Update History

Delta-A / TM-4 (revisited): a solid state tremolo in the style of valve tremolos, using a tap tempo LFO. (click on the picture to load larger image) Nicholas Boscorelli’s Project #29 (Tremolo-Matic IV) [1] convincingly emulates a valve-tremolo as known from vintage valve amps. Tremolo is an effect you are not going to hear very often, but when you do, its an ear-catcher. It's luxury to have a dedicated tremolo unit consume precious pedal board estate, and by means of reason you would no doubt resort to a multi-effect device, since those can do everything. But yes, they can do all, but mostly they also sound like all... And when you finally acquired such a thing, you find out that it has other drawbacks. Tremolos can come in a variety of shapes, but this one imparts a tone that feels luxurious and sultry. The optical design is a homage to the beautiful valve amps of times bygone. Delta A in physics is the quantity of an amplitude change.

"Bias tremolo creates a unique harmonic richness, particularly when played with clean tones. This effect can be subtle but adds a captivating texture to your sound. (...) Bias tremolo interacts dynamically with your amplifier’s gain. The effect becomes more pronounced as you increase the amp’s gain, making it a fantastic tool for both clean and overdriven tones. (...) Many vintage amplifiers from the ’60s and ’70s, such as the Fender Vibrolux and Supro Tremolo amps, are known for their distinctive bias tremolo circuits." - Aldi Fajar (allforturntables.com), Unlocking the Secrets of Guitar Amp Bias Tremolo, https://allforturntables.com/2023/09/25/unlocking-the-secrets-of-guitar-amp-bias-tremolo/

So when I saw  such a unit described in Boscorelli's book [1], I decided to make one - just to hear first hand, what all this is about.

Certainly this unit contains no tubes, but it emulates the mechanisms that take place in tube tremolos. It seems they have vanished from modern tube amps, because they were putting strain on the tubes.

"Claims of tube sound from solid state usually need plenty of salt. Not so TM4." (...) The fact that TM4 sounds different from VCA-based tremolos should not surprise the player, since it contains no VCA. Gain modulation by uni-polar squashing is the same mechanism found in certain classic tube amp tremolos."    - Nicholas Boscorelli, The Stomp Box Cookbook, p.135-7

The reader should not be mislead by the fact that there is an OTA at work - it is not used as an OTA. It does not utilize the OTA's voltage controllable gain functionality, so strictly speaking it is not a VCA tremolo. It does use the OTA as distortion unit by shifting its bias point into a region, where the amplifier throttles back - just as it happens in a tube based tremolo. Boscorelli calls this effect gain modulation by uni-polar squashing.

The OTA is biased with a constant current in a way that its gain is about one. You can verify this by turning the depth pot to zero and check with a 'scope.

I used the version with the single CA3080. People have produced clones of the original (now obsoleted) unit that promise to be as good or even better.

There is nothing to improve about that part of the unit, however, I only used the signal part and not the LFO.asdf

Schematic and PCB of Signal board: (click on the picture to load larger image) Here are the schematic and PCB, identical to the original drawings, but for completion, because I have been asked for.

TM4 would certainly work well with the described LFO, but it seemed a bit outdated to use a pot (only) controlled LFO when tap-controlled subsystems are available for cheap. To my ears, tremolos have to be tightly in sync with the music, so tap control offers a new dimension to that. Besides that, it opens up a wealth of possibilities for interconnection.

TM4 employs a clever mechanism in the wiring of the stereo depth pot. By increasing the amplitude of the LFO signal,  progressively more DC voltage is added too, which in combination produces DC bias voltage that is modulated with the LFO signal, which pushes the music signal towards the cut-off region of the OTA. This mechanism is the same in the tube counterpart.

TAPLFO is wired to produce a sinusoidal wave balanced around the center (with some trimming) with appropriate glue-logic, which seamlessly incorporates into TM-4.

Back To Index

---

Configuring TAPFLO

►

An important word ahead. Boscorelli made his work modular. Some assemblies appear identical in other projects throughout his book, such as the PSU, the input amplifier and more. In my renderings I use an adapted PSU using a charge pump, which is described in detail at the pages on SOMA [3]. I suggest you look at those pages first. Since the signal path used is identical to the original document, please refer to that. The LFO part will be explained in the following.

TAPLFO is a bare-bone digital IC that can be configured in a myriad of ways. It is very powerful and you want to study the data sheet closely. What we need here is a pretty straightforward application thereof:

Schematic and PCB of LFO board: (click on the picture to load larger image) Here are the schematic and PCB that incorporates Electric Druid's TAPLFO.  The large IC to the left is TAPLFO.

My design basically adheres to the modular synth design [2], although the given values for the LFO low pass filter were totally unsuitable for the task needed.
With the help of an online calculator for such multi-feedback filters I  quickly found suitable component values that provide a circa 30 Hz steep roll-of that kills all frequencies above the band of interest stone dead.

The 100k and 200k resistors following the depth pot provide different mix ratios into the subsequent OPA. This means that more DC gets mixed in progressively.

