Two headphone amplifiers sharing the same basic MOSFET source follower output stage.

When the source current and source resistance are optimized for the given headphone load and similar maximum output power (~50 mW at 1% THD), the distortion pattern vs. output power is remarkably similar.

One plot below is simulation, the other measurements. The J-Mo 2 simulation closely matched the actual measurements, it wasn't worth my while to generate a full simulated data set when I already had the measurements on hand. No reason to suspect that the Szekeres sim is inaccurate, either.

The take home message is the distortion characteristic of a MOSFET follower is what it is, and unavoidable. Take it or leave it, as it were. However - and this is key - if you don't optimize the stage for the headphone impedance, the distortion for a given output power will increase significantly.

As an aside: Greg did his homework with the original circuit....

I've always enjoyed the sound of the Greg Szekeres' Headphone Driver (buffer) and derivatives sharing the MOSFET source follower output stage.

I've often wondered however, whether it's distinctive sound is because it is unusually free from noise and artifacts, or because its unusually prone to heavy second harmonic distortion.

It's not hard to set this up in LTSpice, but I haven't seen it done before. So, for your education and enlightenment, I present the harmonic distortion vs. output power data for the original "classic" circuit as uploaded to Headwize all those years ago. The LTSpice asc file is also included I you want to play along. The harmonic data is generated by hand, reading the FFT peaks for 10 or so different input voltages.

The Amp boards arrived Saturday, unfortunately, I was just leaving for work so they'd have to wait. Everything looked OK on them.

Initial power up test was done with both LME49990 in place but only one LME49600 mounted on each channel [PSU used was a modified "The Wire" unit meant for testing the A/B Power Amp at +/-25Vdc, now set at +/-15Vdc. It was getting quite hot as it was dropping 10V per rail]. That went Ok, so the rest of the 49600 were put in place and powered up once more... Again, everything OK.

Measured offset:

Left channel: 27mV
Right channet: 8mV

Foam board speakers are progressing, and will be ordering a heatsink this week also.

Update [13-04-30]: Been a little bit worried about the higher offset value on the left channel, though it's not at a bad level overall. Once the amp has been on awhile, it appears to settle at 25.9 mV.

Made a quick SE input>volume pot AC coupled...

I'm a recent convert of the lowest possible impedance of power supply based on my experience of adding caps to my chipamp. So I figure the signal stages can't be harmed by reducing their supply impedance either, particularly at LF.

I note there are a few aftermarket regulators around - I had a look at Paul Hynes and Belleson in the past few days. They're a bit pricey for my tastes, given the cost of the components they're using can't be over single digit $ so I've had a look at "doing it at home, only cheaper".

First off, a simple TL431 is about the best bang for the buck achievable, as the part here is 0.2rmb. But the dynamic impedance is typically 0.2ohms and I was hoping and aiming to go a bit lower than this - perhaps an order of magnitude lower, to around 10mohms. Lower than this and the resistance of the PCB tracks come into play and its also very hard to maintain such a low impedance beyond the audio band as cap ESRs (for the best ones) are of...

Here's the first channel under construction on 2mm pitch double-sided proto board. The opamps in this instance are AD8014 in SOT23-5 packages, hence the choice of the 2mm pitch board. Power supplies will go on another board to be attached to the reverse side.

Changed the collector load on the voltage amplifier stage to a current source (Q3), as per the Marantz SR2285B circuit.

Also increased the resistance of the feedback connection, R6+R8, there seemed to be no obvious advantage in making it much smaller than the typical load (>10k). The compensation capacitor C2 is increased to match, to flatten the HF response.

The circuit can drive light loads to +20 dB. Of course that's not much of a challenge for this general class of circuit.

I'm a bit stumped as to what the next logical step is from here. Seems to me to depend on what you actually want the circuit to do.

Holiday: Bad weather to a point it was near impossible to take a steady photograph due to the gusting winds. Hiding from the winds meant poor vantage points, very much non-idealistic for photographing landscapes.

Digi-key: The order turned up on the Monday morning, roughly 10 minutes before leaving. The PSU boards must have arrived mid week, a neighbour had those...

The parts order was enough components for one power supply and one amp. The PSU is assembled but untested as yet. The toroidal and 10K Alps RK27 are to be ordered from RS Components later today.

PCB issues: C1-C6... this is where trusting the people who generate default libraries for FreePCB falls down and my inexperience shows. Lead diameter on the caps is 28MIL, default part through-hole diameter is 28MIL... Each lead needed a quick file to reduce its width enough to fit the board. No other issues to report.

I don't think I could have made the overall size of the board...

Where was I? Right, back to the workshop and cue the music.

Sorry in advance for my photos. They get the job done, but some of them look pretty bad now. I'd retake them but the amp has long since been returned to it's owner.

Once both channels were running happily with their new parts on the benchtop power supply I turned my limited attention to the chassis. The phenolic insulators sandwiching the RCA jacks in place were starting to crumble so, some shiny new jacks were fitted courtesy of ApexJr.com for 99 cents. The input wiring was done with 22AWG shielded wire.



The pro Hafler models like the P125 and the P230 have a solid aluminum bracket for the ground connections between the main power supply caps. The DH200 and DH220 models only got some tinned wire which isn't very convenient or nice looking. I fashioned up a new bracket with some scrap...

All members who want to use LiFePo4 cells from the leading company A123 Systems
and want a secure source to get no fakes,
could order them at Lipopower Ltd. in Germany.

The original cells are from the German A123 Systems distributors Reva or BMZ.

You could order them with or without solder taps.




Also a Ri selection is possible.

International shipping and paypal payment is no problem!

...

Over the last couple of years most of my interest in audio has been with transistors. I've been slowly teaching myself to read and understand the circuits.

Circuits like this one for example. Not hard, but still a bit too complicated for me to understand without the helpful wikipedia markup attached.

Instead I've looked at primarily at the schematics I have for discrete audio preamplfiers, 1970's vintage typically. Based on what I've learnt so far, I've done up a "test mule" in LTSPICE, shown below.

It's not a circuit you should build. It's for pedagogical purposes, though it does actually work reasonably well - in simulation anyway. Its just a simple starting point to observe how the different parts interact under simulation.