#robotics Logs

Apr 11 2019

02:18 AM mrdata__ is now known as mrdata_
03:30 AM veverak: wow
03:30 AM veverak: zhanx: nice usb hub
06:49 AM tsglove2 is now known as tsglove
08:27 AM orlock: Police confirmed the WikiLeaks founder was arrested for breach of bail, and also 'on behalf of the United States authorities [under] an extradition warrant'.
09:04 AM Wayfarer: morning
06:53 PM Tom_itx is now known as Tom_L
06:55 PM rue_mohr: the anode has the ZnS on it that glows when hit
06:55 PM MLEVeee: and the filament needs to be warm to puff off electrons
06:55 PM rue_mohr: yup
06:55 PM rue_mohr: the electrons are biased at 5V, so a -5V grid will repel them nicely
06:55 PM rue_mohr: aka 0V
06:55 PM rue_mohr: -5V -V 27V
06:56 PM rue_mohr: but its really
06:56 PM MLEVeee: oh ok
06:56 PM rue_mohr: 0V 5V 32V
06:56 PM rue_mohr: grids perspective
06:56 PM MLEVeee: so when the grids get grounded via the uln2003, they appear to be negatively charged to the electrons puffing off the filament
06:57 PM MLEVeee: so they don't want to pass
06:57 PM rue_mohr: yes
06:57 PM MLEVeee: so if you didn't bias the filament, what would happen?
06:57 PM rue_mohr: kinda random, depends on what floating voltage the grid picks up
06:58 PM MLEVeee: ah
06:58 PM MLEVeee: where would it pick up floating voltage?
06:58 PM rue_mohr: most of it would be 60Hz noise from the transformer
06:58 PM rue_mohr: due to the capacitance between the windings
06:58 PM rue_mohr: (pri:sec)
06:59 PM MLEVeee: ok
06:59 PM MLEVeee: so without biasing, the grid's voltage can just drift
06:59 PM rue_mohr: yes
06:59 PM MLEVeee: and cause weird behaviors sometimes
06:59 PM rue_mohr: and I dont think that always work
06:59 PM rue_mohr: yea
07:00 PM rue_mohr: you would probably get pulsing brightness
07:00 PM rue_mohr: or drifting
07:00 PM MLEVeee: i see
07:00 PM rue_mohr: have you ever seen a circuit for a DC radio tube?
07:00 PM rue_mohr: some of those tubes use the filament as the cathode too
07:00 PM MLEVeee: Maybe
07:01 PM MLEVeee: so when we did this before, we had the filament hooekd to DC and it worked
07:01 PM MLEVeee: because it was biased by being connected to DC
07:01 PM MLEVeee: but DC can cause uneven brightness because the filament does not heat evenly?
07:01 PM rue_mohr: radio tubes usually have )]:| (fil, cathode, grid, anode)
07:02 PM rue_mohr: well the dc has the votlage drop across it
07:02 PM rue_mohr: so if you have 5V on one end, and a 3V drop, your only 2V on the other end
07:02 PM MLEVeee: oh i see
07:02 PM MLEVeee: so you get uneven biasing
07:03 PM rue_mohr: some DC tubes are ):| (filament, grid, anode) because the filament doubles as a cathode
07:03 PM rue_mohr: (yes)
07:04 PM MLEVeee: ok, so you need AC to do the heating and you need 5 volts in the middle to get the biasing where you want
07:04 PM rue_mohr: http://1.bp.blogspot.com/-Wa62V6Y1yWk/T1zISi9UMVI/AAAAAAAABDk/aiTZ2_0Srpg/s1600/fil-vs-cathodebias.jpg
07:04 PM rue_mohr: yes
07:05 PM rue_mohr: I'm just reciting what I know, I could be wrong, but so far I can get them working
07:05 PM MLEVeee: cool
07:05 PM rue_mohr: that image, less the other stuff, shows a tube with a cathode-filament, thats how vfd's work
07:05 PM MLEVeee: well, roger has the PIC and the demux chips and the uln2003s all working correctly
07:06 PM MLEVeee: so this weekend we're going to try hooking up the VFD and the higher voltage stuff
07:06 PM MLEVeee: fingers crossed it will work!
07:06 PM rue_mohr: :)
07:06 PM rue_mohr: as for pullups
07:06 PM rue_mohr: I think my diagram says 2.7k, that chews up lots of pwoer
07:07 PM rue_mohr: when I was digging for a display the same as yours, I saw values of 10k and 47k as quite common
07:07 PM MLEVeee: we were planning to use 10K
07:08 PM rue_mohr: ok
07:08 PM MLEVeee: that would only use 0.72 watts to run the grids, and probably about the same to run the plates
07:08 PM MLEVeee: a watt and a half isn't too bad
07:09 PM rue_mohr: and there is nothing saying that 47k wont work
07:09 PM rue_mohr: that I havn't worked much with
07:09 PM rue_mohr: but my resistors sure were gettin hot!
07:09 PM MLEVeee: i suppose it would work, too
07:10 PM rue_mohr: in the datasheet for the screen, there is probably a recommended current... oh well
07:10 PM MLEVeee: because there's really no current flowing except when the uln grounds the grid
07:10 PM MLEVeee: so there would be basically not voltage drop even with 47K while the grid is on
07:10 PM MLEVeee: right?
07:10 PM rue_mohr: the anode pulls current when the segment is on
07:11 PM rue_mohr: maybe measure your 10K voltage drop if you can, I dont even know the order of magnitude for that current
07:11 PM MLEVeee: you were using 2.7K and they were getting hot?
07:11 PM rue_mohr: @24V
07:11 PM MLEVeee: ok
07:11 PM rue_mohr: zippo:/tmp# ohm -r 2700 -v 24
07:11 PM rue_mohr: Wattage is: 0.213333
07:11 PM rue_mohr: Current is: 0.008889
07:11 PM rue_mohr: Voltage is: 24.000000
07:12 PM rue_mohr: Resistance is : 2700.000000
07:12 PM MLEVeee: so i would guess that using a high value resistor for the segments would reduce the brightness
07:12 PM MLEVeee: right?
07:12 PM rue_mohr: a bit
07:12 PM MLEVeee: lowe power consumption, but some loss of brightness
07:12 PM MLEVeee: ok
07:12 PM MLEVeee: so what to account for when picking a value of resistor?
07:12 PM rue_mohr: oh, for brightness control, rodger can put pwm on that enable of the 74138
07:13 PM MLEVeee: aha. that would work!
07:13 PM rue_mohr: well, the anode will only pull so much current, so there is a payoff between resistor power dissipation and segment brightness
07:14 PM rue_mohr: as I dont know the order of magnitude for the current, I cant really say
07:14 PM rue_mohr: "Heated cathode"
07:14 PM MLEVeee: since neither of us are EEs, i think we will just use 10k and see what happens
07:14 PM rue_mohr: yup I think it'll work fine
07:15 PM MLEVeee: ok, i have to drive home
07:15 PM MLEVeee: thanks, rue!
07:15 PM rue_mohr: and its not like we will ever see a datasheet for that screen
07:15 PM rue_mohr: :)
07:15 PM MLEVeee: byeee
07:15 PM rue_mohr: "directly heated cathode"
07:15 PM rue_mohr: see ya!