What your LFO does while you're not looking

[Williamblake]

Blessings to you all!

I would like to start a conversation about the low frequency oscillator in tube amps. There are a lot of threads about thumping but i am begging you to let this conversation be about the LFO signal generation more than about appliyng it to signal. Exceptions apreciated where necessary.

There is Merlin and Aiken givving the basics and mostly you have a triode, sometimes a pentode generating the oscillation.  Some circuits do several stages of high pass before applying it to the guitar signal.

Reading Merlin i expected a sine but ever since putting my scope to work i know a sine wave is not the first signal being generated by an LFO circuit.

I am hoping for LFO enthusiasts posting their circuits and a screenshot of the LFO generated.

There are tremolos that work well enough but my point is to have slower usable signal and to evaluate what does what. So right now i can go very slow but my oscillation looks more like a heartbeat and nothing like a sine wave.

Sorry for not starting with the fodder i want, but i am throttled right now.

Would love to see some LFO scoped for consideration. So what rocks your boat and what does it look like?

Regards
Jens

[sluckey]

Here are a few. Filename is the name of the project...

[sluckey]

I have others too, but these are the only ones that I had scope pics for. All are sinewaves, some prettier than others. I have none that look like a heartbeat. I would like to see a schematic and scope pic of your oddball. Indicate on the schematic where the scope is connected.

[sluckey]

Still waiting to see your 'heartbeat' pics.

[pdf64]

If tuned for a lower than standard frequency range, LED biasing is a really good idea.

[Williamblake]

Have a good year everybody, thanks for looking into this and your input, too.

As you may or may have guessed i am totally whacking it up but here is the circuit and stuff that got me extrapolating way above my weight and what happened then.

Looking at again, i think the oscillator has the wrong bias voltage.

The oscillator is oscillating on itself, not having a signal output.
The LED is red.
I am happy, i didn`t hurt the scope.

Voltages are (measured):
B+ 300V
Pin 7 Anode 150 VDC
Pin 8 Grid 2 130 VDC
Pin 3 Cathode 3 VDC
All measured while oscillating. (How did i do that?)
Cathode current is approximately 1mA, so EF80/6BX6 shouldn´t be too happy.

The motivation to use an EF80/6BX6 as an oscillator is because they are plenty and cheap and maybe could do the job. I could swap this for a 12ax7 with a cathode follower and be happy but then again EF80s/6BX6 come raining down on me.

Regards
Jens.

[Williamblake]

Also i hope you could get a good laugh while i was messing about with the scope and you probalby noticed how the oscillation amplitude changed.

[Williamblake]

And the variable resistor in addition to the LED is turned to zero.

[Williamblake]

And to tell noting but i am not a heart scientist.

[HotBluePlates]

... have slower usable signal and to evaluate what does what. So right now i can go very slow but my oscillation looks more like a heartbeat and nothing like a sine wave. ...

Does the EF80 really have a 1µF screen bypass cap, or is that 0.1µF?

I ask because the screen bypass cap sets a low-frequency limit on a pentode, just as a cathode bypass cap does.  I am unsure how the LED interacts with the screen bypass, so I might take that bypass string to ground (at the ground-end of the cathode LED).

The 47nF - 6MΩ R-C sections are pretty massive, and probably contributing to your issue.  They should be -3dB around 0.56Hz.  But if your screen bypass was actually 1µF, it is -3dB with the 470kΩ at about 3.4Hz. 

That would mean the tube gain is way down before the range of the feedback circuit.  I suspect the oscillator circuit doesn't "fire" until that screen bypass charges up, but I haven't thought through it thoroughly.

Try reducing all those caps to 0.01µF (10nF), or at least 0.022µF (22nF).  That will bring the oscillating frequency up, but perhaps more in-line with the low-frequency roll-off of the tube.  Alternatively, you might need to switch to a 12AX7 with an LED (and no series resistor).

[Williamblake]

Thanks for your input. I almost ditched this because even if it would be working it wouldn`t justifify the current draw a 6BX6/EF80 tipically has but i gave this this the twelth chance we all wish we had and this is what is is looking like now. No sine in the oscillator still but maybe it could be.

