Help debugging an LFO

[tubegeek]

Does this wave shape mean anything to anybody? The rate changes with rate pot movement. Near one extreme the oscillation dies. Then it rejuvenates with the knob back the other way. Then it stays alive and changes ƒ for most of the adjustment range, until it dies back at the other end where it died before.

Schematic is the LFO section from this, the Fake Fender Vibe from GEOFEX.



I built it including the 1 uF output caps and the 1M pulldowns at the outputs, the circuit ends there and I sent it into a recording interface to get a look.

Here's the waveform I am getting from the two outputs. The top (small sine) trace is from the emitter of Q9, the bottom trace (big ragged wave) is from the collector. Two zoom levels, same waves. Does this trace mean anything to anybody? I took a pretty careful look at my perfboard, nothing jumped out at me so far.






Voltages:

Q7 CBE: fluctuating 0.33-0.38 ; 0.63 ; 0.45
Q8 CBE: 9.4; 2.93 ; fluctuating 3.5-3.8
Q9 CBE 6.92; 2.66, 2.39 (readings around Q9 were fluctuating, stopped moving, I assume it's the oscillator stopping?)

The only resistance that looks weird to me is that Q7 Emitter reads short to ground, not 100R. I'll try reinstalling a new resistor in the morning - I don't see any path for that bedsides through the resistor, maybe it's actually a 0R jumper (?!) or failed short from soldering (?!!¿)

[PRR]

Yes, find and fix the solder blob emitter to ground.

You need gain roughly 30 to oscillate. You need a higher gain to start, 1.5X to 2X. Wild swings will de-bias the transistor to the gain needed to just sustain oscillation.

Starting gain of Q7 should be near 10k/(60+100) or 62. Without the 100r it is 166. The large excess gain forces huge de-bias and pretty wimpy operating point when/if it stabilizes.

The waveform is not much use to the inexperienced (me) because the phase-shift network can modify clipping into many sorts of rounded waves.

[j_flanders]

The only resistance that looks weird to me is that Q7 Emitter reads short to ground, not 100R. I'll try reinstalling a new resistor in the morning - I don't see any path for that bedsides through the resistor, maybe it's actually a 0R jumper (?!) or failed short from soldering (?!!¿)

Digital MultiMeter? Reading the resistance or continuity check?
My DMM 'beeps' on anything below 50 Ohm, so a 50 Ohm resistor beeps as if it were a piece of copper wire when checking continuity.
Obviously, when switching to resistance/Ohm reading, it shows 50 Ohm.

Maybe some DMM's (yours?) do it to 100 Ohm and below?

[ElectricDruid]

The waveform is not much use to the inexperienced (me) because the phase-shift network can modify clipping into many sorts of rounded waves.

Additionally I've been caught out in the past trying to use a computer to look at LFO waveforms. Turned out the computer's audio input was AC-coupled (as I'd expect if I'd thought about it) so there was an effective highpass filter on the input which played hell with my LFO waveforms.

[tubegeek]

The only resistance that looks weird to me is that Q7 Emitter reads short to ground, not 100R. I'll try reinstalling a new resistor in the morning - I don't see any path for that bedsides through the resistor, maybe it's actually a 0R jumper (?!) or failed short from soldering (?!!¿)

Digital MultiMeter? Reading the resistance or continuity check?
My DMM 'beeps' on anything below 50 Ohm, so a 50 Ohm resistor beeps as if it were a piece of copper wire when checking continuity.
Obviously, when switching to resistance/Ohm reading, it shows 50 Ohm.

Maybe some DMM's (yours?) do it to 100 Ohm and below?

Reading ohms in the 200 range, I can go to 100R (single digits even) with this meter. And re-checking today I get 100R. WTF?

[tubegeek]

The waveform is not much use to the inexperienced (me) because the phase-shift network can modify clipping into many sorts of rounded waves.

Additionally I've been caught out in the past trying to use a computer to look at LFO waveforms. Turned out the computer's audio input was AC-coupled (as I'd expect if I'd thought about it) so there was an effective highpass filter on the input which played hell with my LFO waveforms.

I'm using a decent usb interface's instrument level input, no pad. I'll look at the spec but I think it's speced for below 20 hz. Good thought though!

.... manual says 10Hz - 50K ... hmm .... that could DEFINITELY be a factor, a trem LFO is below 10Hz for much of its adjustment range, right? Just weird that the two outputs read so different in level...

You guys are gonna make me get the scope up out of the basement aren't you? I know I'll need it for this project but
• it's frickin' heavy
• it's in an awkward spot to get to.

Oh well, gotta man up and schlep that thing upstairs.

[ElectricDruid]

No, we were going to make you walk to your friend's house round the corner where they've got a nice scope already set up!

T.

[tubegeek]

No, we were going to make you walk to your friend's house round the corner where they've got a nice scope already set up!

T.

LOL!

OK, so I compromised: I brought the scope up but not the vintage Tek trolley it sat on, which was kind of wedged in the corner. A nice item but not worth the effort to rearrange all the junk vintage parts boxing it in.

The scope says the same thing the interface pic said only with actually less detail - the interface has more gain and the display on the recording app has better clarity. One trace very low amplitude and sineish shape, the other trace larger but still small amplitude and ragged shape. Something isn't working right and so far I can't find it - all the resistances to ground and to 9V that I've measured look spot-on. There must be a wrong connection within the phase shift oscillator loop because the transistors all seem to be biased sortof OK and almost nothing's coming out. I'll check AGAIN tomorrow.

[tubegeek]

OK Working much better. Reflowed some solder joints. I paid closer attention to the caps (because all the resistors seemed kosher) and while I was thinking about the caps I remembered to check the polarity of the one electro. Sure enough, I found that C15 electrolytic was in backwards. Also found that I may have had the Depth trimmer down all the way.

Now I have much bigger and symmetrical waves. The oscillator still has a tendency to stop running sometimes when the Rate extremes are reached, I'll work on that next chance I get, increase the series resistors to 2K or whatever.
Cool oscilloscope display with the slow, opposite waveforms.