:23 katsmeow i know :-/
02:23 rue_shop3 I need to design an inductor for 50% duty at 250Khz, say, 40V, 6A
02:24 rue_shop3 I dont suspect it will come out that big
02:26 katsmeow a pulse being 2us ? do you really need that high a freq? the fet capacitance will be huge current drain
02:26 rue_shop3 ok, I think I'll put the DAC and error amps on one board, and then the pwm gen and fet drivers with buck supply on another
02:27 rue_shop3 250Khz is the latest crazy
02:27 rue_shop3 er,. craze
02:27 katsmeow ok
02:28 rue_shop3 IRF4905
02:29 rue_shop3 3.4nF
02:29 katsmeow pchan
02:29 rue_shop3 yup
02:29 rue_shop3 I dont want to drive the power stage against the output
02:29 rue_shop3 causes undesired output signals
02:30 rue_shop3 how does that power calc work for caps?
02:31 rue_shop3 P = C*f?
02:31 katsmeow something like that
02:32 rue_shop3 I would say that Z = 1/2PifC but its a square wave
02:33 rue_shop3 245nJ/cycle
02:33 katsmeow use the rise time of your square wave, x4 is the hf wavelength , 1/x would be the freq
02:34 rue_shop3 0.06125 watts
02:34 rue_shop3 if the online calcs did that right
02:35 katsmeow so you'll be driving the gate with 15,312.5 watts
02:35 rue_shop3 hu?
02:35 rue_shop3 1 cycle is 245nJ, at 250k cycles/sec
02:36 rue_shop3 245nJ/s * 250000 ?
02:36 rue_shop3 is .06w
02:36 rue_shop3 ?
02:36 katsmeow you didn't refresh yur http://designerthinking.com/images/DSCF6660m.jpg
02:36 katsmeow i can tell, because that's 404
02:37 katsmeow it's http://designerthinking.com/DSCF6660m.jpg
02:37 rue_shop3 I saw it was 404,
02:37 rue_shop3 ah perfect
02:37 rue_shop3 what happens if the difference is mor than .7V on the inputs?
02:38 rue_shop3 it blows?
02:38 katsmeow undefined
02:38 rue_shop3 oh, the old 'goes the wrong way'
02:40 katsmeow many opamps don't care that much, you can drive the inpus several volts apart
02:40 katsmeow on the olde Norton opamp, yo can gnd either one and put the other on vcc, it doesn't care
02:40 rue_shop3 I was thinking today about a bunch of amps that use local feedback instead of a bunch in one big loo
02:40 rue_shop3 p
02:41 tsmeow ponders a bunch of 741's in one big
02:41 katsmeow p
02:42 katsmeow Two X100 250MHZ Oscilloscope probes up to 1.2KV A3125
02:42 katsmeow $59.28
02:42 katsmeow Was: $76.99
02:42 katsmeow Buy It Now
02:42 katsmeow Free shipping
02:42 rue_shop3 I was playing with the op-amp gains, 200 to 20 didn't make a diff, but I presume stability would be all over the map
02:43 rue_shop3 I found some 100Mhz 1/10 for about $5 ea
02:43 rue_shop3 hard to find cheap probes
02:44 katsmeow those are 1200v 250Mhz
02:44 katsmeow there's a;lot of $6 100Mhz on ebay
02:45 katsmeow well, maybe they are $8 now
02:45 rue_shop3 I wonder if the feedback dividers should be on a 3rd board
02:45 rue_shop3 its not realy part of the error amp and its not part of the power driver
02:45 katsmeow New 4 100MHZ x10 x1 Oscilloscope clip probes Agilent HP RIGOL Tektronix OWON Acc
02:45 katsmeow $21.48
02:45 katsmeow Was: $22.61
02:45 katsmeow Buy It Now
02:45 katsmeow Free shipping
02:47 rue_shop3 http://www.aliexpress.com/item/1Set-P6100-Oscilloscope-Probe-DC-100MHz-Scope-Clip-Probe-100MHz/32292499407.html
02:48 rue_shop3 I'm not gonna say they are good ones, not bad, and cheap
02:51 katsmeow that's what they are selling on ebay, same thing
02:51 rue_shop3 :)
02:51 rue_shop3 I think lots of ebay things come from aliexpress
02:52 katsmeow someday you need to coach me on how to buy from ali
02:52 rue_shop3 make account, search, search, search, buy, pay with credit card, wait, wait, wait
02:52 rue_shop3 you just cant filter like you can on ebay
02:53 katsmeow ok
02:53 rue_shop3 if you try to filter by price on aliexpress, everything dissapears
02:53 rue_shop3 ali excells for large volumes of things, for 1 or 2 usually ebay wins
02:54 rue_shop3 so, 500 transistors, ali
02:54 rue_shop3 for 4 probes, ebay (usually)