The ratio of LFO vs. DC bias is important. I played with the level trim pot on the LFO board until it sounded best. My setting was between 6-7Vpp, which coincides with the targeted value of 7Vpp.

A high efficiency LED can be tacked directly onto Pin 7 (clock out) for a visual representation of the beat. Adjust the series resistance for suitable brightness.
A 10k resistor appeared low enough, although the output can source or sink up to 25mA, as Tom Wiltshire (the designer of TAPLFO) confirmed. The LED is hard wired and not, as in other projects, activated with the bypass foot-switch. This way the throbbing of the LFO is visible even in bypass mode. This enables you to tap in the right frequency before activating the effect, say, during a performance.

Also, as maybe different to other effects, a tremolo effect sounds unusual with one pulse one per beat, it should be for most cases two pulses per beat. With the x2 multiplier (1/8th note) selected, two taps encompassing a quarter beat will result in two LFO pulses per quarter beat. This can be adjusted to taste.

The application note shows an 79L05 negative regulator, that is not finally connected to the zero adjust section of the second OPA. Instead, it uses the negative supply directly for reference. This is fine as long as you have a regulated negative supply - which is not the case if you use a charge pump converter like I did – those produce an approximate voltage only, dependent on load. I had not thought about that beforehand and had to retro-fit a zener diode circuit (5V1 or 5V6); while a regulator may be superior, it is also an overkill in this position.

The zero-offset trim pot is best chosen to be a multi-turn pot.

All pots appear as trim pots, so does the speed pot. This is handy for evaluation, but can be replaced by front panel controls. The speed pot may not be connected to the LFO board at all in case of using a TRS connectivity. In this case it may better be wired directly to that assembly.

For the tap-speed stomp-switch the best choice turned out to be a soft-touch switch, because their low activation force helps to stomp in timings more accurate, without the need to overcome the threshold of a clunky tactile switch (as much as I prefer them otherwise). Getting that right takes a bit of training.

Warning! The preliminary circuit to the right is a raw sketch of circuitry to control a similar external device. I left it there but it is going to need some more work. I suggest to leave it away for a straight build.

Back To Index

---

Controls, Functions And Connectivity

Schematic and PCB of TRS assembly: (click on the picture to load larger image) The actual jack used comes from Amphenol, which has a stand-off of a few millimeters. Some components (not visible here) are located under the jack, but on the top side. This keeps the dimensions small. Towards the front you see the jumpers for rotation sense.

The current box uses a 1/4" TRS jack for an external expression pedal and a 1/4" telephone jack for synchronization.

• TAPLFO accepts an external sync beat. This is a stock (grounded) jack.
• The speed pot attached to the controller and the tap tempo are not tracking each other. Whichever one gets activated, takes priority.
  
• The last tempo value is not saved. Upon power-up, speed defaults to the pot's setting.
  
• The controller uses all linear pots, so a front panel 10kB pot can be used to set the default speed.
  

►

On the PCB the footprint for a trim pot is used. This helped during the evaluation phase. It was subsequently replaced by a front panel control.

• Since it uses the pots in a potentiometric manner rather than as a variable resistance, its center pin can be looped through a further (isolated) TRS jack contact, so that an external device like a passive expression pedal plugged into the TRS jack can intercept the control line and impart its own control voltage onto the controller.
  
• Such a device would need a reliable reference voltage too, so 5.1V to GND, generated by an on-board zener reference is supplied to the TRS jack sub-assembly.
• I added two jumpers that allow for voltage reversal and thus reversal of the motion sense, but many expression pedals have that built into them.
• control of external devices (other TAPLFO units) is possible, reverse control too.

Back To Index

---

Enclosure

Unlike a different Boscorelli design I made, this one does not have many front panel pots, which on their own would demand a big enclosure, but it uses many additional jacks for external control. The TAPLFO sub-assembly with all its entourage is bulky too.

I wanted to keep the design modular for

1. re-purposing the modules and
2. in case of a total design failure of a certain module (like the LFO) to only have to exchange the respective module
3. some necessities have arisen on the course and were added afterwards.
   
4. although this has not evolved into a case of TMK (too many knobs), it has become a case of TMJ (too many jacks), so again, there was no chance to fit this all into a case smaller than a 1590XX using through-hole technology.
   

The advantage of such large cases is that you can use beefy knobs and are enough space to add visual embellishments. A worthy interior demands a worthy outward appearance. Honor where honor is due...

Back To Index

---

Using

The unit is surprisingly quiet for a OTA build, but actually not so surprising if you consider that the OTA runs at a gain of one and any initial working gain is later counteracted by a corresponding attenuation on the output – Boscorelli calls this quasi-companding. Distortion is audible, but that is meant to be so; the ear tolerates large amounts of distortion before objecting. The  tube counterpart is no different.

The unit sounds sultry and luxurious. Cudos again, Nick.

A short demonstration of the unit. A small hair of echo has been added to augment the luxury. The following equipment has been used: Proprietary speaker cabinet with built-in omnidirectional capsule microphone Recording device: PC interface into DAW Volante echo all recordings untreated

Back To Index

---

• Mar. 14, 2026: added signal part schematic
  
• Mar. 13, 2026: first release