I have connected my pre-amp's tape monitor rec-out directly into an amplifier, then placed a Niles volume attenuator in-between the speakers and the amp to control the volume.

It seems to be working just fine, but I have heard 2 different opinions about this configuration:

1) BAD IDEA! The tape monitor rec-out sends a signal to the amplifier that is WAY too powerful, and will obviously over-load the amplifier. Configuring your system this way is sure to over-load and damage your equipment.

2) IT'S FINE! Regardless of the strength of the source signal being sent into the amp, it is actually the speakers that 'pull' the power from the amp, and ultimately determine what the 'load' on the amp is. By placing the attenuator in-between the speakers and the amp, you are changing the load that is being demanded from the amp. Attenuating the speaker has the same effect on the amp as attenuating the source signal.

What's the correct answer and why?

This setup appears to be functioning normally. The amp stays very cool, rising in tempurature, normally, when the volume is turned up for extended periods of time, then automatically cooling down when the volume is lowered.

Am I going to damage my equipment by configuring my equipment in this way?

If you are attenuating the input to the amp and the impedance matches alright it is OK. If you overdrove the amp you would hear it as gross distortion. If you fed it through a pre-amp you would essentially be doing the same thing: putting an attenuator between the tape out and the amp, the attenuator being the pre-amp's volume control.

This setup will work just fine as long as the volume attenuator is rated to handle the maximum rated output of your amp. After all, that's exactly what Niles attenuators are designed for: to be installed in a wallbox, being fed the output of an amplifier that is being fed by the straight line-level signal of a CD jukebox or source selector, and then allowing you to attenuate that level down to a normal listening level in the room.

Yes, but I am NOT attenuating the INPUT to the amp. I am connecting the tape monitor rec-out DIRECTLY to the amp. The source to the amp is running with a full, un-attenuated signal. The volume attenuator is just a speaker volume attenuator that is connected on the other side, in between the speakers and the amp (using the speaker wire, not the rca cables). So I am NOT attenuating the INPUT at all - I am instead attenuating the OUTPUT signal from the amp to the speakers... does this make sense?

You are feeding the power amplifier a full line level signal at all times. This is the same as volume control at maximum level. You are driving the amplifier at full level, to produce the maximum power output, more or less.

Any attenuator in the speaker leads can only work by dissipating the difference between full amplifier output and what the speakers use. In general that will be most of what the amplifier puts out moment to moment.

I would think that the amplifier would be driven into clipping fairly often but I suppose its overload headroom might be sufficient so that this happens less frequently. That attenuator is like an extra old refrigerator, using up a lot of electricity for very little benefit relative to its cost.

If he were using a purely resistive attenuator, such as an L-pad, that would be true, but the Niles attenuator is an autoformer, which is basically a multi-tap inductor. I'm still not exactly sure how it works, but I have used them in the past with moderate-power (60W/ch) amps and they have always worked very well.

It is possible to run into problems if you have mismatches in amplifier input sensitivity and source output level, i.e. high input sensitivity + high source output, or a very powerful amp (200W/ch) driving just one attenuator and one pair of speakers, but if everything sounds fine, then generally speaking, it is fine. Distortion is an audio system's way of telling you that something is wrong.

So... if the amp is running full-tilt, because I fed it a full line-level signal, and I am listening at a low volume, where is all the heat going?

Normally, when I run the amp at full volume, it gets really hot! If the attenuator is working to dissipate the unused power, shouldn't IT get really hot?

When I run the system at low volume levels, there does not appear to be any additional heat being dissipated anywhere. The amp remains cool and the volume attenuator does not seem to be warm in the least.

It only seems logical that if the amp is really running full-tilt, as you suggest, heat would need to dissipate out of the system somewhere...?

Per your suggestion, I went and checked with the manufacturer's to confirm BOTH the amp's input sensitivity AND the pre-amp's source output level:

INPUT SENSITIVITY: Adcom GFA-535 amplifier, 97 milivolts for 1 watt output, rated at 60 w/ch into 8 ohms.

