leok

shaman, I am very happy with 2A3s (Sovtek or RCA) driving RF-7s. did you try the 300B amp(s) set for 4 Ohm output? If not, give it a try. In fact, I recommend that with any tube amp. Also, I prefer parallel feed output ("parafeed") to the more conventional, but they're not that easy to find. Leo

Ray, If you're beginning to listen to "classical" I suggest you pay attention to the distortion at low power, of any amp you consider. There is a lot of low level detail in classical recordings. If your amp does not handle low power well, much of the more delicate sounds (strings, harpsichord, even trumpet) will sound piercing and harsh. One way to say it is: power gives you volume, low distortion at low power gives you quality. Some types of amps that handle low power well are: many tube amps, Class-A transistor amps, "digital" or pwm amps (pwm stands for pulse width modulation). If a transistor or pwm amp claims under 0.05% distortion at 0.5W or lower, it's probably pretty good. If distortion is only specified at "rated output" it could be pretty bad. With tube amps, It's hard to say because it's so dependent on implementation. Unless you have a reliable endorsement or return policy, listen to an amp on your speakers before you buy. Also, try a few out to get some feel for the differences. Good luck, Leo

Duke, The affect of limited low level dynamic range on ambience is, because it is distorted and noisy, it becomes just that: noise. The recording looses spatial information and sounds flat. There is not defineable "room" in which the acoustic event takes place. Also, distorted and noisy harmonics mix with the distorted and noisy ambience and instruments loose their horizontal positioning with pieces of their spectrums appearing at various places on the sound stage Instruments and vocalists tend to jump around, or at least some of their harmonics do. Sound also tends to glare. As the lower power is cleaned up the sound stage becomes more easily defined and 3-dimensional. Instruments stay in their places right left front and back. Echoes are distinct from the dry or direct originals. Mark, Pretty amazing. I'm considering the 180 instead of the National. Leo

Congratulations. Several years ago I gave up waiting for a pair of CF-4s and went for the RF-7s. Not many CF-4s around and people who have them don't tend to give them up. Leo

Duke, It depends on the source. The THD & Noise levels of the Tripath chip I use in my amp are below .03% from 0.5W to 6W. For that technology that seems OK. I suspect the Tripath distortion starts up below 0.5W but probably isn't over 0.1% at 0.1W. I hear a little edge now and then that I don't get with the tube amps or Crown D-45. So, I'm simply applying similar criterium to any ss linear or switching design (below 0.05% THD&N is probably OK), and I'd prefer that go down to 0.1W. The Crown D-45 takes it lower than 0.1W and I hear no edge in that amp at all. Tubes are different. In no-feedback triode designs, the distortion of the tube itself approaches 0% as power approaches 0W. This provides a low power advantage. Since that region is important to me I tolerate problems with high output impedance and rather benign high distortion at moderate power. However, With the Crown D-45, and what looks like a few additional ss linear and switching designs I can have a clean enough low end, low output impedance, and clean higher power (finally, the original promise of ss without the trashed low end). Leo I missed the where part: Mostly higher frequencies and ambience. Actually, there is a lot of wood and bow noise on a cello that gets "enhanced" way out of proportion with distortion. I really don't like the gritty "enhancement" of sibilances. Ambient information is so much better with low distortion and noise at low power. My wife and I were listening to a recording: Schubert for Two by Gil Shaham and Goran Sullscher. On the first track we noticed that the two instruments were recorded with completely different ambient information. It sounds like a second room is overlayed on the first when the violin opens. It was totally clear with the Chorus-IIs driven by the Crown. Now that's just an amusing example, but ambience helps me place instruments on a sound stage and that's important to me. Placing a flute 2/3 of the way back in an orchestra, hearing the echoes on the back and side walls, is a fair amout of low level detail that I think requires very low levels of distortion and noise at low power.

Mark, Specs. on the UcD180 are remarkable. The amp appears to have effective dynamic capability from below 10mW to something on the order of 100W, depending on where one decides distortion gets too high at the top end. I'm used to amps with about 2 orders of magnitude of low distortion power and having to carefully place that effective power where I listen. This amp has 4 orders of magnitude of low distortion power, appearing to extend, as shown, one order of magnitude beyond each end (low and high power) of the region I consider most important. This thing is clearly worth a listen. Have you heard one? Leo

Mark, Of course, I didn't think of it .. I wouldn't have to run at max supply voltage. Thanks for the data sheet link, I'm looking it up right now. Leo

Mark, I've read a few articles, but haven't looked at data shets or specs. I'd be interested in any tech. references you have. Thanks. Leo I checked a few references. 120W @ 8 Ohms is the lowest power module. Interesting technology, but not exactly optimized for where I want the amp's dynamic range placed. Looks like a good path for those interested higher power.

