Another thread on preamp 'gain'

[Deang]

I wanted to start a new thread on this, and will probably start a new thread on this every time it comes up until I really understand what is going on with this.

In Eric's/ Ed's Moondog thread below, Eric said, "The gain of most tube preamps is really more than the couple of volts the Moondogs need to push them into the clipping region."

I know when Leo was getting ready to buy his Moondogs, he had some concerns about this as well.

I'm having a really rough time understanding what is happening here. I mean, if the volume is turned all the way down on the preamp, and there is no sound coming out of the speakers -- there is no 'gain', right? So then you turn it up, there is 'gain' -- which makes sound. Why does it just seem to me that the only thing that should be impacted is how far the volume control knob is turned before before the amp is overdriven? I'm clearly missing something here.

Most preamps supply between 18 and 20db of 'gain'. Is this number a constant then? I just always figured preamps start at 0 gain, and then as you turn the volume up -- the gain is increased until one reaches the 18 to 20 db figure. Hmmm, let me guess -- it doesn't work this way does it?

On the other end of this equation is something like Mike Sanders' Horn Monos. He specifically designed these for Big 'Ole Horns, and they are advertised as being "low gain".

Are low powered amps by their very nature "low gain"? What constitutes low gain in an amp, what constitutes low gain in a preamp, and how does one go about determining what should be matched with what? High gain preamp with low gain amp, low gain preamp with high gain amp, low gain preamp with low gain amp, high gain amp with high gain preamp -- did I leave anything out?

Can these things be explained in such a way as not to cause my brain to collapse in on itself?

How about one of you genius' make up a nice chart for the rest of morons telling us what numbers work best with what numbers.

Oh yes, and then there is the output impedance of the amp and the input impedance of the preamp (or is it the other way around) -- doesn't this relate to this whole issue of 'gain'?

Thanks to whoever takes the time to help out here.

[elan120]

Dean,

Let me give it a try to answer your questions.

First, when you have a preamp with a gain of 20dB, that means the voltage gain is 10X. Therefore, if you have a 1V input to the preamp, it will give you 10V at the output. Next take a look at the input sensitivity of the amp. Lets assume that is 1.5V to full output, and the preamp gain is 20dB, then to get to the maximum output from your amp, you will only need 0.15V at the preamp input. Following that, you can take a look at the attenuator. Obviously, you don't want to listen at maximum output from the amp, so some amount of attenuation is needed. Here, lets assume you need 1/5 of the maximum power from your amp, so the preamp input requirement is now 0.030V. Typical CD player output is at 2Vrms max., so to use that value as base to calculate the amount of attenuation required, it is equal to about -30dB. Lets say if you have a step attenuator that is capable to do the attenuatiion from -44dB to 0, that means from the -44dB position, when you turn up around 5 clicks(assume -3dB per click), you will get the preamp input voltage you need.

All these are just a sample. Actual result will depend on the maximum amplifier output power, speaker sensativity, room size, CD player output voltage, the recording level, listening loudness level, and the type of music you play...etc.

I hope this make sense. If anyone sees any mistake I made, please feel free to correct it.

Kevin

[Deang]

Welcome to the forum!!

Excellent first post.

Now I just need a little while to figure it all out.

Thanks!

[edster00]

I was hoping to see more response to this thread but it appears elan120 said all that needs to be said on the subject!

FWIW, I thought I would mention that I have always had little excursion of the volume control on any preamp when driving the larger more sensitive Klipsch speakers. While I was talking to Mick Malone (builder/desinger of Supratek products), I mentioned this "problem" to him. He employs a dial on the rear panel of his preamps to fine tune the "gain". He put an extra pair of switches (one for each channel) in mine as a primary gross adjustment and then rear panel dial.

Here is a shot of the internal switches (mounted next to the while thingys marked Low and High):

here is a shot of the rear panel dial (located on the right side) next to the power cord connector:

I have the Syrah's gain set as low as it will go.

PS: I edited the text to correct an error in the description of the rear panel gain control dial!

