#robotics Logs
Nov 13 2017
#robotics Calendar
01:56 AM Jak_o_Shadows: The LED light distorted the plastic on the torch
01:59 AM Jak_o_Shadows: http://pasteall.org/pic/show.php?id=120326
01:59 AM Jak_o_Shadows: Those scratches weren't visible before ejther
01:59 AM rue_shop3: sweet, I got an all-transistor version of the switching constant current controller working
02:12 AM rue_shop3: so, in 6 transistors I made a osc, comparitor, and RS flipflop
02:12 AM rue_shop3: in the crudest form of each
02:26 AM rue_shop3: I wonder if I can get an avr to software simulate a oversampled 1 bit adc
02:27 AM rue_shop3: and use a 1 bit dac to get the value back
02:44 AM rue_shop3: http://ruemohr.org/~ircjunk/tutorials/elex/1bitanalog/p1080671.jpg
02:45 AM rue_shop3: I wonder what I fast way to simulate the rc filter is
02:45 AM rue_shop3: like, in 10 cycles
02:49 AM rue_shop3: lookup tables?
02:50 AM rue_shop3: state machine?
02:50 AM rue_shop3: its either charging or discharging
02:51 AM rue_shop3: one of the things I learned with the 1 bit is that if your regeneration RC values are out, you still get your output, its just scaled a bit
02:52 AM rue_shop3: an atmega8 has enough eprom for 2 8 bit tables
02:53 AM rue_shop3: but
02:53 AM rue_shop3: something dosn't add up
02:54 AM rue_shop3: its almost like you want to just walk linear thru the table values and just use the table for working out what the value of your index is so that you can compare it to the value you want to output
02:54 AM rue_shop3: value = table[index];
02:55 AM rue_shop3: if (value>control) step = 1; else step = -1
02:55 AM rue_shop3: value += step;
02:56 AM rue_shop3: and your 1 bit output is step
02:56 AM rue_shop3: not optimized, but conceptual
02:57 AM rue_shop3: OH GREAT, NOW I HAVE TO TRY IT
03:03 AM rue_bed: so, I need a 256 byte table of RC charge values
03:32 AM rue_bed: ok I have a charge values table
03:32 AM rue_bed: how does this work
03:36 AM Jak_o_Shadows: I was converting a location dependent velocity to a time dependent velocity todaya
03:36 AM Jak_o_Shadows: took me far too long
03:39 AM rue_bed: ;)
03:40 AM rue_bed: is 1:39 am here
03:40 AM rue_bed: I have a able of voltage values for an RC charge, I want to use it to simulate RC changes
03:41 AM rue_bed: charging is easy, you just start at zero and increment thru the table
03:41 AM rue_bed: but, switching to discharge at a random point is hard
03:49 AM rue_bed: its almost like I need a table that tells me how to jump across the first table
03:49 AM rue_bed: but no, the values have to be scaled
03:49 AM rue_bed: arg
03:50 AM rue_bed: no
03:51 AM Jak_o_Shadows: I roughly wordly explained an algorithm for some form of differentiatio
03:51 AM Jak_o_Shadows: well. I know the value of the closed integral, and one of the end points. Find the other end point. You're integrating over discrete data.
03:52 AM rue_bed: RC charge
03:52 AM rue_bed: e^(-t/RC)
03:53 AM rue_bed: er
03:53 AM rue_bed: 1-e^(-t/RC)
03:58 AM Jak_o_Shadows: Yeah. was just talking about the problem I had today.
03:58 AM Jak_o_Shadows: Would be a fun little problem to do with the minimum interpolations
04:01 AM Jak_o_Shadows: you could reuse integratiosn and stuff whilest searching
06:26 AM SpeedEvil: https://www.youtube.com/watch?v=opsmd5yuBF0
06:26 AM SpeedEvil: Autonomous Drifting using Machine Learning
07:06 AM robotustra: morning
12:21 PM rue_bed: drifting?
12:26 PM SpeedEvil: rue_shop3: manoevering a vehicle with substantial orthogonal motion relative to the tyres.
12:35 PM robotustra: manoevering?
