#garfield Logs

Nov 23 2020

06:44 AM rue_bed: BS170 might be an analog switch option
06:50 AM rue_bed: ooo 2N7002 is better (lower input capacitance)
07:53 AM rue_bed: "imagine if you had something programmable like an avr, but with an IO expander with as many pins as an FPGA?... oh wait..."
08:38 AM polprog: ive gotta figure out that s&h i posted
08:38 AM polprog: the paragraph below it talks about it briefly but i dont get it fully
08:56 AM rue_mohr: switching diodes, but I cant quite see how it cancels out
08:57 AM rue_mohr: normally there wont be bias for the diodes
08:57 AM rue_mohr: the sampling pulses bias them on
08:58 AM polprog: they say, "a small amount of charge moves the potential in the point between C5 and C6
08:58 AM polprog: "
08:58 AM rue_mohr: c4 and c2 cancel each other to a voltage divider, same with c1,c3
09:00 AM rue_mohr: ah, so d1 and d2 are noramlly reverse biased by R3, r4
09:00 AM rue_mohr: the sampling pulse forward biases them
09:00 AM rue_mohr: but
09:01 AM rue_mohr: 3MR,94nF isn't much time
09:02 AM rue_mohr: lets say the supply is +-10V
09:02 AM rue_mohr: and that d1 and d2 are germanium
09:02 AM rue_mohr: so 0.2V drop
09:03 AM rue_mohr: so, 19.8V across 94.2k
09:04 AM rue_mohr: 210uA
09:07 AM rue_mohr: +10V (r3) 0.121V (r1) 0.1V {interesting} ... I didn't do it right
09:08 AM rue_mohr: oh I see
09:08 AM rue_mohr: nope
09:09 AM rue_mohr: I think I need to print it :)
09:10 AM rue_mohr: the diodes, when on, will have 0.2V ea
09:10 AM rue_mohr: the 100R resistors are dropping about 20mv
09:11 AM rue_mohr: c5 and c6 will change their charge, but
09:14 AM rue_mohr: I think its just a switch
09:14 AM rue_mohr: it lets the input signal thru for a pulse
09:14 AM rue_mohr: then goes back to zero
09:15 AM rue_mohr: I have an exam in 15 mins
09:16 AM rue_mohr: so that circuit would be for PAM
09:22 AM rue_mohr: oh, and you said right there, sampling.
09:30 AM polprog: yep
09:31 AM polprog: hmm so.. yeah the pulses bias the diodes open, but then what
09:31 AM rue_mohr: it kicks the input of the op-amp for just the pulse time
09:32 AM polprog: ah
09:33 AM polprog: the opamp is ac coupled
09:33 AM polprog: makes sense
09:33 AM polprog: and probably the opamp input voltage change is proportional to the input voltage?
09:35 AM polprog: best of luck on the exam
09:35 AM polprog: we have another ee one on thursday
11:19 AM rue_mohr: yea
11:19 AM rue_mohr: 78%
11:19 AM rue_mohr: my lowest score so far
11:19 AM rue_mohr: there was a lot of stuff he didn't cover
11:56 AM Tom_L: happy monday!
11:56 AM Tom_L: is that still passing?
11:59 AM rue_shop2: yes
12:03 PM Tom_L: short day today
12:05 PM Tom_L: how much clearance does your arbor press have?
12:05 PM Tom_L: i noticed the one at HF is too short for anything useful
12:16 PM rue_shop1: 7 inches
12:16 PM rue_shop1: might be 7.5 be advertizing
12:16 PM rue_shop1: (you could squeeze in something 7.5)
12:20 PM rue_shop1: ok, I made scope cables that plug into the breadboard from old audio cables
12:20 PM rue_shop1: bad ones too
12:20 PM rue_shop1: chineese, just a few strands for a shield
03:25 PM aandrew: hahahaha https://i.imgur.com/n0jv775.png
04:00 PM rue_shop1: testing a bunch of 4066
04:13 PM polprog: what have i done :D
04:14 PM polprog: meanwhile i have another ee exam on thu
04:14 PM polprog: this time impedances and 4 port nets
04:39 PM rue_mohr: :)
04:39 PM polprog: hmm, you are checking the Vcc current inrush
04:39 PM polprog: try the iput current now
04:39 PM polprog: :)
04:40 PM rue_mohr: yea, I'm looking at the shoot thru
04:40 PM polprog: oh unless you are checking the noise generating functions
04:40 PM polprog: :P
04:40 PM rue_mohr: I figued 10hz is slow enough I should be able to see anything
04:40 PM rue_mohr: at 10Hz, its not much of a rush
04:41 PM rue_mohr: the input current should be 0 right the way thru
04:41 PM polprog: yeah
04:42 PM rue_mohr: oh god yea twitter is exploding from that
04:42 PM rue_mohr: do you really think I'll get anything from a input current test?
