Passive pre amps

I tried a Nobsound switcher/ pre with a volume control with my Coda stage amp last night. I was surprised to find it better than my long time IRD preamp. More transparent and smoother sounding through my Klipsch Quartets. In fact, I think it overall sounds better than my tube integrated. :( that kind of bums me out. The quartets have always sounded better with lower gain and my pre, which has variable gain, was best on the zero gain setting. Im certainly not missing dynamics or drive with this little Nobsound.
 
I tried a Nobsound switcher/ pre with a volume control with my Coda stage amp last night. I was surprised to find it better than my long time IRD preamp. More transparent and smoother sounding through my Klipsch Quartets. In fact, I think it overall sounds better than my tube integrated. :( that kind of bums me out. The quartets have always sounded better with lower gain and my pre, which has variable gain, was best on the zero gain setting. Im certainly not missing dynamics or drive with this little Nobsound.
Detailing and superb imaging, it shouldn't be this affordable or easy. Best with an adequate power supply of ca 19-24V/ 4A.
 
Impedance matching is a very specific concept and has absolutely nothing to do with audio. Not sure what term is appropriate, so I'll use ICOIIAOI (intelligent choice of input impedance and output impedance). OK, that's never going to catch on.

Sometimes you'll hear the rule of thumb of using at least 10X the loading impedance as the output impedance of the previous stage. That's also nonsense, assuming you have flat impedance with frequency. As an example, you can load a 50 ohm output signal generator with a 50 ohm load and it will work perfectly from DC to daylight. The problem would be if a power amp input impedance varies with frequency. Then a passive attenuator might give you rolloff or other non-flat response. That's never been the case for any amp I've measured, but then that's a very small subset of all amplifiers ever built.

You want your passive attenuator (pot) value low enough so that the output impedance isn't too high at mid volume, so response remains flat and added noise is kept to a minimum but high enough that it doesn't excessively load the signal source. I've used a passive stepped attenuator since back in the days of the Ace Audio "Zero Distortion" preamp of the 1970s. All they did was pull the buffer stage out of the design and give it the zero distortion moniker. I've never found I needed the additional gain or buffering of an output stage. Every time I've added one, I soon took it out as detrimental to the sound. Regardless, I still need a powered RIAA phono stage, but when not in use I don't even bother plugging in the power cord.

Self suggests an interesting idea for attenuators in his small signal book. It turns out if you drive a very low value pot (1 kohm or less) with a modern opamp, you can get lower noise than a higher value pot would generate all by itself. It will also have low output impedance because it's a low value pot. I'm thinking the ultimate version would use a stepped attenuator with lower than normal resistor values.

At first glance, the idea of putting an unnecessary transformer in the signal path seems nutso, but transformers transform impedance along with voltage. They transform impedance as the square of the turns ratio, and voltage directly as the turns ratio.That means they can provide a stable low impedance drive to the power amp at lower and mid amplitudes where most of us old folks listen. Good ones are expensive, so experimenting with those will have to wait until my ship comes in.
 
The term "preamp" when applied to a passive controller does make sense when viewed with some historical perspective. What we now usually refer to as power amplifiers were simply called amplifiers back in the 50's. Anything that came before the amplifier in the signal chain was known as a pre-amplifier. It's therefore perfectly logical to use a simple passive device before the power amp if nothing else is needed and refer to it as a preamp.......

Yep, agree, and even today the definition of an integrated amplifier is preamplifier and amplifier combined in a chassis. Unfortunately, gain is so cheap, and sources as CD players (cheap gain stages, gain of two, or 6db), or amps with high ~28db gain with no volume control/selector switch, that hum problems, from different sources develop, although not impossible to rid. That CD player would still output of 1 vrms without that extra gain stage, for better fidelity.

-----------

Of course, too high of passive impedances (resistor types), combined with ic capacitance, and highs tend to be lost. Too low of impedance and bass is lost (capacitor coupled output). Kind of a balancing act imo.