[LarryC]

Dean: Let me try from a rather amateur standpoint to see if I can help out a bit. While I don't know electronics, I have had to deal with this subject to accommodate K-horn efficiency in my system setups. I copied in below what you wrote, so let's see:

I'm pretty sure that a "gain" control is actually an attenuation control, and that top volume is wide-open full gain. It seems to me that -65 dB has been a customary range for volume controls (other forum folks will be far more knowledgeable about this) However, this has primarily been an issue for me when trying to match a pre-amp with an amplifier and the K-Horns, and have the volume control sit somewhere between 7 or 8 o'clock and 12 or 1 o'clock. With the K-horn's very high efficiency, some amp/pre-amp setups will cause the volume control to sit at or near the bottom of its range, very limiting in trying to set a low volume without cutting it off (I understand that some volume controls also lose quality at the bottom of their range). This can be a particular problem with fixed-step volume controls with only, say, 24 positions (like my CAT pre-amp), especially because these can be "tapered" to have only large (6 or 8 dB) steps at the bottom of their range!

I have long felt that, with this problem, a key way to address it is with a low-sensitivity amp. In my experience, the typical sensitivity of 0.5 volts for full power out, if that power level is 100 w. or more, raises compatibility issues, because it takes so LITTLE voltage for such high power out. For many years, I drove my K-horns with Mark Levinson ML-2's, which had a sensitivity of 0.675 v. for its rated 25 watts -- I believe this would be equivalent to 1.35 v. for 100 w, a lower sensitivity and far more gentle and compatible level of amplification for K-horns. The rated gain of the ML-2, incidentally, was 26 db, which I doubt is very low -- the sensitivity for X-number of watts out seems to work better for me, than the gain figure by itself.

Anyway, I've had a heck of a time trying out 0.5-volt amps on my system -- I can't turn down the volume control enough. My current amp probably has close to a 2-volt sensitivity for its rated 100 watts out, much better.

One brand of amp I have heard of with good compatibility with K-horns is Quicksilver (I've never heard one); I'd note that the sensitivity of the Q/S Mono 100 is 1.5 v. for 100 watts out -- again, less sensitive -- and the Horn Mono, mentioned elsewhere in these forums, requires an astonishing 9 (!!) volts for its rated 50 w. out, as low a sensitivity as I've heard of. Now, THAT should get those volume controls run right up there!

There is sometimes another, related factor: the amplification of tube/electronic noise after the pre-amp's volume control, especially by a high-sensitivity amp plus K-horns. This may not be a problem with most pre-amps, but has been with mine. In that case, reduced amp sensitivity seems to be the only way for me to deal with that issue. The significance of the volume control location is that, while it raises/attenuates prior signals, the circuitry following the control is unaffected and can be quite noisy in some pre-amps. Some other pre-amps like an earlier ARC I had appear to have two volume control locations, one apparently near the end of the circuitry, which can tone down the final output circuit noise quite a bit.

Finally, one may have to combine a reduced sensitivity amp with reduced sound source levels to get everything properly pulling together for typical 9-11 o'clock control settings. For example, I have a moderately-low output (.35 @ 5cm/sec) cartridge, have set my CD player at only .4 v. or so output (well below the standard), and got Magnum-Dynalab to change my tuner output from 1 v. to 0.1 v.

Hope all this rumination helps, and my apologies for being so expansive.

"I'm having a really rough time understanding what is happening here. I mean, if the volume is turned all the way down on the preamp, and there is no sound coming out of the speakers -- there is no 'gain', right? So then you turn it up, there is 'gain' -- which makes sound. Why does it just seem to me that the only thing that should be impacted is how far the volume control knob is turned before before the amp is overdriven? I'm clearly missing something here.

"Most preamps supply between 18 and 20db of 'gain'. Is this number a constant then? I just always figured preamps start at 0 gain, and then as you turn the volume up -- the gain is increased until one reaches the 18 to 20 db figure. Hmmm, let me guess -- it doesn't work this way does it?

"On the other end of this equation is something like Mike Sanders' Horn Monos. He specifically designed these for Big 'Ole Horns, and they are advertised as being "low gain".

"Are low powered amps by their very nature "low gain"? What constitutes low gain in an amp, what constitutes low gain in a preamp, and how does one go about determining what should be matched with what? High gain preamp with low gain amp, low gain preamp with high gain amp, low gain preamp with low gain amp, high gain amp with high gain preamp -- did I leave anything out?

"Can these things be explained in such a way as not to cause my brain to collapse in on itself?

"How about one of you genius' make up a nice chart for the rest of morons telling us what numbers work best with what numbers.

"Oh yes, and then there is the output impedance of the amp and the input impedance of the preamp (or is it the other way around) -- doesn't this relate to this whole issue of 'gain'?

"Thanks to whoever takes the time to help out here."