Duke, I re-read your first entry just above. The op-amp is the major source of gain. But the additional power circuitry bolted to its output has to remain effective at low power for the entire unit to function correctly at those powers. This may not have been the case with older designs. The use of class-a drivers as output devices at low power helps to maintain gain and speed in that region. And, in keeping with the subject of this thread, this provides the low end capability of the amp allowing dynamic range to include low level detail. Leo

Shawn, I was refering to the small single chip versions I've used. In those, there is no feedback, in a traditional sense, after the pwm modulation. Duke, Funny thing about the feedback. I think with enough of it, fast enough, and making sure it remains effective through the zero crossing (low power) range it can work. The LM2876 data sheet has an interesting sentence or two about "crossover distortion." It refers to the potential loss of gain due to low bias in this area. This would reduce feedback's ability to keep the system linear. I think a lot of ss designers get into trouble here and for my use of horn loaded speakers, it is the most critical range of operation. The effective designs appear to use very high gain, and bias the entire circuit for speed, and continuity over the zero crossing region. Feedback is very much an "engineer's" kind of solution Rather than try to eliminate distortion (use a device or algorithm that is inherently linear like triodes or pwm) they ask "How low is low enough?" Well, it seems, at least for me the Crown D-45 has it low enough. Plus it's low enough in the power range that I use. Leo

There does seem to be some improvement in recent ss designs wrt low power performance. But maybe knowing where to look helps too. The D-45 was quite a surprise. As a result I am considering construction of an amp based on the National LM2876 (smaller but similar to the National ICs used in GainCard/Clone amps). They all act like big, very high gain, op-amps using tons of feedback with all devices turned full on (except the power transistors). Also, the driver stages provide some or all low power output current depending on design. PWM design (Tripath for example) is a completely different approach to the problem. Linearity is inherent in the algorithm and output to input feedback is not used. But PWM does have its limitations and the National specs. are significantly better at low power than the Tripath (which itself is a lot better than very many older ss designs). I'm curious as to whether or not I will hear an improvement. Of course, if I don't hear an improvement it could be I've reached my own threshold of low power distortion detection. If it's worse, that's another can of worms, including my particular implementation. As for the Crown D-75, It may be the D-45 and possibly the D-75A are different in that Crown directly targets low power applications (they refer to the applications as "medium" power). Some of their literature mentions that these amps are appropriate for use with horn loaded speakers. Leo

Nothing much to add here, just agreement. Loud isn't of much interest to me anyway, but lower level timbres and acoustic ambience are important. As I've reduced distortion and improved signal to noise of both systems many recordings have taken on a three dimensional quality creating a very stable sound space between and behind the speakers with recorded echoes helping to establish instrument locations in that space. To me, as has been said above, dynamic range starts with 0dB set at a level I want to listen to and my system can accurately reproduce and extends downward into low level detail. I believe by choosing amplifiers that are close to their limits at the volumes I prefer for 0dB, I am maximizing signal to noise and dynamic range .. geting as much of the low level detail as the system permits. Leo

I use the Belkin F5C980-TEL, available at Staples for about $80 .. maybe less. It won't rob your amp of dynamic range because the inductors are too small to affect the 60Hz power .. they only remove RF. What it will remove, if your system has the problem, is RF which can modulate the audio producing what one might interpret as improved dynamics or "slam" but is actually just distortion timed with the sound itself so one might think it is part of the sound. Many people are so used to hearing rf modulated audio, they think something is wrong when the rf modulation is removed. Leo