12:38 PM SpeedEvil: https://www.youtube.com/watch?v=opsmd5yuBF0
12:38 PM SpeedEvil: Autonomous Drifting using Machine Learning
12:44 PM robotustra: not so interested, thinking about building my mill
12:47 PM rue_bed: tom is inspiring you eh?
12:48 PM anonnumberanon: yo i would so do this for my car
12:48 PM anonnumberanon: make an automatic drifting function
12:50 PM polprog: that would take all the fun away
12:51 PM SpeedEvil: It is an interesting unaddressed aspect of autonomous cars.
12:52 PM SpeedEvil: If in an accident situation the question no longer becomes 'can I stop in time'. But 'can the equivalent of the worlds best stunt driver who has had a week to try and plan this stunt avoid an accident'
12:53 PM rue_bed: heh
12:53 PM weyland|yutani: driftbot is drifting #YOLO
12:55 PM weyland|yutani: and for the record a will be first coining the name Douchebot for a very obnoxious robot
12:56 PM robotustra: rue_bed, no, I was thinking about cnc lathe already for half of year
12:57 PM robotustra: yesterday I put all stuff on the workbench and started to measure https://imgur.com/vXO0IWO
01:19 PM rue_mohr: ah, thats a batter pile
01:19 PM rue_mohr: I wonder what the price validity is of converting an old lathe
01:20 PM Ultrasauce: so what's the runout on that spindle
01:20 PM rue_mohr: whih one?
01:21 PM rue_mohr: robotustras looks like about 8-10 inches right now
01:21 PM Ultrasauce: the clamp handle :D
01:21 PM rue_mohr: :)
01:21 PM Ultrasauce: also hello it's been awhile
01:21 PM rue_mohr: we didn't notice
01:22 PM rue_mohr: so yesterday I designed all sorts of switching current regulators
01:22 PM Ultrasauce: moved on from all-linear designs finally?
01:27 PM rue_mohr: I'm wondering if I can do an isolated dc-dc w/regulated output
01:28 PM rue_mohr: its hard to beat the simplicity of pushing a lm7805 into a circuit that needs 80mA of 5V
01:28 PM rue_mohr: and its hard to find a switching regulator that can beat the noise level
01:28 PM Ultrasauce: yeah of course that's the primary tradeoff
01:29 PM rue_mohr: robotustra, hah, your steel is canadian
01:30 PM rue_mohr: even with lots of filtering, those little dip8 sized switching regulators from china have horrid levels of noise
01:30 PM Ultrasauce: i think the go-to technique there is to use a linear regulator on the output
01:30 PM Ultrasauce: as long as the voltage difference is large enough to be worth it
01:31 PM Ultrasauce: but yeah low noise + high efficiency + low cost is basically unattainable
01:33 PM rue_mohr: https://www.aliexpress.com/item/10pcs-lot-Mini360-DC-DC-Buck-Converter-Step-Down-Module-4-75V-23V-to-1V-17V/32437031041.html
01:33 PM rue_mohr: these ones
01:33 PM rue_mohr: horrid noise, even with lots of filtering
01:33 PM rue_mohr: anyhow, for current regulation, I'v done up a few circuits now
01:33 PM rue_mohr: one of them is just 6 transistors
01:34 PM rue_mohr: I'll post some images later today I think
01:36 PM robotustra: rue_mohr, it's marked as canadian, but I'd found a stamp "USA" on some of them
01:36 PM rue_mohr: interesting
01:36 PM robotustra: 2.5x1.5x0.25 rectangular tubes
01:37 PM rue_mohr: cause, I dotn think we have steel mills here
01:37 PM rue_mohr: were big on aluminum
01:37 PM robotustra: so I think it's just canadian reseller
01:38 PM robotustra: I decided to make it solid, that's why I go mainly with steel. some housing for bearings I'll make from aluminium
01:39 PM robotustra: and I got 4 jaw chuck, so hope on 1 runout ;)
01:39 PM robotustra: and I got 4 jaw chuck, so hope on 0 runout ;)
01:40 PM Tom_L: won't be 0
01:40 PM Tom_L: but should be minimal
01:40 PM robotustra: yes
01:41 PM robotustra: I have 2 taper bearings
01:41 PM robotustra: to hold spindle
01:41 PM robotustra: so I hope on ~0
01:44 PM robotustra: ordered a 1.5 kW motor for it
01:44 PM rue_mohr: what type?
