Tremolo Design Question

[makinrose]

For some time I've been using this LDR circuit for signal varying tremolo: http://sluckeyamps.com/trem/trem.htm. I've found it to be less noisy than the Fender version.  I've also used the Wide Range Trem that works well too.  I'm wanting to try a tremolo that wiggles the bias on a pre-amp tube by pulling in and out the cathode bypass cab like the Garnet example in TUT-2.  Could the LDR circuit I'm using be used to do that by hooking the negative end of the cap to it?  Before a start a prototype are there any design considerations I'll be needing to make?  Thanks Guys!

[K O'Connor]

Hi Makinrose

That circuit can just as easily use the Silonex LDRs, or any discreet assembly of a LED and light-dependent-resistor cell in heat shrink. It might be worthwhile to add a resistor in series with the diode to provide a current limit.

The author suggests the trem frequency range is wider but I always found the Fender range to go far too fast. Greater depth is nice to have.

As far as your proposal to link the LDR into a preamp stage via the cathode bypass cap of that stage, certainly that will work. The depth of effect might not be high if the basic gain of the stage is on the high side. You might also be able to replace the expensive LDR with a less expensive p-jfet. This might work better with the entire trem oscillator elevated by a volt or so to extend the gate drive range into the jfet.

Note that in the trem author's bandmaster circuit, the "improved" bias circuit is not much of an improvement. Were the link on the bias pot removed and a safety resistor added,things would be better. Also,a clamp for the bias filter caps would protect them. And as always, parallel bias pots for push and pull control allows hum-balancing of the output stage.

[makinrose]

Kevin,

Thank you for the thoughtful reply. I think you might have missed something here though. That circuit is not a bias vary trem. It for the Bandmaster with the signal vary tremolo. . As cool as those output stage bias vary trems sound I've experienced first hand why you discourage their use. It really tough on an amps....

I agree that this circuit does not have a wider range. The main advantage is the lack of ticking and noise issues compared to the Fender LDR ciruit. Also the footswitch is little more convenient since it does not require negative voltage from the bias supply. It is rather fast like the Fender circuit so I increased the value of the timing caps to slow it down.

You have an excellent point about the gain level of the stage I'd apply it too. That's definitely in need of consideration. The J-Fet solution really would be ideal....I'll try it out. Thanks!

Thanks!

[K O'Connor]

Hi Makinrose

I did not think it was bias-modulated - you could not use it that way due to the limitations of the resistive cell.

My comments about the bias supply are with respect to the linked article's further link to the amp the trem was installed into.

Reduced noise and ticking are big pluses. In traditional amps, those noises are the result of poor grounding and are easily abated.

Have fun

[makinrose]

Oh I got you.....sorry about that! Yeah wouldn't hold the rest of info on the amp up as prime example of good design.

At this stage I always design layouts for my amps from scratch using the TUT projects as a guide and spacing to fit the components I like. Adding proper bias circuits, decoupling, and grounding is just a given now. Buy has it made my life easier. It seems the vast majority of amp layouts and "improvements" fail to make common sense improvements.

[makinrose]

In case someone want to use this design, I wanted to note on this thread that the Vactrol VTL5C1 LDRs are no longer being produced. The replacement sold by CE Distribution by Xvive and the Chinese made ones on AliExpress say they are same on the data sheets but do not have the specs that are listed. They won't work properly. The best solution appears to make your own or to buy NOS versions.

[Sherlok Ohms]

Good day Makinrose

All good information! I suppose if one wants a "vactrol" with data available and that actually conforms to the data sheet, the only option is Silonex, at least as far as current production goes? If you make your own you have the option to use a real light bulb with a filament

Some of the alternative interfaces get a bit complex or expensive. The p-ch jfet option is running out of jfets to choose from, at least those with leads. I imagine one could use a more common n-ch jfet if there were an interface between the tube oscillator and the jfet gate, but this would also require a small negative voltage supply. The latter is easy enough to do using any of the less-than-ideal bias circuits since this is a noncritical application.

If the circuit point is elevated by a small positive DC voltage and capacitively coupled in and out, then an n-ch jfet could be used without the hassle of a negative rail. Voltage is "relative", so this is just like a cathode-biased tube where the cathode is more positive than the grid and the grid is therefore negative with respect to the cathode. Here, the jfet source is more positive than the gate, so the gate looks negative to the cathode. One could DC or AC couple the tremolo signal into the gate. I believe Mr. O'Connor showed something like this in TUT4.

Other options involve rather a lot of sand being added to the valve circuit.

Cheerio

[makinrose]

Good day Makinrose

All good information! I suppose if one wants a "vactrol" with data available and that actually conforms to the data sheet, the only option is Silonex, at least as far as current production goes? If you make your own you have the option to use a real light bulb with a filament

Some of the alternative interfaces get a bit complex or expensive. The p-ch jfet option is running out of jfets to choose from, at least those with leads. I imagine one could use a more common n-ch jfet if there were an interface between the tube oscillator and the jfet gate, but this would also require a small negative voltage supply. The latter is easy enough to do using any of the less-than-ideal bias circuits since this is a noncritical application.

If the circuit point is elevated by a small positive DC voltage and capacitively coupled in and out, then an n-ch  jfet could be used without the hassle of a negative rail. Voltage is "relative", so this is just like a cathode-biased tube where the cathode is more positive than the grid and the grid is therefore negative with respect to the cathode. Here, the jfet source is more positive than the gate, so the gate looks negative to the cathode. One could DC or AC couple the tremolo signal into the gate. I believe Mr. O'Connor showed something like this in TUT4.

Other options involve rather a lot of sand being added to the valve circuit.

Cheerio

So far only the Silonex ones seem to have the right specs.  I'm getting some parts in for DIY LDRs so hopefully I'll have an easy replacement if someone else wants to try this circuit. I'm sure a JFet is good idea but they are gettingn hard to source and the one from China are kind of iffy.  With this circuit  I like it because it sounds good, uses one triode, and has no ticking issues.   I've also seen a Mosfet driven version of it....