The Butte Headphone Amplifier by Apex Hi-Fi

I'm very pleased to have Todd the Vinyl Junkie nearby. Thursday nights he has listening sessions of newly arrived vinyl and serves imported Belgian beer in proper glasses ... so that's cool. But in this case, I'm pleased because I get first dibs listening to Pete Millett's newest creation for Apex Hi-Fi and TTVJ: The Butte headphone amplifier.

The Apex Hi-Fi Butte ($495)
The Butte holds the affordable slot in the Apex Hi-Fi line-up of very fine Pete Millett designed headphone amps, which also includes: the all-tube Pinnacle ($10,000); the tube/FET hybrid Peak ($1,395) and Volcano Power Supply ($750); and the solid-state Arête ($1095).

The Butte is a model of simplicity. The front panel includes: a power button; a blue LED power indicator; 1/4" headphone jack; input selector; and volume control. The rear panel has: RCA input; 1/8" miniplug input; and 12VAC power input. The amp comes with a CUI 12VAC wall-wart.

Circuit Description
This is a very simple and apparently nicely designed, DC coupled, class-AB headphone amp.

Inside the amp is a two-layer circuit board, nicely laid out and populated with parts appropriate for an amp in this price range. The 12VAC from the wall-wart is rectified and energy stored at about +/- 19VDC (unloaded) in the unregulated section of the internal power supply in 8 (4 per rail) 68uF Nichicon CL-series aluminum electrolytic caps. These are good, low-impedance capacitors.

Rough power is regulated by National Semiconductor LM2940 and LM2990 for 12VDC positive and negative supplies to the audio section. The regulated supplies have 10uF storage caps at the regulator, and have local decoupling with parallel electrolytic and unidentified chip caps at the supply pins of the audio op-amps.

The RCA and 1/8" mini jacks for audio input look to be of good quality. The circuit traces from the audio inputs are nicely routed to the front panel selector switch with each trace separated by a shielding ground trace. Signals are then attenuated by the volume control before being sent off the non-inverting pin of the op-amps. I would have liked to see a better pot used in this amp, but knowing Pete, he probably had pretty good reasons to make the parts compromises he did. Most of the parts used in the amp otherwise seemed to me very good choices for this design and price.

All gain and power for each channel is provided by a single Burr-Brown (TI) OPA552 high-voltage, high-current operational amplifier integrated circuit chip. This is a very nicely spec'd chip delivering 24V/uSec slew rate, and 14nV/rootHz noise. Resistors show the amps set to a gain of 11 ... if I got my math right.

Amplified audio outputs run to a nice, high contact-pressure headphone jack. Also in this area of the board are two Claire CPC3703 N-channel FETs, which are used as constant current sinks to force the OPA552s into class-A operation.

Four-layer boards makes for really good ground planes and shielding. On two layer boards, a variety of techniques are needed to optimize for low crosstalk and noise and get around the geometric limitations of two layers. Pete has obviously spent a good bit of time on board lay-out for audio gear, and on the Butte it shows. Care has been taken to ensure properly separated current paths on the ground plane, and there's one corner of ground plane near the input selector where the ground plane would not have been optimal but for a well placed jumper. Good stuff.

So, how'd it sound?

Apex HI-FI
Todd the Vinyl Junkie
405 2nd Ave E
Three Forks, MT 59752

jay's picture


How do you force the opamp into class A? The data sheet shows the surface mount version he is using in the video, after soldering the tab and and pin 3 for heat sinking, only has one pin left, the flag pin used to indicate shutdown. Does the current sink attach to this?


excellent review.

Tyll Hertsens's picture
It's a little complicated but I'll give it a quick shot.

Class AB op-amps are configured such that the positive side of the amplifier amplifies the positive half of the signal, and the negative side of the amp the negative half of the signal. There is an area around zero where both halves, or neither half, is doing the amplification depending on how the amp is biased.

Using a constant current source/sink to draw current continuously out of an op-amp forces this "cross-over" region to become wider such that both sides of the amp are running for more of the time, or operating in class-A.

The current sink attaches directly to the output of the op-amp.

There is a full description here:

deckeda's picture

Hello Tyll,

The URL in the link (Company Info section) is buggered. It currently resolves to "http://https//"

Nice review.

Tyll Hertsens's picture
Thanks! Should be fixed now.