[artto]

Just to try & clairify (simplify?)this subject (Great Job Elan), its about ATTENUATION. The "Volume" control (or as it should more properly be called, a "Gain" control) is actually an attenuator. It ATTENUATES the (amount of) INPUT signal.

Any amplfier can be driven beyond its rated output, even with the "volume control" set at a very low level, given enough input signal (provided nothing lets go (blows out) before that happens).

My ARC SP6 preamp has a 10Db gain reduction switch on the back (the ARC owner's manual even mentions the Klipschorn as a use for it).

Fortunately, for me, my Lux MB3045's seem to sound the best with the volume (input attenuator) at 12 o'clock (half of full rotation). So I've never had the too much gain problem.

[Erik Mandaville]

What Elan said! Great post!

Plus:

There are a couple of ways to 'tailor' the preamp/amplifier interface: If you're having to barely touch the volume control on the preamp to get really loud music, it's possible to add series resistance on the input of the volume control. Say, if you have a 100k potentiometer, adding a fixed resistor of the same value will reduce the output by several dbs.

Also: Reduction in the value of the volume control also will reduce the overall sensitivity, since more of the input signal is shunted to ground. In this case, a 10K pot will be less sensitive (meaning it will need to be turned up higher for the same level of loudness) than the 100k control.

And: Series resistance is sometimes used on the output of a preamp (I wonder if this is what Ed may have on his Syrah...). This would be a fixed resistor that comes immediately after the coupling cap and connects to the RCA output jack. Depending on the value of the resistor, this CAN reduce the output of the preamp, but can also mess up the frequency response, causing roll-off of high frequency information.

More to come!

Erik

[Deang]

Thanks everyone -- much appreciated.

Mark Deneen tackled this one some time ago for me, but I couldn't remember the details of his post, and didn't have the sense to bookmark it.

[LonestarBlues]

I picked up my first tube preamp from Audiogon. A used factory assembled Welborne Labs Revellie to go with my Moondogs. I really like the preamp except for one thing-the dual volume Goldpoint stepped ladder attentuators. There are 24 steps & the first step from off-no volume to the first step is too large db wise. The manual states there are 2 db increments between steps. There is no such thing as low level listening with this preamp. I am used to low level listening late night when I was using my Mcintosh C 38 preamp.

I am glad I live in a house. The Revellie & Moondog combo through Cornwalls or other Klipsch speakers would probably not be apartment friendly. This is my first experience with stepped

volume control. The other tube preamp I was looking at was the Wright WLA 12. It has a switch on the back of the unit that reduces the volume level by 8 db. Something like that would be nice

in my opinion on stepped volume controls. But I don't have the technical smarts to know if something like that would work with stepped volume controls. And if it would work how much db level decrease would be needed to work.

Any of you knowledgable guys care to comment. Thanks.

Regards, Mike

[LarryC]

Mike: My CAT pre-amp had the same problem (24 steps, 6 or 8 db first step at the bottom) until CAT (Ken Stevens) installed a factory control with the reverse taper -- small bottom steps, maybe 2 dB apart, and big jumps at the top where I couldn't possibly use it. The switch idea sounds great -- can the preamp's manufacturer help?

[Kudret]

I came across this old thread which describes the same problem I am

having

with my Cary SLP70 preamp. The problem is, the Khorns get too

loud as soon as I slightly increase the volume pot from the off

position.

I was wondering if someone has found a solution to this problem. Does

anybody have any schematic for building a switch to attenuate the

input, or

if I changed the volme pot from 25K to 10K (I was planning to upgarde

the volume pot anyway) would this help?

Kudret

[sfogg]

Put a set of attenuators between your pre-amp and amp. Something like this:

http://cls.audiogon.com/cgi-bin/cls.pl?accstwek&1133095960

Or if you are at all handy with a soldering iron you could make the same thing yourself. It literally is just a couple of resistors in those plugs. Very inexpensive to DIY.

Shawn

[Erik Mandaville]

Hi, Kudret:

I haven't seen the link Shawn sent, but I can offer maybe just 1 or 2 cents worth on what has helped for with hyper-active volume controls in the past.

So, you have the Cary driving your very fine Seth monoblocks, correct? And you are finding that the volume control on the Cary has a sort of hair-trigger sensitivity?