01:45 PM rue_mohr: brushed, brushless, 3 phase, single phase?
01:45 PM rue_mohr: diesel?
01:45 PM rue_mohr: :)
01:46 PM Ultrasauce: waterwheel
01:46 PM robotustra: https://www.ebay.com/itm/1500W-1-5KW-Air-Cooled-CNC-Spindle-Motor-Inverter-Converter-13pcs-ER11-Collet/302490507750
01:46 PM robotustra: will reduce the speed may be 5 times
01:47 PM rue_mohr: brushelss 24k (rpm)
01:48 PM robotustra: I use similar one on my cnc mill, I never go higher 12000 rpm
01:48 PM rue_mohr: wait a sec
01:48 PM rue_mohr: hold on....
01:48 PM rue_mohr: uh, ER11 collette
01:48 PM rue_mohr: 1.5kw
01:48 PM rue_mohr: uh
01:49 PM robotustra: my chuck is 160 mm in dimeter
01:49 PM rue_mohr: ER11 is smaller than this ER16...
01:49 PM robotustra: I don't need this ER11
01:49 PM rue_mohr: ah
01:49 PM rue_mohr: k
01:49 PM robotustra: I'll put a pulley on this spindle
01:49 PM rue_mohr: cause,t hat dosn't add up
01:50 PM robotustra: on the shaft
01:50 PM rue_mohr: its sleeved in
01:50 PM robotustra: what?
01:50 PM rue_mohr: somehow, the ring comes off and that vcollette holder must come out
01:50 PM rue_mohr: it looks like
01:50 PM robotustra: no
01:50 PM robotustra: it's not
01:51 PM rue_mohr: k
01:51 PM robotustra: the hole is in the shaft
01:51 PM rue_mohr: either way, ncie motor
01:51 PM rue_mohr: 300usd
01:51 PM rue_mohr: hmm
01:51 PM robotustra: without invertor it's 150
01:51 PM rue_mohr: geez a fellow could spend a lot of money
01:52 PM robotustra: this lathe will cost me like proxxon lathe, but I will have better
01:52 PM robotustra: actually I want it to lathe, grind and polish
01:53 PM robotustra: and it should be compact to live below the workbench
01:54 PM robotustra: like this https://imgur.com/JtctfBE
01:56 PM robotustra: now I have to think a little bit about tool post
01:59 PM robotustra: not sure I want timing belt or V-belt
02:29 PM Tom_L: robotustra better make sure the spindle pulley is balanced
02:30 PM Tom_L: at those rpm
02:30 PM Tom_L: and you definitely want timing belt
02:30 PM Tom_L: HTD series
02:30 PM Tom_L: the round 'teeth' not the square ones are best
02:32 PM Tom_L: and btw rue_mohr, why are you off today?
02:35 PM robotustra: yeah, timing belt is the last turn
02:44 PM rue_shop3: REMEMBERENCE DAY HOLIDAY
02:49 PM robotustra: amnesia prophylactics day
02:51 PM rue_shop3: OH SORRY CAPS
02:51 PM rue_shop3: off
02:52 PM rue_shop3: yes, thats better
02:56 PM robotustra: btw, who knows what are adequate sizes of cutters for the small table lathe like 7x10"?
02:59 PM robotustra: 3/8?
03:01 PM rue_shop3: 6.34mm, dont get any of that 6.35mm junk
03:04 PM robotustra: 3/8 is approxiamately 10 mm
03:04 PM robotustra: what is the next size?
