Low-distortion Audio-range Oscillator

[1audio]

Originally Posted by zigzagflux

Photoresistor has around 15mV across it, as it is in parallel with a 15 ohm resistor. Sounds fairly low to me..

I looked at the attached part of the schematic many times since its confusing. I think the LDR is in series with the 100K resistor (limits its maximum effect and the voltage across it) and the combination is in parallel with what appears to be the 3.6K resistor and something I can't read on the schematic but I think is the 15 Ohm resistor you mentioned.

[stocktrader200]

save a ton of time and effort
iFunGen on the App Store for $1 20 - 21Khz triangle, sine, Square generator
for iphone or itouch

[1audio]

If the iphone were capable of low distortion it might be interesting.

Sent from my myTouch_4G_Slide using Tapatalk 2

[zigzagflux]

1audio:

Yes, it is confusing. I noted in post 174 that the schematic in the article had a number of errors, among them the missing node on either side of the photoresistor. So consider the junction of the photoresistor, R18, 100k, and 3.6k to be a node. It is not shown that way in the schematic, which is in error.

Did make some progress: the sidebands are due to distortion from too high an output level. At the time I was testing this earlier, I only checked the amplitude modulation with output level, which had no effect. Once I bring the output down to 3Vrms, I get much better results as you can see.

Annoying is the fact the amplitude modulation is still occurring, at +/- 12mV. Maybe this is the best I can expect, not sure. I have run the LED current from 6 mA (too low to control startup) to 15 mA (too high to control startup). In between works well, with 12mA settling quicker than 8mA. But each has almost identical modulation of amplitude.

I have scoped at the base of the transistor, and was able to smooth out the error signal by increasing the 100nF integrator cap to 3 uF, but this still did not resolve the output bounce (but does make it better). The 8.2K resistor right next to it does nothing, even when shorted out.

Pics also included of the sine output and output of the FW rectifier. Note the 2nd harmonic present at the output of the FW rectifier; every other valley is at alternating DC levels.

[richiem]

Just seems like the AGC control range is too small. I'd try making the 18R resistor larger, say 27R. I think I'd make the 10uF at the base of the transistor larger too.

[Terry McGee]

Nothing immediately obvious in the PCB layout. I assume you have something appropriate from the GND connection to the groundplane. I note R1 in the Wein Bridge is 54.9R rather than the schematic's 75R.

Can you tell us what the approximate frequency of the amplitude modulation is? Should be measurable or observable at TP1.

Terry

[zigzagflux]

Right, there are a few differences in the PCB. I hand-matched the input RC constant to be equal to the feedback RC. If you go through the numbers from the PCB it works out. The resistor from GND to circuit common is up for experimentation once I get the output correct. Could be resistor, capacitor, jumper, or open.

Frequency of modulation is only a Hz or two. It is a little inconsistent; it holds for a bit, then bounces around and settles, then starts up again. Checked the power supplies; rock steady.

I will play around with the 15/18 ohm resistor tonight or tomorrow. It was my thinking the less effect the photoresistor has on the gain the better, but now thinking about it you may be right; higher might be better.

[Terry McGee]

Oh, and I note you are using two separate diodes rather than the BAV99 pair. But I can't imagine that resulting in a loop stability issue.

By the way, why 2kHz? Remember, I've been wandering in the wilderness for some years. Has 2kHz suddenly become the new black?

Terry

[richiem]

@Terry -- well, it keeps the important harmonics inside the audio band for 44.1kHz sampling... Other than that, I haven't got a clue.

[zigzagflux]

Good question. I figured, based on the results from the article, that all PID control issues were specifically geared for a 2 kHz frequency, so best to leave well enough alone. Otherwise, I would have geared for 1 kHz. I had received some updates from the author over email, which are incorporated in this design, hence some differences from the original schematic. Had I obtained excellent results from this design, I have 2 more PCBs, and would have attempted some other frequencies.

Yes, I can't imagine those diodes requiring extra special attention, but I do note some unevenness in the rectifier output.