I did connect the osciloscope differently to wehre i would want the output of this circuit want to be, it is a 1Meg load and this suits fine, as i want to apply this to grids.

[Williamblake]

Did forget to post the voltages while the oscillation signal is connected to ground.
B+ 303 VDC
Pin7 Anode 219 VDC
Pin8 G2 224 VDC
Pin3 Cathode 5,5 DC
Also the jump in oscillation signal disappeared.

[sluckey]

Are you looking for a pentode LFO? Or are you just experimenting with that EF80?

There are some proven pentode LFOs out there. Here are a couple examples...

     https://el34world.com/charts/Schematics/files/Silvertone/Silvertone1482.pdf

     https://el34world.com/charts/Schematics/files/Silvertone/Silvertone1474.pdf

[PRR]

I think your G2 bypass could be 10uFd, not 0.8uFd. While the G2 resistor is large the dynamic internal resistance at G2 is much lower, even 20k, maybe less. And you do want gain to be constant as trem rate is changed. 10uF at 20k makes 1Hz, which is maybe generous, but not expensive, and covers much unknown. At least tack a 10u-47u in for a try. (The 47u may take a minute to come-up from zero.)

[Williamblake]

Thanks again for your input everyone. I would really like a good use for EF80s because they are plenty. Did try to find pentode oscillation examples but couldn't find them. Very much appreciated but will save those tubes to build PRR's third Watt amp. Will go ballistic on the screen bypass now. Boy this is fun.

[Williamblake]

Thanks again for looking into this. It has been already hinted at but i didn't get it. Output impedace is key. This is what it looks like now:

Screen resistor is 150k and plate is  40k.

[sluckey]

Adjust the scope triggering!   

[Williamblake]

So this is what i arrived at.
Increasing the first coupling cap immediately distorts the sine wave and lowering the cathode voltage does the same whith the amplitude increasing some. While i would call this working it looks finicky to me and maybe a new use for 6BX6 is found but it may not be species-appropriate and other tubes do the same job better.

Plate is at 263 VDC  1 mA
Screen is  255 VDC .3mA
Cathode is 6.4 VDC

Also the range is not as large as you would expect from a dual ganged pot i think.
But i`ll keep it, maybe i stumble on another improvement sometime and i want to hear what wobble this circuit can do before i call it a failure.

[Williamblake]

And this is the same circuit you saw in the video above with the bad triggering. While i found a knob called Niveau on my scope it did not alter my mind the way i expected and i did not record the wave again. Looks like a sine to me, the prove is in the pudding, i want to hear it ticking. Or maybe i don`t.

[PRR]

> not as large as you would expect from a dual ganged pot

Because the three R-C products are not close to each other.

Best you can do with two 7:1 Rs is 3.6:1 of frequency. But you are 4:1 off at the first R-C (2n vs 8n).

[Williamblake]

I had i had a hard time finding an intermittent connection which i knew would be obvious once i found out where it is but what can you buy with that? It was to the cathode resistor, if you wondered. Anyways, this it is now.

Plate      280 VDC
Screen   275 VDC
Cathode 6,5 VDC
HT         300 VDC

So would you want the plate and screen to be 2/3 the HT in a pentode? I am far away from that and the current is pretty low. The wave gets more sinusidial when i keep it this way, lowering the cathode voltage distorts the sinusidial wave form.
Also notice how the amplitude now again changes across the frequency more than before and what a scope does when you turn the knobs the wrong way.
I changed the screen bypass from 47u to 4u just for size, maybe i will look at that again.

Regards
Jens

[Williamblake]

Sorry for the lengthy video, but i have to show off the dual-antilog-pots i found today. Also i never had that much of a near sinewave across the whole speedrange. I measured 5Hz to 15Hz. Voltage swing is 8V.

[sluckey]

Dual pots make a difference. Triple pots make even more difference. Look at an old Sunn schematic.

     http://sluckeyamps.com/sunn/sceptre1.jpg

[PRR]

Your pentode pin-out is drawn wrong. From plate: G3, G2, G1. Doesn't really matter except it makes kibitzers wonder "is that right?"