03:06 PM rue_shop3: not sure
03:08 PM robotustra: hm.. googling
03:09 PM rue_shop3: so, hmm
03:10 PM rue_shop3: I got a lot of switching current control stuff done, my objective was to make a current controlled motor driver
03:10 PM rue_shop3: so
03:10 PM rue_shop3: I have a lot of stuff thats going to be great for current controlling stepper drivers if I ever need to
03:10 PM rue_shop3: +- control
03:10 PM rue_shop3: I ended up aginst a wall of needing a good dac
03:11 PM rue_shop3: for setting the amount of current
03:12 PM rue_shop3: which is what started me on the 1 bit dac thing
03:12 PM rue_shop3: I'm leaning towards it being a bad idea to use a microcontroller for the active element in the current control loop
03:13 PM rue_shop3: I wonder if I can make a 1 bit dac the same way I made the pwm sinewave generator
03:13 PM rue_shop3: technically, its a adc I want, but with a virtual analog
03:14 PM rue_shop3: the problem with the 1 bit is that the output pattern changes depending on what value your comming from
03:14 PM rue_shop3: IT WOULD SURE BE NICE IF DAC'S WERE CHEAP
03:16 PM rue_shop3: I think 60hz bandwidth would be enough
03:17 PM rue_shop3: "8 bit" would be 1Khz oversampled 1 bit ?
03:20 PM robotustra: hm, it looks like the smaller size for lathe bits is 5/16
03:20 PM robotustra: and bigger one is 3/8
03:23 PM rue_shop3: I wonder how long before we get a power outage
03:23 PM rue_shop3: pretty scruffy wind
03:26 PM rue_shop3: maybe what I need is just a division table
03:26 PM rue_shop3: state machine rue
03:26 PM rue_shop3: hu?
03:26 PM rue_shop3: do it with a state machine
03:26 PM rue_shop3: wouldn't that have to be one HELL of a machine?
03:27 PM rue_shop3: well 8 bits input value, 8 bits output value, 8 bit feedback
03:27 PM rue_shop3: so 16 bits in, 8 bits out
03:27 PM rue_shop3: what? wait, something is wrong with that picute
03:28 PM rue_shop3: no compare output bit
03:28 PM rue_shop3: it implies that every... no what
03:28 PM rue_shop3: its too small
03:28 PM rue_shop3: how
03:28 PM rue_shop3: wel.... not sure, but something dosn't add up
03:29 PM rue_shop3: if you know what value your at, and your know what value you want, you know the next value
03:29 PM rue_shop3: because the comparitor will either be adding or subtracting charge
03:30 PM rue_shop3: of which the previous value dosn't matter
03:31 PM rue_shop3: so, your saying that for any given 8 bit state, with a given 8 bit input, we know what the next value will be
03:31 PM rue_shop3: yea, your not charging and discharging at different levels
03:31 PM rue_shop3: ok
03:31 PM rue_shop3: fixed curves
03:31 PM rue_shop3: exactly
03:32 PM rue_shop3: so its a 64k table
03:33 PM rue_shop3: and if your implementing it on a microprocessor, you can use a compare instruction to get the 1 bit output
03:33 PM rue_shop3: if I'm not, I'd need to either bring it down to 7 bits or add another 64kx1 for the value
03:33 PM rue_shop3: the 688 is a comparitor
03:34 PM rue_shop3: 64k*8 you ninny
03:34 PM rue_shop3: 512k rom
03:34 PM rue_shop3: could we mix 8 bit input with 7 bits output?
03:35 PM rue_shop3: technically it would hunt to correct
03:35 PM rue_shop3: right?
03:35 PM rue_shop3: I dont know, I need a diagram
03:35 PM rue_shop3: ok, zoom out for a min tho
03:36 PM rue_shop3: solution shouldn't be made of a cubic foot of chips
03:36 PM rue_shop3: and what are we using this for?
03:36 PM rue_shop3: controlling the current on a dc motor
03:36 PM rue_shop3: k, so, blackbox
03:37 PM SolG is now known as Solgriffin
03:37 PM rue_shop3: ok, we have a motor output.... 1 lead I supopose, reverse dosn't matter
03:37 PM rue_shop3: and a serial input for the current set
03:37 PM rue_shop3: I think thats it
03:38 PM rue_shop3: yea, so its a serial controlled current source
03:38 PM rue_shop3: well, yea
03:38 PM rue_shop3: so, lets say a t13 for recieving the serial
03:38 PM rue_shop3: it just went overbudget
03:38 PM rue_shop3: I noticed that...
