S.E.X. data

[anthony]

Hello everyone,

   I was just wondering (hoping) someone may have some data saved on the S.E.X. amp.   I'm looking for anything either written or graphical (like frequency response charts etc.) outside of (or on top of [no pun intended]) what is listed on the amp's main description page.  I don't really have access to any test equipment myself, otherwise i would try to get these on my own.

    thanks in advance,

-Anthony

[Doc B.]

Off the top of my head the bass frequency response starts to fall off a wee bit below 100 Hz and is -3dB at around 50Hz with the stock Speco output transformers. It goes way deeper with the MQ iron, probably an octave or more. Also off the top of my head, the treble is flat to something like 36kHz with the Speco and pretty close to that with MQ iron.

[Paul Joppa]

I dug out my notes from the prototype measurements, probably late 2004 or early 2005.

At 1v output on the 8 ohm tap (1/8 watt output), -3dB was 27Hz to 22kHz. These are the specs on the product page.

The "10% distortion at full power" in the specs is an estimate by eye looking at the oscilloscope traces; I'm pretty confident it's no more than that - it did not look quite as good as the 2A3 or 300B amps, which run around 3%-5% at full power. All no-feedback SETs overload pretty gracefully, making it a real judgment call where exactly to put the specified power and distortion.

At 2.8v (1 watt, or 3dB below rated output) there was visible clipping distortion below 35Hz, which would correspond to full power at 50Hz and above. This is a limitation of the Speco output transformer.

I don't recall actually measuring the performance with the Magnequest upgrade iron, but I have measured the plate chokes and output transformers, and I recall posting the predicted improvement some time ago. I can't find it tonight though.

[Grainger49]

Anthony,

Just in case you are unfamiliar with these kinds of specs and are doing some comparisons...

When Doc says an octave below that, that means to 1/2 the frequency stated, that is the amp starts a gentle roll off below 50 Hz.  Your speakers probably do the same thing.

Paul mentions the -3dB points, that is an industry standard for power bandwidth.  And the SEX specs are pretty good.

The 10% distortion is a normal spec for tube amps but not for solid state amps.

I see only 3 posts so I'm not sure how much you are aware of these things.

[anthony]

Hey thanks everyone.

   I, as always, really appreciate the help of the community!  I have yet another question about this amplifier.  Why is there no capacitor at the input? Saving costs? could I put a .01u there or would it be unnecessary?

-Anthony

[JC]

With audio gear, it is presumed that you will only be inputing an AC signal; hence, no real need to have a capacitor at the input because the output of the previous piece of gear should already be utilizing some means to prevent there from being any DC coming out of it.

If by some chance you run across a piece of audio gear with DC on its output, chances are something has failed in that piece of gear or it's not really a standard audio output for some reason.

The effect on the S.E.X. amp will be to mess up the bias on the input stage; depending on how much DC is present, I would expect it to be pretty audibly noticeable.

IOW, there is no real reason to have a cap there, and no cap always sounds better!

[Grainger49]

Like JC says, today's audio gear doesn't have DC on the output if it is operating properly. 

DC briefly appears on the output of capacitor coupled equipment when it is turned on.  Either relays or output resistors dissipate this quickly.

I can't remember the last time I saw a piece of gear with an input cap on it.  But it was truly the standard in the 50s and 60s because the manufacturers didn't trust the preceding equipment.

[Len]

Hey thanks everyone.

   I, as always, really appreciate the help of the community!  I have yet another question about this amplifier.  Why is there no capacitor at the input? Saving costs? could I put a .01u there or would it be unnecessary?

-Anthony

I briefly tried a 3900 pf cap at the input to my Paramours, right before the 279K driver grid loading resistor. It formed a 150 Hz hi pass filter, which I used to take a load of my Lowthers.

[Dr. Toobz]

Those S-5 push-pull kits one sees a lot on DIY audio sites actually do have DC-blocking caps on the inputs - one of the mods is actually to remove them for better sound. Otherwise, I can't think of any other time I've seen them in recent amps.

[Paul Birkeland]

If DC appears at the output of whatever you feed the SEX amp, and for some reason you don't want to repair that piece of gear, you can appropriately increase the value of the cathode resistor in the driver stage to compensate for this appearance of grid voltage.  Of course, this would mean that you would be forced to only use that source with the SEX amp.

A better solution, IMO, would be a pair of 15k:15k or 10k:10k input transformers before the stereo pot (buy a 15k or 10k pot respectively).

I think most solid state gear has capacitors at the input.  Even a brief bit of voltage sneaking into an op amp or transistor where it shouldn't be would cause instant failure of that part.  If you had some grid voltage on the driver stage of the SEX amp, you would turn it on, notice it sounded like crap, stare at it for a few minutes, then hopefully turn it off before the fuse blew.  You could fix the problem and probably still get very good life from the 6SN7's, even though you tortured them a little.

[Grainger49]

I don't know that Anthony has a component that produces DC at the output.  I think that he was only asking about an input coupling capacitor because he has seen them elsewhere.

Anthony, do you have a component that has DC on the output?

[Paul Joppa]

...
I think most solid state gear has capacitors at the input. 
...

That's because transistors A) need a positive bias with respect to the emitter, and B) the base draws current. So the base is usually connected to a supply of the appropriate voltage and current. Even an opamp will need some DC bias current, which will not be available from a source device with an output capacitor.  An input capacitor isolates this DC supply from the input device. Tubes however do not normally demand any current into or out of the grid; they are happy with a small resistance (1 meg or so) to ground.

[Paul Birkeland]

Ah, many thanks PJ.  I had read somewhere about people removing the input caps from solid state gear, but it sounds like that could be a dubious modification.  Can this be circumvented with a dual rail supply?

[Grainger49]

PJ,

I hadn't thought about that.  I do know that some do not have capacitors in the audio path.  It was a selling point many years back.  But it certainly makes sense to block the base's bias.  Come to think of it, my GainClone doesn't have caps in the audio path, well not in the external circuitry.  I think that is a unique case though.

[Paul Joppa]

Yeah, you can get away with no input capacitor and a bipolar transistor by either using a low input resistance (10K is typical) or by compensating the bias current. Both approaches keep the voltage at the input small - a few mV is usually OK. Or of course you can use a FET input. The input capacitor is not always necessary, but it was often used especially in the early days of transistor applications, before bipolar supplies and complementary transistors became common design features.