EA Tremolo, mod to get it smoother?

[fiestared]

Hello! I just built a EA-tremolo from effect pedals kit
Schematic and BOM:
https://effectpedalkits.com/wp-content/uploads/manuals/ea-tremolo-kit-building-manual.pdf

It is from what I understand an improved version of the EA circuit. I lack some smoothness in the swells, finding it hard to find the "fender-amp-like" tremolo that some seems to praise in this circuit. Is there any mods I can do to maybe make it smoother?

[antonis]

Despite the fact "fender-amp-like" tremolo actually refers on a vibrato, issue here is Q3 channel resistance non-linearity..
(you may try to replace it with a 2N5457 or 2N5484 although, IMHO, you'll result into minor improvement..)

[fiestared]

Despite the fact "fender-amp-like" tremolo actually refers on a vibrato, issue here is Q3 channel resistance non-linearity..
(you may try to replace it with a 2N5457 or 2N5484 although, IMHO, you'll result into minor improvement..)

Thanks for your input! I guess what I want is the swells to be "smoother" in lack of a better word.Thankful for all ideas.

[Mark Hammer]

I've often wondered about sticking a back-to-back pair of diodes, perhaps LEDs, to ground between R11 and the Depth pot to round off the peaks of the waveform and make it a little more sinusoidal than triangular.  The details of how to implement that elude me, though

[duck_arse]

fiestared - PLEASE - delete C3 from your build, and link across the empty pads. how many times are we going to see that particular error copied across "versions"?

as for smooths - every time I listen to an un-modded phase shift oscillator in a frem-a-like, I hear yanG-yanG-yanG-yanG in the modulated envelope - is this what you are referring to? you could try and un-distort the output with a pair of back-to-back leds [red are good, doesn't matter size of clear/tinted] across C9 is the oscillator. this should cut the amplitude of the oscillator output, allowing it to remain unclipped.

edit - see reply #25.

[fiestared]

fiestared - PLEASE - delete C3 from your build, and link across the empty pads. how many times are we going to see that particular error copied across "versions"?

as for smooths - every time I listen to an un-modded phase shift oscillator in a frem-a-like, I hear yanG-yanG-yanG-yanG in the modulated envelope - is this what you are referring to? you could try and un-distort the output with a pair of back-to-back leds [red are good, doesn't matter size of clear/tinted] across C9 is the oscillator. this should cut the amplitude of the oscillator output, allowing it to remain unclipped.

Thank you! I think that is what I am talking about, yes! Could you please specify more exactly how to do the "back to back leds"? And you mean removing the Q3 and just solder a wire connecting all all the three holes/soldering points?

[duck_arse]

C3 only shows as two solder holes on your build docs. it is not an important thing, but it annoys me whenever I see it. you can remove the cap or simply solder a wire across the cap pads with the cap still fitted, it will work the same.

as for the leds - I can't guarantee it will work, but it does on variations of that oscillator I've tried. it's a bit tight on your board, too. you just get two leds, add them so the cathode of one points to C7 and the anode to C8, and the second led goes cathode to C8 and anode to C7, just like clippers in a distortion.

[fiestared]

C3 only shows as two solder holes on your build docs. it is not an important thing, but it annoys me whenever I see it. you can remove the cap or simply solder a wire across the cap pads with the cap still fitted, it will work the same.

as for the leds - I can't guarantee it will work, but it does on variations of that oscillator I've tried. it's a bit tight on your board, too. you just get two leds, add them so the cathode of one points to C7 and the anode to C8, and the second led goes cathode to C8 and anode to C7, just like clippers in a distortion.

Sorry, misread C3 for Q3. Will try this, and the diods.

[Mark Hammer]

fiestared - PLEASE - delete C3 from your build, and link across the empty pads. how many times are we going to see that particular error copied across "versions"?

as for smooths - every time I listen to an un-modded phase shift oscillator in a frem-a-like, I hear yanG-yanG-yanG-yanG in the modulated envelope - is this what you are referring to? you could try and un-distort the output with a pair of back-to-back leds [red are good, doesn't matter size of clear/tinted] across C9 is the oscillator. this should cut the amplitude of the oscillator output, allowing it to remain unclipped.