03:39 PM rue_shop3: how about a 595 dac on a 393 pwm circuit w/ second gate doing the current limiting
03:39 PM rue_shop3: you overdid it
03:40 PM rue_shop3: right, 595 dac on a 393 current comparitor
03:40 PM rue_shop3: you could also mix with the transistor curuit
03:41 PM rue_shop3: you should go hang some milkjugs off the tent
03:41 PM rue_shop3: s
03:41 PM rue_shop3: lunch would be good to
03:41 PM rue_shop3: didn't you get a set of exponential smt resistors?
03:41 PM rue_shop3: no
03:42 PM rue_shop3: well 11.11 would have prolly been a good day to do it
03:42 PM rue_shop3: awe shutup
03:42 PM rue_shop3: :) the savings seem to be just as good on any other given day anyhow
03:42 PM rue_shop3: you mean the 10x against NA prices?
03:42 PM rue_shop3: that too yea
03:42 PM rue_shop3: lunch
03:43 PM rue_shop3: yes, what do you not want to not eat?
03:43 PM rue_shop3: I want the noodles that will kill me
03:43 PM rue_shop3: we didn't buy any, would you like to know why
03:43 PM rue_shop3: hey, you asked
03:44 PM rue_shop3: well how about we use one of the sucky supliments we got
03:44 PM rue_shop3: pass
03:44 PM rue_shop3: how about perogies
03:44 PM rue_shop3: pass
03:44 PM rue_shop3: you want to eat, right?
03:44 PM rue_shop3: ..... pass
03:44 PM rue_shop3: what do you want for lunch
03:44 PM rue_shop3: I want to be a robot
03:52 PM rue_shop3: I still want to know how to make a 512k serial eeprom into a 1 bit dac
03:53 PM rue_shop3: adc
03:53 PM rue_shop3: technicallity
03:53 PM anonnumberanon: rue_shop3, who are you talking to?
03:53 PM rue_shop3: ... cant you see him?
03:53 PM anonnumberanon: no probably ignored
03:53 PM rue_shop3: :P
03:53 PM anonnumberanon: i don't remember ignoring anyone ever on #robotics
03:54 PM rue_shop3: how much is a 512k eeprom anyhow
03:54 PM rue_shop3: less than $1
03:55 PM rue_shop3: hmm, and a 595?
03:55 PM rue_shop3: about 5c
03:55 PM rue_shop3: wow, that is a price gap
03:55 PM rue_shop3: how about the i2c things
03:55 PM rue_shop3: na, you cant bus up on them
03:56 PM rue_shop3: 8574?
03:56 PM rue_shop3: pca?
03:56 PM rue_shop3: pcf
03:56 PM rue_shop3: chips are 35c
03:57 PM rue_shop3: well, smt, dip is over $1 ea
03:57 PM rue_shop3: 595 just won didn't it?
03:57 PM rue_shop3: oh yea
03:57 PM rue_shop3: so
03:57 PM rue_shop3: yea, buy an exponential set of resistors
04:00 PM rue_shop3: stay within 35mA
04:00 PM rue_shop3: I feel a program comming on
04:00 PM Tom_L: robotustra, i use 5/16
04:01 PM Tom_L: rue_shop3, just don't spill any
04:01 PM edmoore_ is now known as edmoore
04:30 PM rue_shop3: click
04:30 PM rue_shop3: I remember working on this before
04:30 PM rue_shop3: I worked out that finding exponential values is impossable
04:31 PM rue_shop3: so it finding resistor ladder networks for dac
04:31 PM rue_shop3: doing R-2R is reasonable, and I got the resistors for it
04:56 PM SpeedEvil: Or just 1-bit datastream and a filter
04:56 PM SpeedEvil: appropriate filter does sigma-delta reconstruction
05:07 PM climbing_up: Anyone here ever ordered from Seeed Studio? How was it?
05:07 PM rue_shop3: thats funny
05:07 PM rue_shop3: the
05:07 PM rue_shop3: the smt resistors cost WAY more than the 10 turn thruhole ones
05:08 PM rue_shop3: climbing_up, I get everything from china
05:08 PM rue_shop3: SpeedEvil, the problem is, generating the 1 bit stream from a virtual value
05:08 PM rue_shop3: I prolly need about 500Khz
05:08 PM z64555: climbing_up: like any other online electronics retailer I've dealt with. I haven't done any returns so I can't vouch for, well, anybody
05:08 PM rue_shop3: for a descent signal at 60Hz
05:09 PM SpeedEvil: rue_shop3: what are you trying to do
05:09 PM rue_shop3: serial controlled switching current controller
05:10 PM rue_shop3: whats a t13 worth again
05:10 PM rue_shop3: ?