02:55 rue_shop3 ok, so
02:55 katsmeow ok
02:55 rue_shop3 I have 2 current sensors and 2 voltage sensors
02:55 rue_shop3 there is a electronic current sense for the feedback, and an ammeter for the user
02:56 rue_shop3 there is a voltage feedback for the regulator, and a voltmeter for the user
02:56 rue_shop3 I'm gonna say that if the users meters are off a little its not biggie
02:56 rue_shop3 cause they aren't even mirrored
02:57 rue_shop3 so I want the electronic current sensor AFTER the voltmeter
02:57 rue_shop3 and the ammeter before the voltage sense
02:58 katsmeow if you put it at the collector of the Q3 you'll be ok
02:58 rue_shop3 that should keep the error out of things
02:58 rue_shop3 oh
02:58 rue_shop3 wait, thats not my power stage tho
02:59 katsmeow the darlington has a gain typically over 2000, so the base drive is negligable, specially if you remove the Q3-BE resistor
02:59 rue_shop3 and my drive pulls against the drain of the power transistor
03:00 katsmeow if delivering 2amps, the base drive is only 1ma , not even going to show on the meter
03:00 rue_shop3 I really want to use a switching power stage, my output is 5A/channel
03:01 rue_shop3 looking back, I could have done more than 5A/channel with my 10A supply if I did power SOA limiting
03:01 rue_shop3 ah well
03:02 rue_shop3 this IS raking togethor a bit
03:02 rue_shop3 oh I can just shutdown the pwm system to do output disable...
03:03 katsmeow usually, if you are bumping into SOA curves, you have heat dissipation issues anyhow and need more pass elemts in parallel
03:04 rue_shop3 with the switchedmode I'm really good, I'm only 50% duty on each fet, buck mode
03:04 rue_shop3 its a 2phase buck converter
03:04 rue_shop3 (gets around gate drive issues)
03:07 rue_shop3 ok, 3 baords to make
03:07 rue_shop3 regualtor, power convert, and feedback
03:10 rue_shop3 5A, I suppose I should be 5000uF on the output
03:11 katsmeow for 60/120hz yeas
03:11 rue_shop3 how about 250Khz?
03:11 katsmeow for dc filtering or 250khz you can get much smaller
03:11 rue_shop3 dosn't your output kill go the wrong way?...
03:12 katsmeow but for 250khx the ESR becomes VERY important
03:12 rue_shop3 ENOHNELET
03:12 rue_shop3 HFORO
03:12 katsmeow huh?
03:12 e_shop3 squints and roates his head ar
03:13 rue_shop3 OROFH TELENHONE
03:13 rue_shop3 ORDER TELEPHONE
03:15 rue_shop3 RGOT KILON LM 6482
03:15 katsmeow you can prolly get by with 10u at 250khz, but it will still cost a fortume due to esr rating
03:15 rue_shop3 eX^2/2
03:16 rue_shop3 DSZAN/1XelP
03:17 e_shop3 stret
03:17 katsmeow don't break DON"T BREAK
03:17 rue_shop3 *crick*
03:17 tsmeow fre
03:18 rue_shop3 so 25x 10uF then?
03:18 katsmeow that would work
03:18 rue_shop3 if I put them all in parallel in a line, they will blow one at a time from the fet end
03:18 katsmeow yea,, that's a concern too
03:19 katsmeow the pcb trace will resemble s inductor
03:19 rue_shop3 I supoose the ESR causes heating the dries the cap from the inside out
03:20 katsmeow [02:55] <katsmeow> counting up my /testequipment/ directory.........
03:20 katsmeow [02:56] <katsmeow> 170 GB, 47,846 files in 13,182 folders
03:20 katsmeow heating yeas
03:20 katsmeow caps boiling over can ruin your day
03:20 rue_shop3 the little chinese baords used 470uF
03:21 katsmeow haveto go big to get lower esr in one cap
03:21 katsmeow prolly lower than 250khz too
03:22 rue_shop3 how about 4x100uF + 8x0.1uF
03:22 katsmeow ok
03:23 rue_shop3 I suppose you can backfeed into a buck converter and it wont care, its not synchronous
03:24 katsmeow err, umm, well, ok
03:24 rue_shop3 ro should I do a 2 stage LC filter and go light on the first caps, like ~4uF in something sturdy
03:25 rue_shop3 some big non electrolytics that can throw heat well
03:25 katsmeow depends on what you need to do
03:26 rue_shop3 filter 250Khz to dc :)
03:26 rue_shop3 how MUCH ripple... how much...