C3, not Q3.  Q3 stays where and as it is.  C3 is redundant.

Thank you! I think that is what I am talking about, yes! Could you please specify more exactly how to do the "back to back leds"? And you mean removing the Q3 and just solder a wire connecting all all the three holes/soldering points?

A pair of red or maybe even green LEDs can be run from the input lug of the Depth pot to ground, to "clip" the LFO waveform a bit and round off the top and bottom of the waveform.  Again, I do not know if that is sufficient, or "best".  I'll leave that up to those with far more electronics knowledge.  I will just note that more sinusoidal modulation of pitch and volume tends to sound more "musical" to us.  One of the crucial features of the old Boss CE-1 that made it so desirable was that it used some diodes to "round off" the LFO when the unit was set to vibrato, yielding a sweeter modulation.

[fiestared]

[/quote]
A pair of red or maybe even green LEDs can be run from the input lug of the Depth pot to ground, to "clip" the LFO waveform a bit and round off the top and bottom of the waveform.  Again, I do not know if that is sufficient, or "best".  I'll leave that up to those with far more electronics knowledge.  I will just note that more sinusoidal modulation of pitch and volume tends to sound more "musical" to us.  One of the crucial features of the old Boss CE-1 that made it so desirable was that it used some diodes to "round off" the LFO when the unit was set to vibrato, yielding a sweeter modulation.
[/quote]

Thank you! Is this instead of the previous C7 diods addition or can i do both? And I am new to this but will like to learn. Can you please explain in detail how this is done?

[fiestared]

C3 only shows as two solder holes on your build docs. it is not an important thing, but it annoys me whenever I see it. you can remove the cap or simply solder a wire across the cap pads with the cap still fitted, it will work the same.

as for the leds - I can't guarantee it will work, but it does on variations of that oscillator I've tried. it's a bit tight on your board, too. you just get two leds, add them so the cathode of one points to C7 and the anode to C8, and the second led goes cathode to C8 and anode to C7, just like clippers in a distortion.

Just to be sure here, i dont need to remove c7 and c8 and then add the diodes, i JUST add the diods to the same soldering points as the C7 and C8 already are?

[EBK]

I just recently built one of these from this layout:

http://tagboardeffects.blogspot.com/2012/02/ea-tremolo-modified-rog.html?m=1

I'm too tired to check if I also have a C3 problem.  Mind taking a peek?

[Mark Hammer]

C3 only shows as two solder holes on your build docs. it is not an important thing, but it annoys me whenever I see it. you can remove the cap or simply solder a wire across the cap pads with the cap still fitted, it will work the same.

as for the leds - I can't guarantee it will work, but it does on variations of that oscillator I've tried. it's a bit tight on your board, too. you just get two leds, add them so the cathode of one points to C7 and the anode to C8, and the second led goes cathode to C8 and anode to C7, just like clippers in a distortion.

Just to be sure here, i dont need to remove c7 and c8 and then add the diodes, i JUST add the diods to the same soldering points as the C7 and C8 already are?

NO!  C7 and C8 are absolutely necessary to producing the LFO.  What I suggested (and again, let me emphasize that it is untested and may need to be altered a bit) was to shape the waveform that comes from Q4, via R11.

It seems that JD Sleep may have been the initiator of the erroneous redundant capacitor, discussed here as C3.

As for the ROG version of the circuit, it only uses 3 transistors, sending the input signal directly to the variable-gain transistor, sidestepping the issue of redundant caps between a first and 2nd stage as in the GGG drawing.  So you should be good, Erik.

[fiestared]

I tried two red LEDs, the positive end of one LED combined with negative of the other LED, between the two positive sides of the C7 and C8. (Nothing soldered between the Caps negative ends). I din this on the underside of the board. When i tested, I got no effect, just clean signal. So i removed them. Did i do it wrong? Any other input to get morw sinusoidal/musical tremolo effect?