With transformer types, one has to be careful of resonance/ringing problems involving the capacitance of the ic and input of the amplifier, output of the source, resistances. Optimum response is when the Q of the circuit (preamp output, ic, control, ic, amp input is ~0,7. In either resistor or inductor passive, does not take much to alter the sonics.

The input impedance (z) of an amp usually does vary with audio frequency. The criteria are the dc input resistance and input capacitance of the amp. For simplicity sake let's keep simple. Input resistor is 100k ohms. 50pf input capacitance results in reactance of only 160k ohms at 20khz, 320k ohms at 10khz, or 640k ohms at 5khz,,,, in parallel with 100k ohm amp input resistor. Complexity increases with dc coupling and low impedances, so something to consider from that standpoint.

keep on truckin

joe
 
Last edited:
Sometimes you'll hear the rule of thumb of using at least 10X the loading impedance as the output impedance of the previous stage. That's also nonsense...
"Nonsense" is a bit harsh. The idea of a low output impedance and high input impedance is to maximise the voltage transfer. The output impedance of the source and input impedance of the load form a potential divider and a ratio of 1:10 gives minimal loss - just under 1dB. I presume that's where the 1:10 "rule" originates, ie keeping the signal loss below 1dB.

BTW, Douglas Self is quite scathing about passive preamps and while there's no doubting their technical limitations my experience has been that they often sound great. What "problems" they have tend to be quite benign, whereas the distortion introduced by a poor active preamp can be a lot more insidious, as fredgarvin describes in post #21.
 
Yes, if you want maximum signal, don't load the source! Can't disagree with that. It's just that some have implied that there will be quality loss if you go below 1:10. IMO, the odds of that are probably greater with tube equipment where the preamp might be fussier about load.

I have great respect for Self, but he does prefer that his electrons be well controlled and not much left to chance so far as loading. As I said above, every time I add a buffer stage, I end up with too much gain and a perceived loss of sound quality, for no benefit that I can see. It's counter-intuitive, but I've always had more trouble building a well behaved circuit with no gain, than the typical gain of 10 or so. It's typical to include 100 ohms or so in the output of most solid state preamps. IMO, that's not much different than using a lower value stepped attenuator at the lower settings, with something solid driving it, though that goes right back to the problem of building a good X1 buffer. Maybe I'll just go back to using a bird with a sharp beak, a la Flintstones.
 
I have great respect for Self...
It's counter-intuitive, but I've always had more trouble building a well behaved circuit with no gain, than the typical gain of 10 or so.
That's interesting. D. Self's Small Signal book goes into quite a bit of detail about bipolar transistor emitter followers and the dangers of such circuits oscillating at frequencies higher than many 'scopes can see. I wonder if that was that the problem?
 
Interesting thought, as at the time I was using a slow scope. I've got a faster one available now, 350 MHz. I had put the buffer on a PCB that I could easily drop into the preamp. I'd like to try it again, but alas, I seem to have lost it. Happens with more stuff than you might think.
 
I chose the Pass buffer because I trust him to get it right with the least parts count in the chain. I'm leaving it in for a while, considering a switch point ladder passive as another alternative if I continue to experiment.
 
At first glance, the idea of putting an unnecessary transformer in the signal path seems nutso, but transformers transform impedance along with voltage. They transform impedance as the square of the turns ratio, and voltage directly as the turns ratio.That means they can provide a stable low impedance drive to the power amp at lower and mid amplitudes where most of us old folks listen. Good ones are expensive, so experimenting with those will have to wait until my ship comes in.

from http://audiokarma.org/forums/index.php?threads/high-end-separates.666925/page-7#post-9068389
There are 26 preamps in Stereophile's Class A list this issue (October 2015). In alphabetical order, with prices listed, they are:

Audio Research Reference 5SE, $13,000
Audio Research SP20, $9,000
Ayre Acoustics KX-R Twenty, $27,500
Ayre Acoustics K-XE MP, $4,350
Bespoke Audio Company Passive, $12,000
Boulder 2110, $55,000
Classe CP-800, $6,000
Convergent Audio Technology SL1 Reference, $9,995
D’Agostino Momentum, $32,000
DarTZeel NHB-18S, $31,700
Lamm Industries LL1 Signature, $42,790
Luxman Classic CL-38U, $4,200
Music First Audio Baby Reference, $8,590
Nagra Jazz, $14,500
Parasound Halo JC 2 BP, $4,495
Pass Labs XP-30, $16,500
Placette Audio Active Line Stage, $6,995
Promitheus Audio Reference TVC, $890
Shindo Masseto, $13,500
Simaudio Moon Evolution 850P, $30,000
Simaudio Moon Evolution 740P, $9,500
TAD Laboratories C600, $42,000
VAC Signature SE, $19,500
VTL TL5.5 Series II Signature, $8,000
Ypsilon PST-100 Mk2, $37,000
Zesto Audio Leto, $7,500

Total $466405
Average $17,942.50

When I see the average price of "Stereophile Class A" preamp at nearly $18K, it just tickles me pink to know that I have sound that plays in that ballpark, for $580. :yes: I must live with only one volume knob, however. Just thought I'd spread the word!

http://www.audiokarma.org/forums/showthread.php?t=649914

No affiliation. YMMV. Have a banana.

:banana:
 
I have no problem with using a passive volume control at low to medium levels. A 50K pot is still going to present at least 20K impedance at up to half volume. As an impedance ratio to the typical source output impedance, this is still going to be at least 30:1. As you go to higher levels the impedance ratio will fall rapidly. This can cause excess current demands on the output of the source component which could lead to clipping. This may be a problem with people using low power tube amps (I mean the need to turn them up).
Mind you, a lot of active preamp front ends (especially when incorporated into integrated amps) still have input switching followed immediately by a volume control. This arrangement falls into the same variable impedance trap. A well designed active preamp should have an input buffer amplifier, preferably with a very high input impedance (several megohms is good), followed by a volume control and then a high current capacity output buffer amplifier. This arrangement always presents a constant and very high input impedance at the input meaning that the component output voltage amplifier (in CD, tuner, phono preamp etc) will never be required to produce excess current due to driving into a low impedance load under any circumstances. It is also a good method for impedance isolation/control of the source and power amplifier. This is also where ultra low noise OPamps are useful. They can offer very high input impedance, far better PSRR than most discrete designs and also a symmetrical output which negates the need for DC offset blocking caps in the signal line.
 
Has anyone used a "passive pre" with a tube amp? If so, what were the results?
Yes. They tend to be more passive friendly with their typically higher input impedance. My VTLs are 130k ohm while some SS designs are as low as 10k.
 
I have no problem with using a passive volume control at low to medium levels. A 50K pot is still going to present at least 20K impedance at up to half volume. As an impedance ratio to the typical source output impedance, this is still going to be at least 30:1. As you go to higher levels the impedance ratio will fall rapidly. This can cause excess current demands on the output of the source component which could lead to clipping.
I don't understand this logic. Surely the input impedance will depend on whatever follows the pot. If the 50k pot was followed by a high impedance circuit (the 1 meg input of a valve amp for example) the impedance seen by the source would be 50k (very nearly) regardless of the position of the volume control.
 
I don't understand this logic. Surely the input impedance will depend on whatever follows the pot. If the 50k pot was followed by a high impedance circuit (the 1 meg input of a valve amp for example) the impedance seen by the source would be 50k (very nearly) regardless of the position of the volume control.

I agree with you Roth. A 50k pot, at mid resistance, and using a 6dj8 tube's input capacitance will have a minimum frequency response of ~110khz, -1db. Above or below that control setting, the high frequency response rises. I don't see the need of a buffer between the volume control and gain stage.

One also does not need an output buffer stage unless the output impedance (z) is, say, above 5,000 ohms, or the ic to the amp has high capactance, say over 250-300pf. Frequency response is in the hundreds of khz unless one has high capacitance ics (long) between preamp/source and amplifier.

However, with three stages, there are many more sonic robbing parts, solder connections, layout problems/less channel separation etc. So, Power, I don't see how 3 stages in a preamplifier is advantageous except in lower fidelity full house scenarios.

keep on truckin

joe
 
Last edited:
Back
Top Bottom