Tremulus Lune Improvements

[liquids]

I realize the Tremulus Lune is an old standby circuit and I haven't been around that long.

I've had it breadboarded on and off for months now, and think I've got a few improvements...but I respect Dan Green and his logic.  Thought I would put them other there for others to pick apart, or maybe utilize.

1) It seems Dan wanted the output and input signals to be in phase...  I'm not quite sure why that's necessary, but ok.   I see that for the second op amp stage to work properly, it needs to invert, so the input buffer is likewise inverting.  However, such an inverting input buffer is either quiet but dark due to impedance--as I found the pair of 220k as on the 4ms website right now...and the Tonepad version is 47k...lets not go there.  Upping the values to 1M makes the stage potentially clear but noisy.  So unless THIS buffer stage is being fed another  (non-inverting) buffer stage, it's a tonal compromise, not ideal.  And If there's no room for another op amp, why not just add a simple emitter/source follower first, and then we can use the 47k or 10k resistors  in the inverting op amp buffers, preserve tone and in-to-output phase?

Actually, instead, I chose to keep it simple and accept the phase reversal, so I went with making the input buffer non-inverting.  I couldn't get my layout any smaller than it was, so I accepted the 'phase issue' as a better compromise than noise, size, or tonal quality.  YMMV.

2) Why put DC right on the LDR resistor?  Generally speaking, isn't voltage on any variable resistor in the audio path bad design?   The second stage is already seeing VREF via the non-inverting input....  Just putting a cap between the buffer op amp's output and the LDR solved my ticking issue, plain and simple.  Why isn't there one, design wise?  It definitely seemed to help ticking...there must be a reason, even if a poor one...

3) Spacing pot:  Why wire this up as two resistors and a VARIABLE resistor?  What a headache...if I recall, turning it with the stock arrangement not only changed the spacing but seemed to interact heavily with other pots. I remember being pretty confused about what the control was even doing due to the interactions it caused.   Why not change the values of the two resistors, and make the pot more ideal as a constant resistance splitting between one side or the other for biasing, with the WIPER feeding the LFO op amp input?   I used 220K from positive to one side of the pot, and 150k going to ground from the other side of the 500k pot.  I initially did the math to make sure the 'stock' ratios were covered, and then also expanded the range a bit for my ear, so it can be more heavily 'off' now than before, to a point I found useful, but with greatly reduced interaction.   Seemed easy to me....  

4) Note that I moved the gain pot to external and made it 25k, since going so far between into "mostly off' territory makes the perceived volume low, and I needed real-time adjustment to recover gain over an internal trimpot adjustment.  Nothing special there.

5) the 10nF cap from inverting to output on the first LFO op amp stage is there to help ticking, it seems.  But it also slows the LFO, compared to it not being there at all.  If you pull it, the LFO speeds up.  But, ticking is then audible.  Simply changing it to 1nf in conjunction with the additional cap in the audio path as mentioned above, the ticking is quieted, and it seemed to affect the LFO speed minimally, if at all.  An easy compromise, I thought.

6) two pots for speed is silly to me for those of us who have access to reverse log pots.  I've got a 'C100K' pot I'm going to use, in conjunction with a 1uF cap and a 15k 'set' resistor for the range.  C (reverse audio) pots make more sense for LFOs like this, for more natural curve, I've come to believe.  Using only one speed pot conceivably frees up some space to add an external toggle, to switch in another cap in parallel with that for the LFO.  this offers more range of speed, rather than just 'fine' adjustment.   Now I have a 'fast' range with just the 1uF in there, and 'slower' range adding a 3.3uF cap in series (for a total of ~ 5uF). This could even be wired to a footswitch for fast/slow, and of course, can be tweaked by ear for what speeds you'd personally use via making the 'set' resistor or caps smaller or larger...

These are all simple tweaks, I thought, but not one's that are 'drop in,' which means that I couldn't use the 'classic' vero layout, nor the Dan Green PCB, which I might have purchased, as a thank you to Dan Green.   So that is a clear disadvantage.

Anyone have thoughts on these concepts?   Are these all old-news mods?  

Note that I've yet to box it up, this was all on the breadboard.   It's an amazing tremolo, and I think it's end all be all tremolo for MY needs, now, after tinkering on and off with different ideas for, oh, a year or so.  

[Skruffyhound]

Great timing for me that you posted this, I'll build the Lune soon, but now I'll hold off etching a board and breadboard it first so I can try out some of this. Thanks, they all seem like sensible mods, I had just assumed that it had been tinkered to perfection after being around for so long.