I think your approach on changing the value of the potentiometer is absolutely on the right track, but it doesn't sound like it's going to be enough to go from a 25K to a 10K ohm control. There is a formula for this, but I can't remember off hand (I'm horribly at math, anyway!). In any event, it's normally thought better to pad the input into the preamp rather than the output after the coupling capacitors. The reason for this once again has to do with that old imped. relationship equation, where adding series resistance on the outpu would certainly lower the gain going to the amps, as well as improve your S/N ratio, but you would probably need something like 100K ohm or more in series with the input of the amplifier in order to bring the level down to a more tolerable....well.....level. I did this awhile back with some other preamp, maybe it was my Foreplay, and the gain was indeed reduced, but at the cost of performance, which was really bad! From talking in the past, I know you also like to hear good clean detail and transient response (I think the main reason we all like Klipsch speakers so much), but those series resistors killed all of that, and turned the sound to mush.

Instead, try padding the input between the selector switch on the preamp and the volume controls (if the R/S signals come off the center poles of the selector switch), starting with the same value as your control -- or 25K ohms. Just tack solder in the resistors, experimenting with higher or lower values as needed until you have an improvement. As long as you don't go higher than maybe 10K ohms or so (this is with an input z on the amp of 100 K) you could also try some series resistance on the output of the preamp, and maybe get decent results without grossly influencing the sound.

One can purchase ready-made inline filters that will do this, and you can accomplish the same thing by using these on the input RCA jacks on the Cary. I remember another time I had this problem, and I installed a volume control on the Moondogs in order to find a good compromise.

Erik

Oh -- this is probably a chance in a million, but you might check to confirm that your potentiometers are log rather than linear taper controls. Those linear pots get hot very early in the rotation.

[Erik Mandaville]

Ok, I see that Shawn has indicated the outputs here. He is of course right that will work, but you just might want to be careful with the amount of resistrance you're using there. Maybe with the Cary, it won't even be an issue as it was with me.

Good luck.

Erik

[Kudret]

Thanks, Shawn.

What should be the value of the resistor?

Kudret

[Kudret]

Hi, Kudret:

Instead, try padding

the input between the selector switch on the preamp and the volume

controls (if the R/S signals come off the center poles of the selector

switch), starting with the same value as your control -- or 25K

ohms. Just tack solder in the resistors, experimenting with

higher or lower values as needed until you have an improvement.

As long as you don't go higher than maybe 10K ohms or so (this is with

an input z on the amp of 100 K).

Thanks, Erik. This sounds like a good idea to try.

Which approach would have the least influence on the sound quality? If

a 10K pot will give me a noticable improvement (reduction in gain), I am tempted to choose

this option.

Kudret

[Erik Mandaville]

Kudret:

Sure, I agree! changing the volume control would be easy. What's better might depend somewhat on the specs of the preamp. If the preamp has a low-ish output impedance, you could get by with a combination of both the plugs on the output, as well as shunting more of the signal from your source to ground with a lower value pot.

Also: do you happen know the value of the input resistor to ground in the Seth/s? some amps seem to really want to have say 100k or more there (tube amps, that is...), but you might try another value of possibly 68K or so?.

One other thing that might be overly involved but is effective: You could install, within the chassis of the Cary, two trimmer potentiometers, linear taper pots, of say 50K ohms. You could send the signal from the selector switch to these trimmer pots, and use a fine screwdriver to experiment with the rotation that works best. I'm sure you've seen these things, Kudret -- they are small, and take up very little room. They would work as input level controls that, once set, can be left alone, allowing more flexible use of the main volume control.

Erik

[Marvel]

A lot of these ideas will change the impedance of the circuit. I posted

a small program some time ago that calculates values for pads, three

resistors for each channel. It allows you to choose the reduction in db

and the input and output impedance.

Isn't this what the stepped attenuators do anyway? Keep the impedance the same but pride the gain adjustment.

I think Erik got this at one point. I have matched some things up with

this and gotten very good results. I would be happy (as I said before)

to calculate the values if anyone wants the info.

Bruce

[Erik Mandaville]

Marvel:

Yes I did! Good ideas, too.

Erik

[PrestonTom]

It is important to pay attentuation to what Marvel said.

The cable from the pre-amp to amp will have a small amount of capacitance (especially if it is not short). Changing the resistance by swapping the pot or adding resistors MIGHT inadvertently produce a low-pass filter (single pole with a corner at 1 / 2*pi*R*C). The output impedance of the pre-amp and input impedance of the amp are not arbitrary.

Good luck,

-Tom