Szekeres vs. PRR-SOL amplifier

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello, I did try to search the forum for SOL amp, but with no avail. I sort of remember this project from the past of heydays of Headwize. Can someone tell me what is the difference, from design perspective, between Szekeres and SOL? The first thing what stands up is PSU voltage and signal output place. Are they both buffers?
Image:pRR-SOL.png - Wikiphonia
Class A MOSFET Headphone Driver | HeadWize

Thanks
 
Last edited:
No, the SOL as pictured is a common source amplifier with a voltage gain of about 20 (26 dB) and an output impedance of about 120 ohms. It was presumably intended to drive 600 ohm headphones, like the rather insensitive K240M/DF.

Not sure how well something like this works; I imagine that with such a relatively large FET and Miller effect in play, high-frequency linearity may not be that great.
 
Thank you very much for the response. I will try to build it. Still, it should perform better than Szekeres, isn't it?
That depends on what you need it to do. A source follower buffer like the Szekeres usually features lower distortion, lower output impedance and higher input impedance (which equates to lower input capacitance in this case) - but obviously does not provide any voltage gain. It is better-suited for low-impedance headphones and/or sources that are plenty loud (e.g. a CD player).

Here's what you can expect from a Szekeres with a 30 ohm load.
Even the version without a current source shouldn't produce any audible distortion up to about the 10 mW level or so, which is ample for most any regular headphone though I wouldn't attempt to drive a HE-6 with it.

Obviously it's very inefficient - 200 mA of idle current (2.4 W of total power dissipation per channel all the time) to output a decent 20 mA. That's the price you pay for the very low circuit complexity. (I bet a measly cMoy with a JRC 4556 could match that. Looks like it does, actually. At a typical 9 mA idle - for both channels. Obviously there's no less than 16 active semiconductor devices per channel tucked away in the lil' chip. Component count for the DIYer is quite similarly low though.)

You can see how distortion drops by almost an order of magnitude upon adding a simple current source (for a whopping two semiconductors in total). Then there should be no audible distortion up to the point of clipping at ~50 mW.

Incidentally, the active ground version with a 3rd channel can be assumed to be performing worse than the basic capacitor-coupled one. At the very least, the ground ought to be able to sink at least as much current as both channels can source (so a 3rd identical channel is underpowered), plus output impedance is very critical here since it may deteriorate channel separation considerably. See NwAvGuy's measurements of the Mini3 for what can happen in constructions like that.

EDIT: I've now simulated the SOL, and you can get about 10 Vpp into 600 ohms out of it with acceptable distortion, at a gain of an effective 22 dB and worst-case high-frequency -1 dB point at ~19 kHz. That 120 ohm resistor has to be dissipating a whopping 6 watts though! :eek:
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.