SOURCE LEVEL OUTPUT: Parasound PSP-1500 pre-amp/processor, line-level Rec-out, 500 Milivolts

I don't understand how to interpret these stats to be able to tell if there is a mismatch in the input/output sensitivities. But I can confirm there is not any distortion, so I assume their O.K. ?

Also, I noted that the amp has relatively low power (60 watts/channel), and the Niles volume attenuator is rated for loads up to 100w/channel.

The amp is an Adcom GFA 535 rated at 60w/channel. It has a thermal protection circuit to shut it down if it overheats, and a distortion/clipping indicator on the front of it. The amp has never clipped or overheated in its current configuration (in fact, it does not even really heat up at all when I play it at low volume)

The volume attenuator is rated to handle up to 100 watts/channel. The amplifier it is connected to is an adcom GFA 535 rated at 60 watts/channel. Looks good to me! From the sounds of your post, this setup is working just fine!

the 500 millivolt thing... is that peak or nominal? 500/97 ~= 5.15 watts. that's not much (though definitely audible!) if your amp can do 60 watts. however, if the 500 millivolts refers to 0 db VU = 500 mV, then expect peaks at +10 dB VU, giving 51.5 watts. that's still not too bad - your amp wont be clipping. however, a particularly loud recording may well hover around +20 dB VU... you'll definitely hear this as distortion.

though i may be talking crap. it's been a while with the maths...

I'm not sure if it is peak or nominal.

The products owners manual did not even list the output sensitivity of the rec-out section of the pre-amp at all in the specifications section. I had to call the company and speak with their technical support person to get an answer on this. The technical support person could not even say 100% if this was the EXACT specification of this particular model. His response was "I'm sure its gotta be just a standard line-out sensitivity, like any of our line-outs, which is typically 500 milivolts".

Should I go back and get more clarity on this? The technical support person did not seem to think there was much variation between what he called a 'standard line-out' sensitivity from one component to another. Is there commonly alot of variation between the different models of various manufacturers fixed-line-outputs? or are they, as the technical support person suggests, more-or-less 'standard' from one manufacturer to the next?

I think the attenuator works like a variac, this is a transformer with one (1) coil.
The Tap slides over the outer layer of the coil which is denuded for this purpose.



With such a circuit the amplifier sees a much higher impedance on its output when he listens at low volumes, that's why it remains cool.
Nevertheless, a transformer in the power output line always produces more non-linearities than a potentiometer in the input.

This appears that it may be getting more at the heart of the real answer!

I will need to spend some time getting schooled on the basics of electronics to really understand the dynamics that result in your conclusion:

"A transformer in the power output line always produces more non-linearities than a potentiometer in the input"

As far as my question "will I damage the equipment?" - it sounds like the answer is NO, it won't DAMAGE the equipment, but it may not sound as good as it would if I could hook it up in such a way as to attenuate the source instead of the output. I think this is a more accurate answer than simply saying the amp is being sent a full line-level signal which will result in overloading and potential damage.

I don't understand why adding much higher impedence from the speaker-side results in lowering the volume being output from the amp (therefore eliminating the need to dissipate additional heat from any unused power) - but I don't doubt that your on to the right answer!

OK, back to your original question: at least as long as there is a loudspeaker connected, your equipment won't be damaged. When the output is open, the amplifier sees a pure inductive load. I know some Amps begin oscillating when the load is too much capacitive, don't know how it is with inductances.

This also should not damage the Amp, but if it oscillates with the speaker connected your tweeters might get a little bit too warm.

The non-linearities (e.g. harmonic distortion) come from the not ideal magnetic characteristics of the transformer core, which is probably iron. As long as the core is not saturated, this will not sound bad, it will be a bit like tube amplifiers or a tape recording.