The schematic shown here is a small active reverse RIAA circuit with galvanic isolation obtained by means of a small LT700 transformer, originally designed for use in transistor radios and still easily available from many websites. The circuit is a true pickup simulator that allows you to connect a function generator or another audio source directly to a phono input and, thanks to its very low output impedance, it can be connected to both MM and MC inputs.
I ended up building this circuit as an “accessory” for my ISDS2062B analyzer. In fact it is quite difficult to interface the DDS and the probe of the ISDS itself to a phono input, because the DDS and the probe operate at the minimum limit of the amplitude that can be generated and measured, and the background noise leads to inaccurate measurements. Sometimes ground loops also contribute to the problem. I therefore built a circuit that allows the DDS signal to be injected at a sufficiently high amplitude to make the noise acceptable and the signal itself easily measurable by the probe. During the sweep measurement the oscilloscope channel 1 must be connected directly to the DDS signal. The LT700 is used in step-down configuration, so the signal on the secondary reaches levels compatible with a phono input. The galvanic isolation provided by the transformer and the 47nF capacitor connected in parallel with the output ensure sufficiently low noise levels (about 1mVPP) to allow reliable measurements, even if the source (ISDS powered from the USB port) is quite noisy, with white noise around 10mVPP.
The schematic (click to enlarge)
The circuit input is formed by the network of resistors and capacitors that implements the inverse RIAA equalization. The signal from the function generator and the first probe of the ISDS analyzer are connected to this input. The second stage is a source follower built with a BF256 JFET, whose task is to lower the output impedance of the inverse RIAA network and drive the input of the following LM386 operational stage, which in turn drives the primary of the LT700 transformer. The network formed by R9 and R10 is a local negative feedback needed to flatten the frequency response of the LT700, which is already very linear at such low signal levels up to almost 400kHz. For obvious reasons I could not take the NFB signal from the secondary, because I would have lost the galvanic isolation required to keep the circuit ground separated from the phono preamplifier under test. In any case the output stage remains linear down to the required 20Hz.
The whole circuit is powered by a 9V battery to avoid ground loop problems and various kinds of noise. I built the circuit on a small piece of prototyping board, then I printed a small enclosure with a 3D printer and boxed it with RCA connectors, a switch and a power LED.
Below you can see the bandwidth graph showing the reverse RIAA curve of the instrument and the spectrum analysis with a THD lower than 0.05%.