05:10 PM rue_shop3: hmm 27c smt
05:11 PM rue_shop3: considering what the dac is turning into, the t13 might be going back to the cheaper solution
05:12 PM rue_shop3: and put it in the current control loop
05:12 PM rue_shop3: 1/channel
05:12 PM rue_shop3: use its comparitor
05:13 PM rue_shop3: oh, right
05:13 PM rue_shop3: I cant use the comparitor cause I have nothing to compare against
05:13 PM rue_shop3: and the adc is too slow
05:14 PM Tom_L: climbing_up seeeeeeed studio's quality is less than stellar
05:15 PM climbing_up: I'm looking at a lidar I can get from seeedstudio.com and robotshop.com. Seeed seems to have a nicer site and doesn't block my VPN ;)
05:15 PM Tom_L: sloppy mask, poor registry
05:15 PM Tom_L: general chinese junk on the iffy side
05:16 PM rue_shop3: you generally pay 1/10 to get it from china direct
05:16 PM climbing_up: You think I can get a $600 lidar for $60?
05:17 PM rue_shop3: the hardest part is working out wtf the chineese call it
05:17 PM rue_shop3: robot = smart car
05:17 PM rue_shop3: etc
05:19 PM climbing_up: I've been testing the Scanse Sweep. It turns out the range is fine but the sample rate is too low (only 1 kHz). I want to upgrade to a 4 kHz model from Slamtec
05:20 PM rue_shop3: you should just buy a $10000 one
05:21 PM rue_shop3: then you dont have to worry about doing any other projects if it dosn't work or you blow it up
05:21 PM climbing_up: lol
05:22 PM rue_shop3: I need to choose a current range for my digital current sink
05:22 PM rue_shop3: I think 1A
05:34 PM anonnumberanon: but.. 1A is not a range!~
05:40 PM Ultrasauce: https://www.youtube.com/watch?v=giBBQaqbLFc
06:20 PM robotustra: Tom_L, what is the thikness of your lathe bits?
06:22 PM robotustra: Tom_L, is it square?
06:23 PM robotustra: huh, just bolted the vise to the workbench
06:23 PM robotustra: fill much better
06:39 PM Tom_L: it better be square
06:39 PM robotustra: what size of your lathe?
06:39 PM Tom_L: what do you mean what size?
06:39 PM robotustra: 7x10
06:39 PM robotustra: 8x15?
06:39 PM Tom_L: i hope to get 9 x 18 or so
06:40 PM robotustra: but what do you have now?
06:41 PM Tom_L: the sherline i can get about 7 x 9
06:41 PM Tom_L: it's the one with the extended arm
06:41 PM robotustra: mine will be about 8x10
06:42 PM robotustra: it means that I can stick with 3/8
06:42 PM robotustra: I think 5/16 is a bit small
06:42 PM Tom_L: 3/8 what?
06:42 PM Tom_L: cutters?
06:42 PM robotustra: lathe bits size
06:43 PM robotustra: yes
06:43 PM Tom_L: my lathe isn't a sherline
06:43 PM Tom_L: it's an old sears atlas
06:43 PM Tom_L: and i use 5/16 cutters on it
06:44 PM robotustra: I want to make something simolar to this tool changer https://www.youtube.com/watch?v=Yi-YXHsrbfA
06:45 PM robotustra: but for the beginning most probably it will be simple tool holder
06:50 PM robotustra: https://www.youtube.com/watch?v=JfMR7jDzDqk
06:55 PM Tom_L: cute
06:58 PM robotustra: but it has small chuck
06:58 PM robotustra: the guy is doing very nice design
07:00 PM robotustra: I'm designing something like this one but horizontal https://www.youtube.com/watch?v=eipyve8xv24
07:05 PM rue_shop3: its not about size, its about how you use it
07:07 PM robotustra: it's all about capabilities
07:07 PM robotustra: I will have 40mm through hole in spindle
07:07 PM robotustra: chinise ones have only 19mm
07:08 PM robotustra: wtf
07:08 PM robotustra: it's one of the reason I'm bulding my own
07:46 PM rue_bed: what I need is a current-to-freq converter
07:46 PM rue_bed: skip the analog feedback signal all togethor
07:47 PM rue_bed: tho, it introduces a nontrivial delay
07:47 PM rue_bed: hmm, adjustable threshold or external conversion
07:48 PM rue_bed: digital pots are too expensive...
