The EAR 864 preamplifier, designed by Tim de Paravicini, is one of the most frequently cited references in the world of high-end tube audio. It is a refined and highly regarded piece of equipment, especially for its top-level phono stage and for its transformer-coupled balanced output, a non-trivial design choice that requires great attention to detail. The result is a preamplifier capable of combining the typical tube musicality with surprising dynamics, achieving a balance that still makes it a sought-after device today and, at the same time, anything but trivial to work on if someone has already tampered with it.

The absurd diagnosis of the “alternative” technician

The client who entrusted me with this EAR 864 told me an amusing story: before bringing it to me, he had taken it to another technician, who diagnosed the fault as a design error. According to him, the output transformer had “burned” because direct current was flowing through it. Yes, according to this self-proclaimed electronics expert, a transformer should never carry direct current, otherwise it gets damaged. Too bad he completely ignores what a single ended circuit is, where by definition direct current is an integral part of the transformer’s operation.

A repair that requires ingenuity and patience

Once I received the preamplifier, I immediately noticed clear signs of tampering: altered wiring, oversized capacitors forced into spaces not meant to contain them, shielded cables where they were not needed. But the real problem was one of the output transformers, with the primary winding completely open.

I removed both transformers for comparison: one working, the other with an open primary. And here began the most complex reverse engineering job I have ever faced: 10 separate windings, 20 leads to identify, windings jumping from one core window to another, and complex routing paths to follow. Hair-thin wire, single ended topology, not push-pull, and a truly convoluted layout.

Colored pencils and reverse engineering

To find my way through it, I had to go to a tobacconist and buy children’s colored pencils. One color per lead, every path marked on scrap paper.

Then, with CAD tools at hand, I redrew the entire transformer map to understand how to rebuild it. In the end, I managed to rewind two perfect clones. I also measured them and compared them with the original unit that was still intact, confirming the success of the work beyond any doubt. The fault? No sign of burning at all, just a point where the extremely thin primary wire had broken. Likely a mechanical failure or vibration, nothing more.

Electronic overhaul and sensible improvements

I then removed the main PCB of the preamplifier. Many electrolytic capacitors showed high ESR values and were replaced. I also found two 1000µF electrolytic capacitors on the cathodes of the output ECC82 tubes. While I understand the intention of achieving strong cathode bypassing, a capacitance of this magnitude is clearly oversized in this specific context. From a functional standpoint it provides no real benefit compared to much lower values and instead introduces an undesirable side effect: at power-on, the tube is forced to deliver a high inrush current to charge these capacitors, a current that also flows through the output transformer, whose primary is wound with extremely thin wire. In a single ended circuit with transformer-coupled output, where direct current is an integral part of normal operation, it is always preferable to limit any unnecessary stress, especially during transient conditions. For this reason I replaced the two capacitors with 100µF units, a value more than sufficient to ensure proper bypassing without excess. The new capacitors were also bypassed with small NOS polypropylene capacitors rated at 160V, chosen to improve high-frequency behavior while keeping dimensions compatible with the available space.

The randomly installed 600V “audiophile” capacitors were far too large and had lifted the PCB traces. They were removed and replaced with 250V NOS capacitors, more than sufficient for the 180V present in the circuit, with superior measured quality and minimal physical size.

After replacing some overheated resistors, I carefully reconnected the new transformers, tested all the supplied tubes with my uTracer 3+, and replaced the three ECC83 tubes in the phono stage that were weak and unbalanced.

After final tests and measurements, the EAR 864 preamplifier was back to perfect operation. A meticulous job, full of obstacles, that brought back to life one of the finest tube preamplifiers ever built. And above all, it prevented it from ending up permanently in the wrong hands.

Instrumental verification and performance

Once the work was completed, I carried out the usual instrumental measurements to verify signal quality and proper operation of the preamplifier. The total harmonic distortion (THD) spectrum, measured at 10Vpp output on a balanced load, shows a low value of 0.18%, with a clean and regular harmonic distribution, indicating linear and well-controlled operation. The frequency response extends from 10Hz to 25kHz with only -1dB attenuation at the extremes, confirming wide bandwidth and good linearity.

Square waves, tested at the classic 100Hz, 1kHz and 10kHz, are well defined, free of overshoot, with rise and fall times consistent with the expected behavior of a transformer-coupled output.