[Mark Hammer]

I don't understand why you are bringing C7/C8 into this.  The suggested mod is shown here (I've rearranged some things and shortened lines to make the image more compact)  The LED pair simply go from the junction of R11 and the DEpth pot to ground.  That's it.

[fiestared]

I don't understand why you are bringing C7/C8 into this.  The suggested mod is shown here (I've rearranged some things and shortened lines to make the image more compact)  The LED pair simply go from the junction of R11 and the DEpth pot to ground.  That's it.

Thank you so much Mark for trying to help a complete newbie. C7 and c8 was mentioned before. I am not use to reading schematics, I tried what i think is right? Wiring the LEDs together (but reversed from each other) between depth lug 3 and r11..

[Fancy Lime]

Sorry to jump in out of left field here but isn't a phase shift oscillator supposed to produce a sine wave anyway? And wouldn't diodes clipping the waveform make it more squarish and thereby less smooth? And shouldn't a triangular wave sound smoother than a sine wave because the zero transitions are less steep? Maybe a triangular wave with slightly cropped peaks would be ideal? Again, sorry, I am really not an expert on LFOs.

Also,, what about putting a 1M linearization resistor between G and D of Q3, so that the LFO waveform translates better to the actual modulation? As it is, I would expect some rather asymmetrically distorted sine wave shape in the volume profile.

Andy

[Rob Strand]

It's not an easy thing to fix.  It will be pot luck if you can get a diode clipper to work.

There two things that happen with the EA tremolo:
- When the depth is low the nature of the JFET gain control it spends more time in the high-gain region (LFO positive) than the low gain region (LFO negative).   The sounds isn't even but it is still smooth. 
- However when you crank the depth control the JFET the sound gets choppy because the JFET is completely cut-off and the tremolo spends more time in the low gain region.    It does that in a less smooth manner.   

If you don't like the first case you might be able to tweak it with diodes.    The second case needs major work since you have to prevent the JFET cutting out.    Once the JFET is cut-off addition of diodes will not help - the JFET has no control over the gain.

The thing to realize on the EA tremolo is the cap between the LFO and the depth control charges up.   It uses the JFET gate as a rectifier which shifts the LFO output on the JFET side down so the positive peaks are at about 0.6V (roughly).    The gate waveform then swings below that and ideally swings between -VP to 0V  where VP is the gate cut-off of the JFET.    However we don't *have* to do that, if we let the gate swing below -VP the will JFET cut-off you get a choppy sound.

The inherent self regulating nature of the positive LFO peak is a big bonus since it is automatically adjusting the LFO output to match the JFET.    The adjustment on the negative side is done by the user.  The result is the unit works and there's no trimpots to adjust.

If you want to prevent the negative LFO swing becoming choppy you need a way limit the LFO swing to VP then on top of that you want to modify the shape of the LFO.      No doubt that will require trimpot adjustments to match the JFET.     To help this case case you might need to also make the lowest gain point lower.   Ages ago member Transmogrifox replaced the 1.2K emitter resistor with a current source.   This helps extend the lower gain side.

[Fancy Lime]

To Rob's point: aJ201 strikes me as a distinctly odd choice for Q3. I would expect much less problems with dropping below Vgsoff if you used a JFET with a Vgsoff that is more in the range of what the LFO puts out. I'd try a J112 or something like it and see where that goes. If you have one sitting around, that is.

Andy

[Mark Hammer]

Like I said,I expected folks with more actual knowledge about these matters to chime in, and it seems they have.  So, thanks. 

Q3 is being used as a variable resistor.  But it occurs to me that in one of the other major ways we use JFETs as electronic resistors, in conjunction with an LFO - namely phase shifters - a bias voltage-current is provided and the LFO takes it from there.  The basic EA design certainly works, but I wonder if it would work better and sound less choppy, or whatever, if a trimmer was used to provide a basic bias to the gate of Q3, and the LFO simply added to that, providing variation in Q3's drain-source resistance that was always smooth and musical.