08:13 PM robotustra: there are chips which generate frequency depending on input voltage
08:13 PM robotustra: put your current through resistor and you'll have what you want
08:15 PM robotustra: http://www.falstad.com/circuit/e-vco.html
08:16 PM robotustra: it's calle VCO
08:16 PM robotustra: voltage controlled oscillator
08:31 PM rue_mohr: no
08:31 PM rue_mohr: I need a current controlled osc
08:32 PM rue_mohr: the less parts the better
08:32 PM rue_mohr: oooooh
08:32 PM rue_mohr: ooooooooooh!!!!
08:32 PM Ultrasauce: http://schmitzbits.de/vco3.html
08:33 PM rue_mohr: transformer osc, w/ the current going thru the core
08:33 PM Ultrasauce: ignore the waveshaper shit all you need is the top right part
08:33 PM rue_mohr: nice but waaay too many parts
08:34 PM rue_mohr: context, last night I made a 6 transistor switching constant current supply
08:35 PM Ultrasauce: i guess it doesnt even need to be accurate/linear, just stable with feedback
08:35 PM rue_mohr: yea, there we go, 1 transistor heartly osc
08:36 PM rue_mohr: it should change its tune when there is 1A going thru the coil
08:36 PM rue_mohr: I can count/measure that in the control circuit
08:37 PM rue_mohr: might have to be a saturating core
08:38 PM Ultrasauce: oh that was the wrong link, this is the one i meant http://schmitzbits.de/vco4.html
08:38 PM Ultrasauce: but yeah if a 555 and an opamp are too much then something else
08:39 PM rue_mohr: :) I think I can do it with 1 transistor
08:40 PM rue_mohr: http://www.electronics-tutorials.ws/oscillator/osc8.gif
08:40 PM rue_mohr: see that?
08:40 PM Ultrasauce: achieving stability is going to be challenging
08:40 PM rue_mohr: imagine another turn going thru the inductor that has the detection current going thru it
08:40 PM rue_mohr: it causes the freq to change
08:41 PM rue_mohr: does the tiny13 has pulse cature?
08:41 PM rue_mohr: timer counter!
08:42 PM rue_mohr: whats the max freq?
08:45 PM rue_mohr: I dont think its limited by the system freq
08:45 PM rue_mohr: so I think its 20Mhz
08:47 PM rue_mohr: that of I use a saturable reactor method
08:49 PM rue_mohr: heartley and the other one is...
08:49 PM rue_mohr: colpitts?
08:53 PM rue_mohr: ok goal a) is to find out if I can get ANY kinda LC osc to work
08:53 PM rue_mohr: me and RF dont get along
08:54 PM rue_mohr: goal b) is to apply dc currents to the cores of them and see what the changes are
08:59 PM sadisticroot is now known as evilroot
09:31 PM SpeedEvil is now known as Guest51474
09:42 PM rue_shop3: ok I found my arrl handbook
09:42 PM * rue_shop3 ties to lift it to the table
09:47 PM BitEvil is now known as SpeedEvil
09:55 PM z64555: "handbook"
09:55 PM rue_shop3: its ok if you also have a handtruck
09:56 PM z64555: now that I can understand
11:32 PM rue_shop3: apparently the ability for my solderless breadboard to oscillate at 43Mhz is throwing off everything to do with making an LC oscillator
11:37 PM rue_shop3: the adc on the t13 is 15ksps
11:38 PM rue_shop3: the counter goes up to 20Mhz
11:38 PM rue_shop3: so I can get, 8 bit 78Khz with it
11:39 PM rue_shop3: is that fast enough?