#avr Logs

May 11 2017

12:17 AM daey_ is now known as daey
01:08 AM l9_ is now known as l9
02:59 AM xentrac: I got a VCO working in simulation: https://is.gd/dumbvco
03:01 AM xentrac: normally people design their VCOs to have predictable oscillation frequencies that don't depend on component value variation or temperature, and that depend on the input voltage in an accurately linear or otherwise predictable way
03:03 AM xentrac: I don't care about any of that, because I can correct those things in software. I just want it to reliably oscillate, and to vary its oscillation frequency with input voltage. I'm not totally convinced that this circuit will always even do that
03:04 AM xentrac: I want to feed the VCO output into a ripple counter with parallel output and use it as an ADC
03:06 AM xentrac: If I drop the capacitance to 5pF I can get it to oscillate (in simulation!) at 260 to 320 MHz, which I feel is pretty decent
03:08 AM Casper: remember, simulators don't like oscillators
03:08 AM xentrac: because they try to find the stable operating point?
03:08 AM xentrac: I'm not convinced it will reliably oscillate because the oscillation seems pretty delicate; it's easy to drive it into non-oscillation by changing component values slightly
03:09 AM xentrac: this circuit.js simulator I'm using doesn't use SPICE and doesn't seek a stable operating point, as far as I can tell; it's more direct-mode
03:10 AM xentrac: like, it starts up your circuit with all the voltages at zero
03:13 AM xentrac: I should probably be simulating using SPICE, both because this thing's transistor models are maybe not very well validated and because I feel like some Monte Carlo simulation would give me a lot more confidence that some small variation in β won't accidentally make the circuit stable
03:15 AM xentrac: what do you mean about simulators, Casper?
03:15 AM Casper: simulators suck, specially the web based ones
03:15 AM xentrac: I don't know very much about any of this stuff, which is why I'm using this simulator
03:16 AM Casper: they have incomplete models and do not simulate anything like noise and component imperfection
03:16 AM xentrac: yeah, this one definitely doesn't do noise
03:17 AM Casper: basically none do...
03:18 AM xentrac: isn't noise analysis a standard analysis in SPICE?
03:19 AM xentrac: Camenzind's book says Simetrix is the bomb here
03:19 AM xentrac: but I've never seen Simetrix
04:04 AM Emil: xentrac: actually
04:04 AM Emil: xentrac: I should file a bug about this too
04:05 AM Emil: but the circuit voltages amd currents dont always reset to zero ":D"
04:38 AM JanC is now known as Guest55043
04:38 AM JanC_ is now known as JanC
04:40 AM markus-k_ is now known as markus-k
09:07 AM yids_ is now known as yids
01:48 PM markus-k: has somebody seen an ATtiny2313 drawing some 6-8mA idle before?
02:16 PM bss36504: markus-k: From the datasheet: Icc @ 8Mhz active, Vcc=5V: 6mA.
02:16 PM bss36504: I assume that's where you're running it.
02:16 PM bss36504: Define "idle". If it's just looping, yeah it's idle but it's still fully active. If it's in a powerdown state, well then there might be a problem.
02:18 PM markus-k: bss36504: it's at 4MHz, just a timer with Fast PWM active
02:19 PM markus-k: running at 5V. I actually had a lower Icc in mind, but still a bit on the high side
02:19 PM Tohin: Hello. Anybody speak Russian?
02:20 PM polprog: markus-k: do you explicitly put it in some power down mode or just loop
02:20 PM bss36504: markus-k: I assume it's more related to the Vcc than the clock, but that's just speculation.
02:21 PM markus-k: it loops on sleep_mode()
02:21 PM polprog: well theres your problem
02:22 PM polprog: read about sleep modes in the datasheet
02:22 PM polprog: i cant tell you anything from the top of my head but there are a couple of them
02:22 PM polprog: going down to several nanoamps
02:23 PM polprog: microamps*
02:23 PM markus-k: it's in idle sleep mode, which should be < 1mA, and only wakes up every couple hundret ms
02:24 PM markus-k: but maybe it's slightly damaged, haven't tried another unit yet and used this one on a breadboard for a while now. It's working fine though
02:25 PM polprog: what does sleep_mode() method look like? can you paste it somewhere?
02:25 PM markus-k: I think it just the sleep instruction
02:25 PM polprog: ok
02:25 PM polprog: i wasnt sure
02:26 PM polprog: odd indeed.
03:10 PM enhering: Lambda_Aurigae?
03:11 PM NoHitWonder^: enhering what is KSP?
03:12 PM NoHitWonder^: oh its kerbal space program
03:12 PM enhering: Is is something extremely addictive...
03:12 PM enhering: yep
03:13 PM enhering: Have you played it?
03:13 PM NoHitWonder^: no but my friend plays it
03:15 PM polprog: i need to get back to ksp
03:15 PM enhering: i need to leave it
03:15 PM polprog: and get up early tomorrow, so good night :) o/
03:15 PM enhering: polprog: China?
03:15 PM enhering: Nope... Romenia
03:16 PM enhering: Or Russia?
03:16 PM NoHitWonder^: poland
03:17 PM enhering: Near...
03:32 PM Emil: https://emil.fi/bashwin
03:33 PM Emil: If only usb worked and I would use it for uploading and coding :/
05:11 PM Lambda_Aurigae: enhering, wha?
05:11 PM Lambda_Aurigae: enhering, now it's time to build a ksp control box...using avr and such.
05:13 PM enhering: hum... interesting idea.
05:13 PM enhering: I have to leave. Be back in three hous
05:13 PM enhering: hours...
05:16 PM Lambda_Aurigae: https://hackaday.io/project/8891-ksp-gegi
05:16 PM Lambda_Aurigae: https://hackaday.com/tag/kerbal-space-program/
05:16 PM Lambda_Aurigae: couple of them.
05:16 PM Lambda_Aurigae: ardweeny based...but
07:09 PM robinak is now known as robink