DIY Wow & Flutter Generator – Turntable Hack

[alvaro-oliver]

As part of my ongoing experiments comparing different methods to measure wow & flutter in turntables, I’ve already shared some results from static tests — measuring the same deck, running normally. This time, the goal is to deliberately introduce controlled W&F, so I can later contrast audio-based, gyroscope, and FG-signal measurements under identical conditions.
After considering (and discarding) a few crazy ideas, the most practical solution turned out to be modifying a direct-drive (DD) turntable. Here’s why:

• The ability to create constant oscillations across a wide range of frequencies and amplitudes depends heavily on torque. DD decks have more torque than most belt drives.
• DD turntables use a feedback loop: a speed setpoint and a frequency generator (FG) that reports the actual speed. The controller dynamically adjusts the motor to maintain speed.
• Many DD decks also have a pitch fader. If I can inject a signal into this control, the turntable’s own servo loop will do the rest.

---

Choosing the Victim​

I wasn’t about to sacrifice my near-mint SL-1600MK2, so I needed something modern, affordable, high-torque, and with a service manual or schematic available.
In Chile, that narrowed it down to three “Super OEM” Hanpin models:

• Audio-Technica AT-LP120XUSB
• Pioneer PLX-500
• Reloop RP-4000 MK2

The Reloop had the highest torque spec, and I found a good online deal — so it became the patient.

---

Step 1 – Tapping the FG Signal​

From the RP-4000MK2 service manual, page 21, the MOTOR PCB shows the FG signal at CN102, buffered by IC101A. Inductors L109 and L108 filter noise and make perfect solder points for FG + ground.

On the scope at 33⅓ RPM: a perfect 50 Hz square wave, 90 pulses per revolution. This is fast enough to capture W&F up to ~23 Hz — plenty for measurement. I’ve already used this signal for both total speed and W&F analysis.

---

Step 2 – Understanding the Pitch Control​

The MAIN PCB schematic shows VR301, a dual-wiper pitch pot feeding ADC pins 78 and 79 on the MCU. Interestingly, the two wipers are inverted — as you move the fader, one output rises while the other falls.

Measurements confirmed:

Fader	R318 Output	R319 Output
-8%	3.3 V	0 V
0%	0 V	0 V
+8%	0 V	3.3 V​

That means the injected “modulation” needs to be two half-sine positives, 180° out of phase, centered at 0% pitch. Conveniently, these 3.3 V signals can be generated by a microcontroller DAC/PWM or, in my case, a dual-channel AWG.

---

Step 3 – Calibrating the W&F Injection​

The pitch range is ±8%, driven by two 0–2.285 V signals.
From the 2-sigma method:

W&Fpeak% = f/3150 × sin(0.95×90°)​

Solving for the AWG output:

Vpk ≈ 0.41189 × W&Fpeak%​

Example: 0.2% W&F → ~0.0824 V amplitude.

---

Step 4 – First Tests​

With the AWG fixed at 4 Hz and varying amplitude:

Amp (V)	W&F Target (%)	Measured Unw. (%)
0.0000	0.000	0.224
0.0823	0.200	0.243
0.1235	0.300	0.294
0.2059	0.500	0.384
0.2883	0.700	0.442
0.4118	1.000	0.524
0.6178	1.500	0.610

Observations:

• The deck’s baseline unweighted W&F is ~0.22%, so injecting below this has little effect.
• Above ~0.3%, the modulation is clearly visible in the spectrum.
• The response is non-linear, with measured W&F consistently below target — likely due to limited motor torque and a motor controller tuned more “integral” than “proportional” (preferring long-term accuracy over rapid correction).

---

Step 5 – Frequency Sweep​

Using 0.3% target deviation:

Freq (Hz)	Target (%)	Measured (%)
0.2	0.299	0.312
1.0	0.299	0.331
2.0	0.299	0.330
6.0	0.299	0.294
8.0	0.299	0.269
10.0	0.299	0.260
12.0	0.299	0.268

Great response up to ~6 Hz, then a drop. Boosting amplitude for higher frequencies:

Freq (Hz)	Target (%)	Measured (%)
8.0	0.698	0.349
10.0	0.698	0.321
12.0	0.698	0.421
15.0	0.698	0.312
20.0	0.698	0.251

Spectral analysis shows that above ~10 Hz, the main modulation energy shifts downward — e.g., a 20 Hz injection produces strong modulation around 1.4 Hz. Again, likely torque or servo loop limitations.