03:26 rue_shop3 .05V?
03:26 rue_shop3 the input ripple is 36V,
03:27 rue_shop3 720:1
03:27 tsmeow is fa
03:27 katsmeow it's 3am
03:29 rue_shop3 -57db
03:30 katsmeow i can tell you formulas will only get you intot he ballpark, you could be off by 20:1 because of oem tolerances and marketting bs and pcb trace and dielectric properties etc etc etc
03:33 rue_shop3 I'm considering using veroboard
03:34 rue_shop3 I'm also thinking I'd be well advised to put the pwm converter in a metal can
03:34 katsmeow you are approaching territory where a ground plane is a really good idea too
03:35 rue_shop3 hmm
03:35 katsmeow you have 100ns rise and fall times, power industors, etc etc
03:35 rue_shop3 build it Rf style on copper clad...
03:35 katsmeow yeas
03:36 katsmeow 100ns x4 ; 1/x = 2.5Mhz
03:36 rue_shop3 hmm the fets will prolly not need much of a heatsink
03:36 katsmeow any ringing will prolly be 10Mhz
03:37 katsmeow so a simple dc power supply suddenly needs a 20Mhz oscope to troubeshoot
03:37 e_shop3 notes to use metal panels on the power su
03:46 rue_shop3 ok 0.1mH + 2uF first stage, then 0.1mH + 400uF second stage is -140db
03:47 katsmeow ok
03:48 rue_shop3 -180 degrees on both stages
03:48 rue_shop3 what are my octives
03:49 katsmeow do re me fa so la ti do!
03:49 rue_shop3 13.5k (250k) 500k
03:49 rue_shop3 er
03:49 katsmeow oh, those octaes
03:49 rue_shop3 125k (250k) 500k
03:49 rue_shop3 62.5k 125k (250k) 500k
03:50 rue_shop3 I wonder how strong the 3rd harmonics are on a square wave
03:50 rue_shop3 probably quite present
03:50 rue_shop3 both filters have poles about the 1Khz mark
03:50 rue_shop3 I think I'm ok
03:51 rue_shop3 I think 0.1mH isn't hard
03:52 rue_shop3 I wonder if I have any litz wire
03:53 katsmeow for 5amps? well, knowing you, you prolly do
03:53 rue_shop3 .. at 250Khz...
03:54 katsmeow use coax braid?
03:54 rue_shop3 ooo
03:54 rue_shop3 shelac it myself?
03:54 katsmeow possibly
03:58 rue_shop3 the error amp is only good to a little over 7khz so far anyhow
03:58 rue_shop3 op-amp limits
03:59 tsmeow
03:59 katsmeow i tend to not buy opamps under 15Mhz
03:59 katsmeow got some 600Mhz ones
03:59 rue_shop3 wuf
03:59 rue_shop3 video grade?
03:59 katsmeow yeas
04:00 rue_shop3 oh yea, I forgot to keep looking at cathode drivers for that last amp
04:00 rue_shop3 or did I? maybe they couldn't go down to 0V
04:01 katsmeow used a 120Mhz opamp as a comparator to make a pwm audio amp in the early 1980's
04:01 rue_shop3 the 1Mhz pwm?
04:01 tsmeow
04:01 rue_shop3 pll to offset the waves, I need to think on this more...
04:02 katsmeow fed 74F14 to make nice sharp edges
04:03 katsmeow easy enough, convert the 1st osc to feed sawteeth to the 2nd one, set the trigger for the 2nd by where on the sawtooth it comparators to
04:07 katsmeow i should go lay down, see if i can sleep
04:08 katsmeow gnites
17:42 rue_shop3 if I drive the fet at 5V, it will push 1W while on
17:42 rue_shop3 if drive at 12V, it will push 0.5w while on
17:43 rue_shop3 but will require a bunch of extra parts
17:43 at^ finally cannot put off having 2 of theses http://www.harborfreight.com/4-ft-x-6-ft-fiberglass-welding-blanket-67833-1.html
17:43 rue_shop3 how many parts are worth 0.5w power savings
17:43 rue_shop3 something catch fire?
17:44 ^kat^ no, and i hope to keep it that way
17:44 rue_shop3 hows the platform, pics?
17:44 ^kat^ need to weld near the plastic barrels
17:44 rue_shop3 whats a P channel current source look like?
17:44 ^kat^ looks pretty much the same, never mind i have burned 5 lbs of rods welding over 50 little bits together
17:45 ^kat^ i never made a fet current source, but pnp bipolars make goot sources
17:45 rue_shop3 isn't mosfet enhancement mode?
