oooh I'm early this time! Your series has been my favourite recently! It's the sort of content I cannot believe I'm watching for free haha

I'll take a crack at solving some of the 1% resistor/trimmer mysteries. Assuming the unlabeled connection goes to V+ and ignoring R81, R83 and R82 form a voltage divider to bias that node to 1/3V+. 1% resistors are used throughout the DC-coupled path starting at IC13 in the Q3/Q4 subcircuits, plus similar circuits elsewhere. I suspect the 1% values were chosen experimentally to compensate for actual offset from a typical 741 opamp. With R84 they're probably compensating for offset between the switch settings since R89 and R90 are nearly the same ratio as R83/R82 and R84/R85, but that doesn't explain the wide range provided by R84/R85, so my best guess is R84 has to be adjusted as final compensation for accumulated asymmetry in the circuit brought about by adjusting the other trimmers. The datasheet for NF510 says VGS ranges from 0.5 to 10V, which is maybe the last piece of the puzzle since it surely requires some compensation to get consistency between units. Since V+ is 15V, if we continue to ignore R81 then R83/R82 bring the voltage to the middle of the VGS range of NF510 at 5V, with IC17 producing -10V when Q4 is conducting, which coincides with the max logic level of NF510. R84 provides about 5V of adjustment range to level shift the output from the upper half. That's about as far as I'm gonna get here I think, but hopefully that's on the right track!

@Ruba1323

Жыл бұрын

I did make circuit based on this. R83 does go to positive rail and is there just to align both parts of the newly formed saw wave. And my guess for R82 is that they used it to compensate the value of R83, so they don't have to use trimpot there. It is quiet sensitive to change of resistance hence the 1%.

Gosh!!!🎛️👌🏻⚡🤖

Amazing content, Aaron! Looking forward to more.

@superpie0000

3 жыл бұрын

i cant belive im not paying tuition for this

Amazing lecture once again Thank you! I get the gist.. but left me wanting more...you did that on purpose. Bergfotron is BIG!!

Awesome!! Thanks alot dude!

Thank you for your effort, professor.

Yessss

I have a lot of xr2209 lying around, voltage controlled oscillators that produce square and triangle waveforms, so it would be great to spice these up with a sawtooth waveform as well. However I don't have any 4066 analog switches lying around 😕 But I do have a couple of DG508A analog multiplexers 🙂 Aren't analog multiplexers essentially just a bunch of digitally controlled analog switches? Can I use these DG508As in place of the 4066? I was thinking to connect the enable pin to the high rail and the squarewave to the 0th control bit pin to select between inverted and non inverted triangle.

What about taking the triangle and inverting it when the square goes low? If the square is in phase with the triangle so the falling edge of the square lines up with the peak of the triangle and the rising edge lines up with the trough, that should result in a sawtooth, I think. Is that a big "if"? That was my solution when trying to come up with a way of getting other waveforms out of a programmable sound generator like an AY-3-8910 (getting the triangle from a square with an integrator).

So if you want a VCO with both sawtooth and triangle outputs, what's the advantage of using a sawtooth core VCO and deriving a triangle wave from it vs the other way around? Sawtooth core + triangle conversion is simpler, but triangle core + sawtooth conversion is more precise since you don't have to discharge a capacitor very quickly each cycle?

@Lantertronics

3 жыл бұрын

Indeed, the fact that you'll see so many of each kind of design, without any one kind coming to dominate designs, is an indicator that they're pretty evenly matched. I think you've hit on the trickiest part of the sawtooth core, which is that discharge time that takes up more and more of the waveform as you go higher up in frequency. That needs to be compensate for in the exponential control-voltage-to-current conversation, which we will probably cover later this week.

@g.d.8065

3 жыл бұрын

There are also differences in the way syncing works with different cores (saw cores discharging when given a sync pulse and triangle cores changing direction when given a pulse).

@AnalogDude_

2 жыл бұрын

Than you should buy the Make Noise STO, i got it and scoped it, the bottom side (direction change) of the wave is angular, but the top is rounded off. but that could be caused by the type of capacitor, not sure., it's probably SMD in the Make Noise STO.

Hey, professor, i found another way to extract the saw wave and it's symmetric, no offsets or fine adjustments needed and only 3 resistors instead of 6, like Buchla did... and doubled the saw frequency. but the cd4066 has to be used in bi-polar fashion. i think you will like it.

@Lantertronics

2 жыл бұрын

Excellent!

@AnalogDude_

2 жыл бұрын

@@Lantertronics thnx, would you do a review on buclha's"symmetry" circuit? redraw? Buchla schematics has quite a bit of stuff around the ....., probably alot of the 259 complex wave generator "device" has alot to do with the led / LDR reaction time in the vactrol.

@AnalogDude_

2 жыл бұрын

@@Lantertronics it doesn't seam to make sense around opamp (IC27), the vactrol, +15V, -15V and GND in between towards the switch (IC23). it seams to be also part of the some sort of the control voltage for the folder circuit. GND is closer to the IC23 while -15V has to pass both resistors in the vactrol. drawing Buchla_2590_3_2''.jpg

@AnalogDude_

2 жыл бұрын

@Lantertronics - Aaron Lanterman Professor, would you do a well prepared "top notch" video explination over the NPN and PNP transistors? but also give the PNP at least 50% of the attention. i did choose electro studies at school, but that was a long time ago. the NPN is clear, pretty much, anyone can handle that and turn on a led of or build amplifier. but i didn't knew that you can also pull on the base of the NPN transistor, the control current coming from an inverted opamp, like in the transistor VCA of the mini brute @ the Hack a Brute on Yusynth's website. normal you put a a tiny current as the base and expect a return of 100+ times at the emitter, hence HFE, gain. but a reverse current on the base of a NPN transitor? i have read the wiki on transistors, mosfet, transistors, diodes, pn junction. but the PNP remains a mystery. for most video on KZread are over the NPN. why is the emitter / collector inverted. why is the current mirror in chips and the "buchla VCO core" a PNP? when you can do the same with NPN's. what benefits the PNP? if you check the "BA6110" datasheet, page 2 and look at the "buffer" part at the end. the job is Q17 is clear, but what does Q18? as the Q8 and Q9 in the CA3080 datasheet. what about the diode (D1)? the CA3080 looks weird compared to the LM13700, BA6110, BA662 the weird current mirrors in the LM13700. here is a online simulator, in case your didn't knew, it works pretty well, but can't make a sine wave from the Buchla circuit, but for the rest it's pretty good, but bummer they don't have a CD4013 there, but much standard tools is there. could you build some most common transistor arrangements in the Falstad simulator, like current mirrors, push pull VCA, switching stuff using bi polar psu's. like going from negative -12V volt to GND, etc, etc. thing you ca encounter. compile to a video and add some comments? i don't have a resources to pay for a professor and have the remaining classes to fully understand electronics, but i sure would ... ... and a video on how calculate the correct capacitor value would be cool as well. I = C x (dv / dt) it's weird, like the huge 220uF capacitor on the Moog Ladder Filter for a voltage divider witch barely consumes. thanks for your time.