---

Conclusions​

• It’s absolutely possible to externally control a DD turntable’s motor setpoint to inject predictable, repeatable W&F across much of the weighted measurement band.
• This low-cost hack works well up to ~6 Hz, reasonably up to 10–15 Hz, but higher-frequency modulation reveals the limits of the motor/servo system.
• Despite non-linearities, it’s good enough to serve as a DIY W&F generator for testing and calibrating measurement methods.

Hope you enjoyed reading

 

[Balle Clorin]

You can also generate W&F on a 3000/3150 Hz carrier by a Python script, have you investigated that option.?

 

[Balle Clorin]

The deck’s baseline unweighted W&F is ~0.22%, so injecting below this has little effect.

Is this about the same as a 2 sigma value? If so it is not a good DD player.

Edit: I saw the 0.35 2S value now, quite horrible result for any turntable even low cost ones. Moste beltdrives are better too. People should rather buy vintage things from garage sales .

 

[alvaro-oliver]

Thanks for the feedback @Balle Clorin ! The ultimate goal here is to compare wow & flutter measurements using three different methods — audio analysis, gyroscope, and FG signal — which I’ll be publishing in the coming weeks. That’s something synthetic test tones can’t fully achieve, so I needed to build this tool.

And just to clarify, the choice of the Reloop wasn’t about playback quality — it was selected for its high torque and ease of modification, making it a practical “W&F generator” for this experiment.

Finally, that “horrible” 0.35% you saw is actually induced by the modified turntable — that’s the whole point of the experiment! Hopefully it will all make sense with my next posts.
Stay tuned!

 

[Balle Clorin]

I am look forward to mor of your excellent and fascinating work. I appreciate it very much,
What was the 2S of the Test TT without modified w&f ?

 

[SSS]

Intersting hack of the PLL. But to verify or calibrate a W&F meter there exist software programs and hardware instruments to modulate a 3150 Hz tone.
I still use sometimes my Grundig GA 1000 W&F Meter. Since I have a Thorens TD124-II TT with heavy mass disk there is no electronics to modulate.

 

[Nick_Sukhov]

to verify or calibrate a W&F meter there exist software programs and hardware instruments to modulate a 3150 Hz tone.

here is test signal @ flac file with 3150 Hz and 1%, 0.1% and 0.01% W&F [ #63 here https://www.patreon.com/posts/77-audio-62733337 ]

 

[DVDdoug]

The response is non-linear, with measured W&F consistently below target — likely due to limited motor torque and a motor controller tuned more “integral” than “proportional” (preferring long-term accuracy over rapid correction)...

...Spectral analysis shows that above ~10 Hz, the main modulation energy shifts downward

I would expect that. You've got inertia. I don't remember ever hearing flutter from a turntable, only from tape. And I've never heard wow unless the turntable was broken, usually a worn or otherwise slipping belt or drive wheel and the slowing was the main symptom... The speed variation is secondary if the it's playing too slow.

I used to read the wow & flutter specs but it wasn't a concern for me. (I don't play records anymore. I occasionally digitize one. and my turntable is direct drive so there are no belts or pulleys to worry about.)

 

[EJ3]

I never heard any wow & flutter from my TT or any others that come to mind (although I have heard it from some cassette decks):

Technics SL-M3

Specifications​

Type: fully automatic

Drive method: direct drive

Motor: brushless DC motor

Drive control method: quartz phase locked control

Platter: 325mm, 2.5kg, aluminium die-cast

Pitch control: +-6% range

Speeds: 33 and 45rpm

Wow and flutter: 0.022% WRMS

Rumble: -82dB

Tonearm: dynamically-balanced linear tracking

Effective length: 238mm

Effective mass: 13g (including cartridge)

Cartridge: moving magnet

Replacement stylus: EPS-33ES

Dimensions: 526 x 205 x 425mm

Weight: 15kg