04:43 PM rue_mohr: you see the current is different on them too eh?
04:43 PM rue_mohr: some of them dip 0.5V, some 1
04:43 PM rue_mohr: accept that last one of course
04:44 PM rue_mohr: https://twitter.com/RueNahcMohr/status/1331002626318614529 <-- tom, there is shoot-thru when switching a 4066, I just ran my set thru the scope
05:04 PM polprog: yeah
05:04 PM polprog: but i specifically remember my 4066 shooting thru to ground when i tried to switch it really slow
05:04 PM polprog: eventually i think i used some other design for that mute function
05:04 PM polprog: the schems are probably on the server
05:04 PM polprog: https://polprog.net/rozne1/mixer-rev-D.pdf
05:04 PM polprog: yep
05:04 PM polprog: ive used a 2n7000 eventually
05:04 PM polprog: not the best design but it was acceptable
05:04 PM polprog: i never built that
05:04 PM polprog: not even laid the board out
05:08 PM rue_shop1: ...gn?
05:08 PM rue_shop1: already?
05:08 PM rue_shop1: Who are you and what have you done with the 4am polprog I know?
05:08 PM aandrew: to be honest I'm feeling really run down too
05:09 PM * Tom_L sanitizes his screen
05:09 PM aandrew: and it's only 6pm here
05:15 PM polprog: well
05:15 PM polprog: i was never up til 4am
05:15 PM polprog: im just a bit tired
05:16 PM polprog: and i gotta be up at 10 doing these 4 port nets with a couple people
05:17 PM polprog: i dont feel like i can solve them under pressure (i know the theory, but never had to use it to prove some bullshit like that the RC first order filter is 3db/oct)
05:17 PM polprog: not that they introduced the concept of an octave in the class
05:18 PM polprog: we were doing an RLC bandpass, and the tutor approximated and plotted the response in log scale
05:18 PM rue_shop1: hmm
05:18 PM polprog: but no word about dB or octaves
05:18 PM polprog: weird approach imo
05:18 PM rue_shop1: sounds like it should have been part of foundation
05:18 PM polprog: yeah. its a mess
05:18 PM polprog: the subject is a mess
05:18 PM rue_shop1: that switch doesn't work
05:18 PM rue_shop1: one fet it always on
05:19 PM rue_shop1: or
05:19 PM rue_shop1: its the diodes in this one
05:19 PM polprog: so i gonna do some of the excercises with a couple friends over screencast
05:19 PM polprog: learn myself better this way
05:20 PM rue_shop1: k
05:20 PM rue_shop1: you know how to do filters, right?
05:20 PM polprog: if they arent thinking im an undercover grad yet lol
05:20 PM polprog: yeah
05:20 PM polprog: "how do you do fellow undergrads"
05:20 PM rue_shop1: :)
05:20 PM polprog: i mean
05:21 PM polprog: i was always looking at filters as voltage dividers
05:21 PM polprog: i didnt know complex numbers so i didnt care about the phase shift much. I knew it was thete and i knew what shape the phase graph had so that was enough
05:22 PM polprog: i know all this stuff, or i know where to find it
05:22 PM rue_shop1: I'm not sure wher eI put my jfets
05:22 PM polprog: i just dont like working under time pressure
05:22 PM rue_shop1: yea
05:22 PM rue_shop1: I dont even care if its LOTS of time, I get bothered
05:23 PM polprog: yeah, exactly
05:23 PM polprog: we arent time strict creatures
05:23 PM * polprog looks at the mixer schematic again
05:23 PM polprog: 2 years
05:29 PM rue_mohr: apparently the trick to avoid the pop is to have a super-low-pass on the signal, and to shunt it to that.
05:29 PM rue_mohr: so, the 2.5V point for ya there
05:30 PM Tom_L: polprog, don't take it from him! he gets his days n nights mixed up all the time :)
05:30 PM polprog: haha
05:31 PM polprog: hmmm
05:31 PM polprog: ill keep that in mind when i revisit the mixer project
05:32 PM rue_mohr: I need to find my jfets
06:10 PM rue_shop1: yea, jfet
06:11 PM rue_shop1: bit of a pain to bias, but yea