Back
[00:01:21] <katsmeow> the 14 is sharpening up the slow moving opamp output, and should have enough hysteresis to do that well enough
[00:02:36] <rue_shop3> pin 1 is clean switch, but pin 3 has some ring as it crosses the ~1.5 and 3V marks
[00:03:40] <katsmeow> ok, put a few k across your led
[00:03:41] <rue_shop3> OH
[00:03:54] <rue_shop3> the ring follows the control voltage on pin 6
[00:04:10] <rue_shop3> the led is to bring the low down 1.4V
[00:04:28] <rue_shop3> as the op-amp dosn't pull low enough
[00:05:11] <rue_shop3> I wonder if I should have a resistor between 3 and 5, the pwm comparitor must be pushing noise back?
[00:05:21] <rue_shop3> as it crosses
[00:05:34] <katsmeow> it shouldn't
[00:05:51] <katsmeow> but that led is an open circuit until it tirns on
[00:06:18] <katsmeow> the sudden turnon may cause the opamp to glitch
[00:06:26] <rue_shop3> 5 cycles in .2us
[00:06:40] <rue_shop3> what about the pulldown?
[00:07:01] <katsmeow> talking about the opamp, not the transistor
[00:07:36] <rue_shop3> but the pullup?
[00:07:42] <rue_shop3> :)
[00:07:52] <rue_shop3> maybe I should number my components :)
[00:08:12] <katsmeow> i asumed the opamp is an OC, i cannot read the part #
[00:08:33] <rue_shop3> it is
[00:08:44] <rue_shop3> lm393 is the only thnig fast enough
[00:09:02] <rue_shop3> so the ring is at about 5Mhz
[00:09:06] <rue_shop3> hmm
[00:11:10] <rue_shop3> I want to know what your seeing
[00:11:23] <rue_shop3> oooh, you mean a series resistor
[00:11:35] <rue_shop3> cause that BE is only .7 and the led is only 1.4
[00:11:55] <rue_shop3> but its all via a 1k pullup
[00:12:38] <rue_shop3> hmm, the '14 is gonna have a pullup too
[00:12:56] * rue_shop3 pulls the '14 and watches the traces
[00:13:50] <katsmeow> i mean across the led
[00:14:36] <katsmeow> but with no scope, i dunno where the glitch actually starts
[00:15:04] <rue_shop3> .. what did I do... the osc failed, cause I'v changed the circuit, I removed the transistor and put in a '14 gate
[00:15:23] <rue_shop3> the glitch happens at abotu 0% and 99%
[00:15:30] <rue_shop3> its a micro low pulse
[00:15:41] <rue_shop3> really short, but enough to screw up the switching
[00:16:03] <rue_shop3> why did I change it to drive the osc with a 7414...
[00:16:27] <rue_shop3> arg, arg arg arg
[00:17:23] <katsmeow> i dunno
[00:17:59] <katsmeow> a vanilla 7474 shold be happy with the opamp output, altho i dunno where you'd get a vanilla 74 anything nowadsze
[00:22:24] <rue_shop3> from the drawr :)
[00:22:29] <katsmeow> oh
[00:23:16] <rue_shop3> ok, lets presume I had a problem with the charge/discharge rate of that circuit being too lop sided
[00:23:16] <katsmeow> so me telling you to use your scope to find the glitch was helpful ? can you put a note to yourself on the wall, to remind to you use the scope again?
[00:23:38] <rue_shop3> when I breadboarded that circuit it worked fine
[00:23:42] <rue_shop3> ! :/
[00:24:14] <rue_shop3> I have no idea how many/what and where to install caps to make the circuit think its still in one
[00:25:17] <rue_shop3> the freq resistor is 10k, assuming designed for 50% as I do, the 5k and the 1k pullup
[00:25:18] <rue_shop3> hmm
[00:26:02] <katsmeow> the signal clocking the 7474 doesn't need to be 50-50, the 74 will 50-50 it anyhow
[00:26:13] <rue_shop3> I think this circuit has to go back to the breading board.
[00:26:32] <rue_shop3> yup, but I think I was trying to keep the osc at ~50%
[00:26:43] <rue_shop3> its not a triangle ramp anyhow...
[00:26:52] <rue_shop3> know a nice triangle ramp gen circuit?
[00:27:12] * rue_shop3 mumbles 250khz
[00:27:47] <katsmeow> current source, charging a capacitor, and a transistor or fet to short the cap at the top of the ramp
[00:28:35] <rue_shop3> not saw...
[00:28:45] <katsmeow> comparator senses the top of the ramp, as you decide it, and kills the cap, which kills the comparator, and the currnet source starts the ramp back upwards
[00:29:13] <katsmeow> ok, ivert the CS and the cap <shrug>
[00:29:49] <rue_shop3> bilateral current source...
