This is a listening comparison of the Through-Zero Linear Frequency Modulation sounds made by four different TZ capable oscillators.

• Dannysound EN129 Thru-Zero Oscillator
• Patching Panda Operat Analog Complex Oscillator
• Weston Precision Audio TZ0 Thru-Zero VCO
• Joranalogue Audio Design Generate 3 Through-Zero Multiphonic Signal Generator

Through Zero Linear FM Initial Offset

Through zero linear FM entails voltage control over the current injected into the VCO core. For practical purposes you need to think about two parameters: the initial current setting and the depth of the audio frequency modulation signal that sums with this initial setting.

The initial setting affects the oscillator’s frequency, along with the exponential frequency modulation settings. So there are two initial frequency settings to deal with. Many combinations of the two initial knobs can result in the same initial frequency. My approach is to set the initial linear offset (bias) and then use the exponential tuning knobs to dial in a specific frequency.

Modulation Depth

Modulation depth sets the range of the main oscillator’s (the carrier) frequencies that the modulating signal will cover. The depth is the ratio of the modulating amplitude to the amount of initial offset. Of course the modulating amplitude is typically controlled by a panel knob that attenuates the incoming TZ modulation signal from another oscillator. The initial offset typically has its own knob, which may be unipolar or bipolar. The bigger the modulating amplitude is in proportion to the initial offset will determine the depth, i.e. the “drama”, of the sound. Thus, keeping the modulating amplitude fixed while increasing the initial offset will reduce the depth. The greatest depth will happen with the lowest initial offset, which can be so close to zero. Exactly at zero the oscillator will stop.

Coupling the modulation signal – AC or DC

The modulation signal may be AC or DC-coupled. Each of our oscillators does this differently. The EN129 and the TZ0 both use AC coupling, which sets a lower limit on the modulating frequency. The Operat is DC-coupled and the Generate 3 has a switch to enable AC-coupling or DC-coupling. The advantage of AC-coupling is to remove external offsets, ending up with a nice bipolar signal. DC-coupling can be useful for slow linear modulation frequencies, such as linear vibrato.

Comparison of the initial offset and modulation controls on the four oscillators

The EN129 has a unipolar offset knob labelled THROUGH ZERO OFFSET. The modulation input and CV attenuator are labelled THROUGH ZERO LIN FM. The offset knob adds zero to +6 volts into the oscillator core, summed with the attenuated voltage on the LIN FM input.

The Weston TZ0 similarly has a TZ Offset knob, a Lin FM input and a Mod Depth knob. The only difference with the EN129 is that the maximum offset is +5 volts. (The TZ0 has a built-in VCA for control over modulation amount, but this feature isn’t used for this test.)

For offset the Operat has a knob labelled SYMM. It is bipolar, with zero volts appearing around 10 o’clock on the dial. Below 10 o’clock, the waveform reverses, hence the symmetry language. But it’s the bias pot. This is a bit confusing. The TZ-FM|PM jack is the modulation input with the THRU-ZERO knob being the associated unipolar attenuator.

Generate 3 has the most confusing controls, and oddly they are less flexible. The one knob 0FM acts as the initial offset when there’s no input to the 0FM jack, simply because +5 volts is normalled into the jack. The BIAS switch when enabled (down position) adds another +5 volts offset. This comes into use when a modulating signal is patched to 0FM. (The BIAS switch always adds five volts to the core when enabled.) The AC switch introduces AC-coupling when enabled, which also removes the normalled +5 volts. The downside of this arrangement means that when modulating linear FM, since the knob now acts as the attenuverter for the incoming CV, you can only choose to offset by zero or 5 volts with the BIAS switch, but no smaller offset, such as +2 volts can be dialed in. This ends up reducing your options for the depth of modulation (see above).

Test Setup

I’ve arbitrarily chosen the frequencies for the three test recordings. I started by setting the Initial Offset to the maximum on each oscillator. For Generate 3 this means enabling BIAS and also AC (both switches are down).

The second step was to tune all of the oscillators to A4 (440 Hz).

The modulation source was the Dannysound Cali Osc sine wave. This signal is passed through a VCA controlled by an Attack-Decay envelope that triggers repeatedly.

On the recordings, the order is always EN129, Operat, TZ0, Generate 3.

Round 1

Modulating frequency is set to 180 Hz. For each oscillator, the volume is turned up, then the modulation attenuator is swept manually from off to full and back, then the volume is turned down.

Round 2

The oscillators are tuned to C4 262 Hz. Initial offsets remain at full. The modulating oscillator is set to 346 Hz, this time higher than the carrier frequency. The modulation depths are left at full and the oscillators are simply faded in and out in the same order.

Round 3

The oscillators are tuned again to 440 Hz, but the Initial Offsets are dialed to their midpoints. Since Generate 3 can’t do this, it’s not included. The modulating frequency remains at 346 Hz. What you’ll hear is that the modulation depths are deeper and more dramatic.

Round 4

This time the initial offsets were all set to zero or close to zero. EN129 and TZ0 were set to zero, while Operat and Generate 3 were tuned close to zero offset. Generate 3 had AC enabled and BIAS off. The initial frequencies were all set to a middle pitch. The modulation signal was set to a middle pitch. For the recording, each TZ oscillator’s coarse pitch was raised manually to the maximum and then back. With low or zero initial offsets, the modulation depths are maximized. This round shows significant differences among the four oscillators. Generate 3 exhibited a thump at the end of each envelope, whether AC or DC coupled.

Concluding remarks

The EN129 and the TZ0 sound very similar, since their offset strategies are similar. I think the TZ0 needs some re-calibration. The sine wave sounds a little buzzy, and there is some carrier frequency change heard at the deepest TZFM.

Clearly the Patching Panda Operat has the deepest modulation ability, especially when the initial offset is lowered closer to the zero point. It also sounds very bright to my ears. It’s my favorite of the four for Through-zero linear FM.

I wish that Generate 3 had an initial offset knob instead of the BIAS switch.