17:45 ^kat^ i decline to answer
17:46 rue_shop3 http://www.pimmlabs.com/web/Image4.gif
17:46 rue_shop3 mrrrrrr!?
17:46 ^kat^ but see with a fet you haveto drop at least 5v, up to 10veven, i never gave away that much voltage with a bipolar
17:47 rue_shop3 oh yea
17:47 ^kat^ plus the fet is soft, the bipolar has a much higher gave for a tiny shift in base voltage
17:47 ^kat^ gaive = gain
17:48 rue_shop3 my idea is to use a current srouce and a low side fet to drive the power fet
17:48 ^kat^ sounds complex
17:49 ^kat^ don't forget, with that mosfet+reistor, you still need a way to turn the idiot thing off
17:50 ^kat^ i dunno, i have no plans to make a current sourse with a mosfet unless i use a bipolar to drive it
17:50 ^kat^ an npn + pchan mosfet szkialia pair would work, sorta
17:51 ^kat^ still throw away all that gate voltage
17:51 ^kat^ if you haveto drop 8v in gate drive, at 10 amps, that's 80 watts you haveto generate and not use
17:53 ^kat^ ok, i go to Bham for welding blankies
17:56 rue_shop3 I just need to create a wave 10V
17:56 rue_shop3 for turning the main fet on and off
17:56 rue_shop3 wait a sec
17:57 rue_shop3 if I have a cmos gate at 10V, whats its Vhi and Vlow
17:57 rue_shop3 my signals comming it go from 0-5
17:57 rue_shop3 maybe I can use the logic levels to stretch it a bit
17:58 rue_shop3 4049 is an inverter isn'tit?
17:59 rue_shop3 at 10V, the low level input votlage is 3V
18:00 rue_shop3 at 10V, the high level input is 7v :/
18:04 rue_shop3 74hc14 can only go to 6V
18:18 rue_shop3 I wonder what current I need to make a class A amp respond to the 250khz pwm
18:23 rue_shop3 kat?
20:47 ^kat^ consarn it rue, whee were you when i needed you?!?
20:47 ^kat^ Canadia or somethin?
20:48 ^kat^ i go in to buy a $21 welding blankie, and i leave after spending $236 !?!
20:49 ^kat^ http://www.harborfreight.com/4-ft-x-6-ft-fiberglass-welding-blanket-67833-1.html
20:51 ^kat^ http://www.harborfreight.com/8-inch-x-12-inch-bench-top-wood-lathe-95607.html <<== can be had for $99
20:51 ^kat^ variable speed
20:52 ^kat^ it's actually 14" between the live center in the tail and the faceplate
20:53 ^kat^ looks like it's made for a bed extender
20:53 ^kat^ needs accessories
20:53 ^kat^ but $100 for that vs $100 to have a piece made for expanding those donuts by .010 ,,, i got the lathe
20:58 ^kat^ i figure if i make a secondary bed with 3" channel, infront of and behind the oem bed, i can add metalworking bits to that bed in the form of grinder and dremel and such
20:59 ^kat^ make up some shafts, faceplate adapters, bearing seats, DOM mandrels, small pulleys, etc
21:57 ^kat^ Friday 6 November 2015 01.36 EST
21:57 ^kat^ Toyota is investing $1bn in a research company it is setting up in Silicon Valley to develop artificial intelligence and robotics
22:00 ^kat^ zhanx , http://www.harborfreight.com/8-inch-x-12-inch-bench-top-wood-lathe-95607.html <<== can be had for $99
22:00 ^kat^ variable speed
22:00 ^kat^ it's actually 14" between the live center in the tail and the faceplate
22:00 ^kat^ looks like it's made for a bed extender
22:00 ^kat^ needs accessories
22:00 ^kat^ i figure if i make a secondary bed with 3" channel, infront of and behind the oem bed, i can add metalworking bits to that bed in the form of grinder and dremel and such
22:00 zhanx K checking
22:00 ^kat^ make up some shafts, faceplate adapters, bearing seats, DOM mandrels, small pulleys, etc
22:00 ^kat^ i went in to get a $21 welding blankie, spend 11x that
22:01 at^ gro
22:01 zhanx Harbor freight does that
22:02 ^kat^ makes people wanna spend money on good deals?
22:02 zhanx Yep
22:03 ^kat^ but $100 for that vs $100 to have a piece made for expanding those donuts by .010 ,,, i got the lathe
22:03 ^kat^ it was a matter of where to spend the $100, and what would i have after i spent it
22:04 ^kat^ Channel 4 renews Humans for second series ahead of season finale
22:08 ^kat^ wierd there is no way to anchor this thing down to the table