My first design! Another tremolo. Feedback appreciated

[sigmoid]

I designed and built my first pedal, the Viking Tremolo (not trying to steal your thunder deadastronaut, just got my documentation done today!)
I wanted something that wouldn't alter my tone, had nice LFO waveforms, and kept the "on" volume constant, meaning no volume adjustment is necessary.
I don't have formal electronics education and little experience, so I'd love to hear any suggestions/criticisms.
I'm not trying to get free EE advice for some ill-advised commercial venture, just trying to learn.


The schematic is below and at the bottom is a pdf with an easier to read schematic and a circuit writeup.
I'm happy to answer any questions.





http://www.aronnelson.com/DIYFiles/up/vikingtrem.pdf

[armdnrdy]

Welcome Sigmoid.

It looks good...do you have a sound sample?

If I may make a suggestion....it would be better to label the pots (speed, depth) and the JFET (type) on the schematic for clarity.

[midwayfair]

Looks pretty cool, sort of a simplified TR-2 but with what seems to be more accurate waveforms.

One thing I've had some trouble with in the past is distortion across Boss-style FETs, but it looks like maybe C7/R8 are to combat that?

[sigmoid]

Thanks armdnrdy! Thanks for the suggestion, here's an updated schematic:



Here's a crappy video I made:

http://www.veoh.com/watch/v71770575TRx4Rf6E

midwayfair, I hadn't thought to look at the TR-2 schematic, definitely similar. You are correct that C7/R8 and also R7 are to correct the non-linearity of the FET. The drain source resistance formula has a -Vds/2 term in it, so by adding half the drain voltage to the gate the fet can become very linear. This page was very helpful: http://sound.westhost.com/articles/vca-techniques.html.

[armdnrdy]

It sounds good.

Good job!

The sine sounds better in a tremolo than I imagined it would. It sort of has its own unique texture about it.

Question:
Any particular reason you went with the J300B?

[mth5044]

That sounds great! Also fun to watch the waveforms fly across the screen.

For UD1, I had read somewhere that the positive side of the opamp should go to somewhere between the power supply rails, such as 4.5V, rather than ground or 9V. Doesn't seem to matter for you as it's working, just thought I'd point it out.

Very cool trem!

[armdnrdy]

For UD1, I had read somewhere that the positive side of the opamp should go to somewhere between the power supply rails, such as 4.5V, rather than ground or 9V. Doesn't seem to matter for you as it's working, just thought I'd point it out.

You are correct Matt,

U1D non inverting input should be at 1/2 V.

It is connected properly for a bipolar supply.

[mth5044]

You are correct Matt,

U1D non inverting input should be at 1/2 V.

It is connected properly for a bipolar supply.

That's a first! You aren't saying it is a bipolar design, right? Seems it could just be connected to the reference voltage coming out of U1A.

[sigmoid]

The sine sounds better in a tremolo than I imagined it would. It sort of has its own unique texture about it.

Question:
Any particular reason you went with the J300B?

I built this one for a friend, but when I make one for myself I might just forgo everything but the sine wave because I like how it sounds so much. I used a J300B because out of the JFETs that I had on hand it had the most convenient Vgs-off. Which is unfortunate because I can't imagine that it's an easy component to find. You could probably substitute many different parts but the depth wouldn't go all the way down. A JFET with Vgs-off of 2.4v ideal if you want to use this circuit unmodified.

For UD1, I had read somewhere that the positive side of the opamp should go to somewhere between the power supply rails, such as 4.5V, rather than ground or 9V.

I had plans for that last op-amp but they didn't pan out and then I wasn't sure what to do with it. I do have it tied to ground as in the schematic so I guess it's probably wasting a lot of current or something. It wouldn't be hard change to 4.5v if that is best practices. Do you know of any references that talk about this or am I best off staring at the TL07x datasheet for a while?

[armdnrdy]

http://e2e.ti.com/blogs_/archives/b/thesignal/archive/2012/11/27/the-unused-op-amp-what-to-do.aspx

[sigmoid]

http://e2e.ti.com/blogs_/archives/b/thesignal/archive/2012/11/27/the-unused-op-amp-what-to-do.aspx

This is perfect, thanks!

[deadastronaut]

not trying to steal your thunder deadastronaut,

no worries man, sounds really good, nice to see and hear the shapes too...

[samhay]

Cool, and welcome to the forum.
Would it not work better with a series resistor between U1B and the FET?

[sigmoid]

Sam that's an interesting idea. Potentially I could attenuate the signal there and get rid of the voltage divider at the input. That would improve the input impedance of the circuit. What were you thinking would be the benefits of a resistor there?

[sigmoid]

no worries man, sounds really good, nice to see and hear the shapes too...

You have no idea how much time I waste staring at the shapes on my oscilloscope 

[deadastronaut]

ha ha, at least you have one to stare at....