[trjones1]

Points 2, 4 and 6 have all been covered in previous posts.  I just built the tremulus with 100k and 10k speed and fine pots.  I found that with the speed knob all the way up, the fine knob gave me as much range of rate as I would really ever want (when combined with a rate doubler switch).  But if for some reason you wanted to explore very slow speeds, then having the two knobs may be useful.

[liquids]

Great timing for me that you posted this, I'll build the Lune soon, but now I'll hold off etching a board and breadboard it first so I can try out some of this. Thanks, they all seem like sensible mods, I had just assumed that it had been tinkered to perfection after being around for so long.

I would have thought so too.  I starting drilling, but went back to the breadboard after reading Aston's (Skruffyhound's) comments.  Suddenly, for whatever reason, I got more ticking with certain combinations of the controls no matter WHAT I did.  Scrapped it and started again...same thing.  Not sure if it was new or I just hadn't tried these combinations before.

I like the breadboard since it is 'worst case scenerio' on a lot of things, in terms of ticking, capacitance, etc.  In cases like this, I figure, if I can get it good on the breadboard, a real build will be cake.

Anyhow, I was using a large 220uf cap across the rails all along.  But Adding a 100n across the rails, and then also a 1n  across the rails basically eliminated any ticking, unless I cranked the gain, in which case it was just slightly audible, but probably twice the volume of of the original signal.  Not completely sure why that should work, but it did.   Give that a try.  Also, this has probably been mentioned before, but a TL062 op amp for the LFO probably just makes sense for current draw characteristics.   The slower slew rate didn't seem to harm anything even at fast speeds.

[Skruffyhound]

Did you finally get it tweaked to perfection, any more recommendations Matthew, I'm starting to work on this now. Thanks

[liquids]

I'm using mine and enjoy it quite a bit!

The opto component you use will play a big part in how this sounds keep in mind.  Also note that some settings are virtually unusable.  For example, While I might use the depth at full when the 'spacing' control, as I have it, is set  to the extreme end of 'mostly on,' when you turn the 'spacing' to the opposite extreme, the depth control should be dialed more subtly.  So do breadboard it and tweak to what you like...and the components you use.

Also, it's just seems safer to add the extra caps across the rails - a large value, plus a 100nf and a small value (2n or so) or something along those lines.  For whatever reason that worked for me....it don't think it should, but hey. =D  I've had no ticking so far, except, it seems, when I have another project on the breadboard sometimes a faint low frequency ticking is audible.  Again, it seems related to my breadboard rig and the wiring which is always at least a bit unsightly with some stray wires.   Otherwise, all is good.

After it was built, the only mod I did, if I recall, was put a small value (2.7k?) resistor in series with the 25k volume control.  Since my tweaks allow it to go so far into 'mostly off' territory, you can see that the indicator LED barely flashes, meaning the optical element is 'barely flashing,' which means it needs a bit of volume to compensate if the 'depth' is extreme....the extra K helps.  Or just put a 50k pot in there.   But again, this is nearing some 'unusable' settings.  likewise With the spacing at max 'mostly on' settings and the depth set subtly, you need little volume to reach unity - the volume pot at 1-3k is good.

Likewise, experiment with cap values and the oscillator.  Even one cap vs to another may make a bit of difference.  I used a different cap type (ceramic multilayer) in my build than what was on the breadboard (panasonic poly), and ended adding a 220n cap in series with the 1uF for 'fast' and then adding another 2.2uF cap in series via the switch for slow.  I Forgot the stock resistor I have in there for the oscillator...15k, or is it 27k?    Anyhow, It meets my needs quite well, but it's all matter of taste if you're not a 'stock is good enough' person...which I'm assuming you aren't since you're replying!  

Feel free breadboard it and report back (here or via PM) if you have ways you want to tweak it if by chance you don't know how.  You want to narrow (or widen) some of the range of the pots into more 'only usable' ranges, etc.  It's a 5+ knob tremolo, so it's tweaker's delight, you know...only so much can be usable.  At it's core, I'm very, very happy with the mods/improvements!

[Skruffyhound]

Great thanks for that, I'll get to it! I've got to etch some boards in my two day mini holiday and get some things boxed so I'll probably take a while to get back to you but good to know that you got a result. My 3 breadboards are all occupied with Sub-mini projects, but now they are on hold waiting for an HV power solution, the Lune can get a look in.Thanks