This page is a work in progress. I will try to answer questions about MagicQuartz that many users have asked …

What types of motors/turntables are supported?

MagicQuartz generates a single phase, true sine wave high voltage AC signal. This allows you to connect the following motor types:

  • Single-phase synchronous or asynchronous AC motors
  • Two-phase synchronous or asynchronous AC motors, when using a phase-shifting capacitor for the second coil. Turntables, which use this type of motor, often have a 50/60Hz switch (for example below the platter). With this type of motor, you will possibly be limited in terms of the usable frequency range (any frequency that is away from the nominal frequency, i.e. 50 or 60 Hz). It may happen that the motor hums, but it depends on the motor. Some motors only use the second phase to power a „small“ helper coil, which forces the motor to start rotating in the correct direction. In any case it’s OK if you use the „live quartz mode“ to fine-tune the speed, as this should only cause frequency changes in the +/- 2 Hz range, but you should be careful when using the speed box function.

The voltage can be configured between 0-240V, however, at lower voltage settings, you would sacrifice dynamic resolution. For 110V motors, a different step-up transformer is recommended (12V:110V instead of the stock 12V:230V one).

Low-voltage AC motors (i.e. around 12-24V) can be powered with a step-down transformer between the “stock” MagicQuartz kit and the motor. However, because it makes no sense to first step-up and then to step-down the voltage again, it should be possible to directly use the generated 24V PWM AC signal. This would involve some additional low-pass filtering which I have to still work on.

MagicQuartz should only be used with IEC protection class II appliances. These are turntables that have a two-pole power plug. Do not connect IEC protection class 1 or 01 appliances to the circuit. These are turntables that have a separate ground wire in their cable and a three-pole power plug. In the unlikely event of a failure inside the turntable, the turntable’s chassis could draw a high voltage, because there is no ground connection to main’s earth (if you use a normal „wall wart“ type power supply as in the original design).

How much electric power can MagicQuartz deliver?

MagicQuartz can drive turntables up to around 20 Watt. I will have to come up with more precise measurements, however, it is capable of driving a Dual 1219 and Dual 1019 in parallel, which are rated 17 Watt together.

How does MagicQuartz compare to “Product X”?

Many people have asked me how MagicQuartz compares to … let me call it “Product X”. If you’re reading this, you should probably know about which unit I’m talking.

When I started developing MagicQuartz in 2010, “Product X” did not exist (at least to my knowledge) and I was not aware of it until recently. So MagicQuartz has been developed totally independently from any other similar product, and it’s targeted to a different user group, compared to “Product X”.

I can’t compare them objectively, because I never had the chance to use one and their technical documentation is rather limited (compared to the MagicQuartz documentation).

“Product X” appears to be expensive and they are targeting it to the high-end market. But I think their price is justified, because they’re selling a finished product, which also looks very nice. In contrast, MagicQuartz is rather targeted to DIYers with upper-class consumer turntables from the 60ies/70ies, that’s why it was constructed it with a low price target in mind. Many parts are of-the-shelf parts. I even encourage you to 3D-print some of the case parts on your own, if you can.

The electronic design / topology also seems to be more advanced, efficient and economic in “Product X”, because it seems to be switching an intermediate high voltage (just guessing from looking at their website images), whereas MagicQuartz is directly up-transforming the generated signal in a huge and “expensive” transformer. A topology as in “Product X”, however, would be much more difficult and dangerous to be assembled in a DIY kit.

From my understanding, the firmware of MagicQuarz is much more advanced. You can basically configure anything in nice, cursor-controlled menus. While this might not be necessary in most cases, it’s always good to have this option.

My documentation also explains almost all details of how the firmware is working. The makers of “Product X” do not, and this leaves you with a “black box”, where no one really knows what it’s actually doing. Not exactly what audio freaks and DIY guys want, but it could be appealing for those who are simply looking for an easy-to-setup speed box.

Also read the other answers below for a deeper insight.

How good is the temporal timing resolution?

MagicQuartz computes with a resolution of 0.0001 Hz/RPM – and it uses some tricky timer programming so that it can actually reproduce such fine frequency steps – however, the typically achieved long-term accuracy in live quartz mode is about -0.0002/+0.0005 RPM (ring-buffered and averaged “long-term” RPM readout).

Typically, MagicQuartz will show something like 33.3335 rpm as average speed. This depends a lot on how well the optical sensor is set up, how the “quartz latency” parameter is configured and the turntable (model, overall mechanical condition, e.g., the lubrication warm-up). From the averaging you can really tell that MagicQuartz is capable of keeping the platter at an average constant speed. This is also possible thanks to the way MagicQuartz is compensating measurement errors when calculating the supply frequency.

I’m planning to come up with some real measurements, where I transfer measurement values to Excel spreadsheets for evaluation.

How good is the (one-shot) speed measurement accuracy of MagicQuartz?

It’s difficult to assess the accuracy without precisely knowing the speed error that is induced by the turntable. There are also other factors, which have an impact (ambient light flicker or noise on the optical sensor signal).

I’ve measured (one-shot, unbuffered) speed measurement variations of around 0.006 – 0.01 RPM on my Dual 1219. On my Dual 701 (direct drive, but not quartz-controlled) I got results well below 0.005 RPM. So the Dual 701 is much more stable than the 1219, although the 701’s speed tends to drift over time when the unit is warming up. On high-end turntables with really heavy platters, the result could be totally different, much better.

Also note that MagicQuatz’s speed correction algorithm was designed to efficiently deal with such speed measurement fluctuations. Even with low-quality speed measurements, it can achieve a super-accurate speed. Low-quality measurements do not directly translate into “high” supply frequency speed changes (this can be configured with the “quartz latency” configuration parameter).

I will try to do measurements with direct-drive quartz-controlled turntables, but the results could be “interesting”, if not shocking. It’s very difficult to find a good reference for accurate measurements.

What’s the difference between DSS and PWM? How’s that implemented in MagicQuartz?

DDS and PWM are part of the same technology. It’s a standard method of generating an analog-like signal (here, a sine wave) with a digital system.

There are some other advanced technologies like noise shaping / spread spectrum, which aim to reduce harmonics (relevant for electromagnetic compliance). These are not necessary in MagicQuartz’ use-case, as long as the DDS/PWM is implemented correctly.

MagicQuartz generates a 9-bit sine wave with 128 steps per full sine wave (that’s the DDS part), which is more than “enough”. The sampling rate is 6400 Hz (for a 50 Hz sine wave), depending on the generated frequency. Again, this is more than enough for a < 70 Hz signal (do the math and remember the Nyquist-Shannon sampling theorem). The PWM part is the switching to reproduce any “analog” value of the sine wave. This happens above 31 kHz, which is beyond the audible range.