[00:29:55] <katsmeow> wants a triangle? use 2 CS, one pulls up, then the otehr pulls the cap back down, then the 1st one pulls it up and thent he 2nd one pulls it down
[00:30:58] <katsmeow> or, use any ole osc clocking a counter with a handfull or resistors making a dac, yo can make ramps, sawtoths, or triangles
[00:31:53] <katsmeow> happy midnight
[00:33:05] <rue_shop3> oh its past bedtime
[00:33:28] <katsmeow> your circuit reminds me of a triangle generator
[00:33:42] <katsmeow> but that led makes it look like it will have problems
[00:34:24] <katsmeow> i need to get some sleeps
[00:34:30] <rue_shop3> this is intersting
[00:34:34] <rue_shop3> my HP fn gen
[00:34:59] <rue_shop3> uses two current sources, and has the square wave comp short them out alternatly
[00:35:07] <katsmeow> yeas
[00:35:15] <katsmeow> [00:06] <katsmeow> wants a triangle? use 2 CS, one pulls up, then the otehr pulls the cap back down, then the 1st one pulls it up and thent he 2nd one pulls it down
[00:35:39] <rue_shop3> yup, its how you switch them that I wasn't thinking of
[00:36:01] <rue_shop3> I usually think series switch, but its a current source, you short it to turn it off
[00:37:32] <katsmeow> easy nuff, if you don't care too much about small errors, just hang a pnp off a +50v line, and a npn on a -50v line, couple 100k R between them, cap where the two 100k meet, then pull the pnp base down, then let it go and pull the npn base up, then let it go, repeat
[00:38:25] <katsmeow> the two transistrs alternately pull up/down on the cap thru the 100k R, and preto, you have a triangle signal
[00:38:43] <katsmeow> my eies keep closing
[00:39:30] <katsmeow> the saw and ramp are cheaper and easier, changing tween the two CS can get tricky
[00:40:04] <katsmeow> with one sharp edge tho, all you need is leave the one CS on full time and short the cap
[00:41:11] <katsmeow> your freq is set hard by another osc, the waveform is made after that
[00:41:48] <katsmeow> usually the waveform amplitude is set hrd too, use a common amp or R network to adjust it after
[00:41:53] <rue_shop3> oh they... used the driver output for the schmitt, so did I...
[00:51:18] <katsmeow> Tom has had 3 more quakes since he went to bed
[00:51:35] <katsmeow> a 3.3 , 3.0 , 3.2
[00:53:02] <rue_shop3> :/
[00:53:12] <rue_shop3> its gonna ripple up to here
[00:55:55] <katsmeow> still nothing on google
[00:58:22] <rue_shop3> it looks like I can fit a 1.6V ramp between the rails of a 5V supply
[00:58:56] <katsmeow> if you are happy with a current source there too
[00:59:13] <rue_shop3> should I amplify the ramp, or compare them as a small signal?
[00:59:40] <rue_shop3> I could amplify the ramp by 3 and get something closer to 5V p-p
[00:59:48] <rue_shop3> then compare at those levels
[00:59:59] <rue_shop3> I dont know if there would be less noise resistance
[01:00:39] <rue_shop3> my orig osc design worked at 12V, not 5
[01:00:47] <rue_shop3> the interface from it was a bit of a mess
[01:00:59] <katsmeow> the lower the signal to start with, the more likely it can have noise
[01:01:38] <rue_shop3> but I should think the ramp would be eaiser to deal with noise than the control signal comming in
[01:01:39] <katsmeow> and the more complicated it is, the more likely it is to be affected by temperature
[01:01:49] <rue_shop3> mmm
[01:02:29] <rue_shop3> I'm really frustrated, most of this is already soldered togethor
[01:02:32] <katsmeow> that is why i suggested the transistor and resistor and a higher voltage than +5
[01:02:34] <rue_shop3> I might have to restart my board
[01:02:46] <rue_shop3> I'd like to operate between the rails
[01:02:59] <rue_shop3> I dont want ot have to have all sorts of supplies
[01:03:11] <rue_shop3> +-15, 12, 5, 2.5
[01:03:43] <katsmeow> ok
[01:04:11] <rue_shop3> right now I have 12 comming in
[01:04:15] <rue_shop3> local 5V regulator
[01:09:28] <katsmeow> within reason, using the highest voltage you can will make the simplest and cheapest, most stable, and most easily toggled current source you can make
[01:10:12] <katsmeow> keep the voltage steady
[01:15:02] <katsmeow> yu can use a LM431 (?) as a current source at unregulated 9v, since you have 12v, make the 9v with a Vdropping resistor, and a capacitor to hold it steady
[01:16:40] <katsmeow> gnites
[01:18:20] <rue_shop3> I get a 3V ramp now
[01:18:41] <rue_shop3> hmm
[01:19:00] <rue_shop3> 431 has a high dropout
[23:59:38] <rue_house> ok
[23:59:39] <rue_house> so
[23:59:41] <rue_house> kat
[23:59:57] <rue_house> the glitch is to do with the 7474 or before