a few updates to my personal S.E.X. amp

[Doc B.]

We're getting gear ready for the Seattle HeadFi meet this Saturday and today I decided that after six years of continuous use it was probably time to spruce up my personal S.E.X. amp with MQ output iron.

We started by eliminating the 120 ohm resistors at the headphone jack since we had heard reports that folks with low impedance phones preferred the sound that way and we intend to show the S.E.X. amp as the best choice for low impedance cans. Not sure we heard much difference with the inexpensive low impedance phones we had on hand, but we left them out.

Then I though about trying a bypass cap on the 1.27K resistor - which I think I left out of the original design partly as a cost cutter and partly because no bypass cap usually sounds a wee bit cleaner. The cap didn't do much, if anything it was worse. Then we tried replacing the 1.27K with an LM431 like we use in the Paramount. In this application the output triode has a lot more gain than a 300B and unfortunately the 431 generated a bit of hiss that we deemed unacceptable, even though it seemed to clean up the bass a bit. Finally we tried an HLMP6000 LED with a 499 ohm resistor under it in place of the 1.27K. This was quiet and seemed to help the bass a bit.

Next, I happened to have a couple of 2200 uF Black Gates on hand and those went in in place of the Xicon 1000uF bypass cap on the cathode resistor of the output triode. Quite a nice improvement, cleaning things up in the midrange and making the background a bit more black.

Speaking of quiet we noticed that one channel had some 120Hz, so Shawn replaced the original filter caps. The S.E.X. amp has been running in my HT for about 5 years now, and the caps probably had at least 8000-10000 hours on them, so this is not out of the ordinary. New caps got the noise floor right back in spec.

Next we replaced the Clarity Cap 0.1uF interstage cap with a .082uF Vitamin Q I had lying around. Surprisingly, this helped the bass, taking a bit of what I would call fizziness out.
We then tried a 1 uF Auricap that I happened to have around as a replacement for the existing 2 uF Clarity Cap. This was a step backwards, creating a bit of a compressed midrange. So the Clarity went back in.

All in all this was a nice sprucing up of the S.E.X. amp, and it really makes for an interesting comparison with a Speedballed Crack. I would say the Crack exhibits the classic OTL plusses of super quick and natural sound and it sounds really tight and punchy with high impedance cans, while the S.E.X. has more of the classic SE qualities of liquid mids and highs, more "3Dness" (better image width and sense of stage depth) and great bass performance with low impedance headphones.

[Dr. Toobz]

Finally we tried an HLMP6000 LED with a 499 ohm resistor under it in place of the 1.27K. This was quiet and seemed to help the bass a bit.

I recently tried something similar to this - two HLMP6000's in series under the cathodes. I would think the result would be about the same as the LED and resistor in series, and my observations concur with yours: better bass and cleaner sound (though not a huge difference).

All in all this was a nice sprucing up of the S.E.X. amp, and it really makes for an interesting comparison with a Speedballed Crack. I would say the Crack exhibits the classic OTL plusses of super quick and natural sound and it sounds really tight and punchy with high impedance cans, while the S.E.X. has more of the classic SE qualities of liquid mids and highs, more "3Dness" (better image width and sense of stage depth) and great bass performance with low impedance headphones.

Couldn't agree more, esp. since I have the S.E.X. and Crack amps in the same room and have compared them side-by-side. I'm running mine with the aforementioned LED mod, just a regular resistor load on the driver plate (I like the extra distortion, it would seem), the MQ iron upgrade, and .1uf/2.2uF Clarity Caps. I do think you should try out CFB with gapped irons sometime - I think it makes the S.E.X. sound even more nuanced and helps unveil the deepest parts of the soundstage. I've not made this a permanent mod, though.

You have some great products!

[Doc B.]

Thanks Angelo,

If I may be so bold, the cathode feedback for S.E.X. idea was originally mine, used on the original S.E.X. amp back in the 90's. In that app it had an interesting spin in that we repurposed the original air gapped output transformer as the plate choke after PJ had added a Speco OT via parafeed. I took the unused secondary of that repurposed transformer and employed it in the cathode of the paralleled output tubes of the original S.E.X. circuit. It got us a lot more power, but I can't recall that it sounded any more refined.

[Dr. Toobz]

Interesting! I wasn't into tubes (or even audio) in the times of the original S.E.X. amp (I won't tell you how young I was, either), but I've read lots of stories about it. In this case, I definitely got less power with the feedback under the cathodes - probably 3dB less gain. But somehow the sound was more balanced or "delicate" (drier treble and less prominent mids, maybe?). I think feedback must be a taste thing, just like capacitors, SET vs. P-P, horns vs. direct radiators, etc. But that's also the fun of this hobby, I'm finding!

[Jim R.]

Doc,

Sounds cool, but I'm a little confused by this:

Next, I happened to have a couple of 2200 uF Black Gates on hand and those went in
in place of the Xicon 1000uF bypass cap on the cathode resistor of the output triode.
Quite a nice improvement, cleaning things up in the midrange and making the background
a bit more black.

I thought the output triode was a grounded cathode and that the 1000 uF caps were grid bias caps?

I'm just trying to get things sorted out because I originally thought they were cathode bypass caps (see my thread on page 3) but I was told there were no cathode bypass caps in this amp.

Also, you don't mention it, but I assume you don't have a CCS in this amp either?

Thanks,

Jim

[Doc B.]

Hi Jim,

Yes I guess that was a loose use of terms. The circuit is an old Western Electric design as PJ described in the thread you mention. The 1000uF cap goes from ground to the negative rail. And yes the driver stage has a C4S on it.

[Jim R.]

Dan,

Thanks, that clears things up.

-- Jim

[Doc B.]

I definitely got less power with the feedback under the cathodes - probably 3dB less gain.

3dB less gain does not necessarily imply less maximum power. These are two different things. Negative feedback is reducing the gain of the circuit, but at the same time it can allow the circuit to be driven with a higher level input signal and produce less distortion at a given power output. In other words, you might have to turn up the volume knob to a higher setting to get the same output level as before, but you might also be able to turn it up even louder than before and still have clean sound at a listening level where you were hearing distortion before the feedback was added.