Ben, Not too sure what you're looking for, but I'll give it a try. My favorite system is the RF-7s driven by the 3.5W 2A3 amp. The system has sufficient power (current and Voltage) for a lot of what I listen too under most conditions. In those cases the system has sufficient "headroom" it doesn't clip or compress the sound. In some cases I may have the volume up a bit higher than usual and notice the onset of compression, "glare," (the sound of a tube amp running out of Voltage or current "headroom"). In that case I can either 1) turn the volume down a click or two or 2) get a more powerful amp (pp, Tripath, or Crown). Headroom is sufficient current and voltage, available in the amp's output to accurately follow the program material. If you need +/- 5V and 1A to accurately drive your speaker and your amp has just that capability then you have sufficient headroom for that occasion with that equipment. If you decide to turn up the volume you'll run out of headroom and will clip (or begin to do the compression thing tube amps do which I consider clipping). If you turn the volume down below the +/- 5V and 1A requirement you have a little excess headroom which is of no particular sonic value but might be fun to "know." If you need 5a and 25V capability then that's what you need. If you have it, you have sufficient "headroom" and the sound is good. If you don't, then you need a bigger amp. If you need +/- 5V and 1A and have an amp that can deliver 5a and 25V you'll never run out of headroom, but you might find the amp has some distortion and noise at the lower levels creating a gritty or harsh sound. In that case you need an amp that better matches your requirements: say, one that can deliver 5a and 25V and still handle the 5V 1A and below OK, or one down to a +/- 5V and 1A that also handles lower levels well. Leo

Fish, I agree with your point, except possibly you know a 12ga wire that is the standard and any other is creating "altered" sound? Possibly the wire you are refering to is creating altered sound and some other 12ga is creating the real thing. How do you know which one is doing the least altering? Does cheapest mean it's the most accurate? Somewhat academic, I know. Leo

thebes, Very interesting. Thanks for posting. I used to have a Sansui 5000A receiver. It was ss, but pretty nice. Leo

HCM Audio sells the audioquest "GBC Delux SST" 12 awg speaker cables, custom terminated, at excellent prices. The build and materials quality alone is worth the price. I use these on both my systems (one stereo, one 3.1). hcmaudio.com With different cables, you can certainly see a difference on a scope (that's why one calibrates scope probes, for example) and I can hear a difference. But, in both cases, the differences are small. Leo

Michael, Right, I wasn't really thinking about the line sources, except that at moderate listening levels you might be more sensitive to distortion and noise. I was thinking more of the amplification, or the environment within which the amplifier is working. Do you use a line filter? Have you tried an amp that's designed to excel at low power? If not, it's worth a try. The inexpensive Tripath implementations, the Crown D-45, or a SET that you borrow. You may not like everything the low power or pwm amps do, but the experiment will tell you if the harshness you refer to is speaker or signal. I thought there was a TEAC amp making the rounds as a loaner. Anyway, I encourage you to give high quality low power a try. Leo

I dont have exactly the same speakers refered to here, but it's my experience that if you drive Klipsch speakers with a signal that's sufficiently free of distortion and noise at low power, they are very detailed, but also free of any harshness. I would look at cleaning up the signal. Now the disclaimer: I have installed Voltage conditioned Hovland caps in the Chorus-II and RF-7s. But the RB-5s are stock. I suspect the Hovlands mostly improved detail and it's the clean signal that removes the harshness. Other speaker technology can be fine. But I don't think it's correcting for Klipsch inability to play effectively at low volume. I think it's correcting for electronics' inability to deliver a clean low power signal to the speaker. Leo

Craig, I don't know if your approach is equivalent, better, or less effective. But, it addresses the same problem the Black Gate addresses as far as I can tell. Leo

I've have tried the Aleph-3 and Forte-II. Beautiful, powerful sound. A bit of noise when idling, but nothing that really bothered me. Leo

Hi Kelly, The Belkin F5C980-TEL (avail at Staples when I got it). By the way, a SET output stage works at constant current, diverting the current to speaker or storage cap (power supply or ultrapath depending on what you have) with the output waveform. Changes in output power requirement (dynamic changes in the music) should make virtually no change in current drawn by the amp. On the Cap: What does the Black Gate cap do? It absorbs high frequency and RF energy (more effectively than cheaper caps), preventing that HF noise from getting to the audio circuits. In more technical terms, at high frequencies the Black Gate has lower ESR than its cheaper siblings. The cap involved in the Philips player is the main charge storeage reservoir following power line rectification. All the switching and linear regulators draw charge from that reservoir. Leo

Chris, I don't know. It seems the 5.1 outputs use better buffers, or possibly a more direct path (no channel mixing stages). There is a substantial improvement over the stereo outs. Someone else on this BB actually discovered the difference and posted here. Of course I can't remember who it was... Leo

For clarity, I'd go with a 2-way .. something in the Reference series. I haven't tried any of the older Klipsch 2-way speakers.they may be as good as the Reference. Leo

Mark, Big surprise .. I was surprised too. Do you have a power line filter on your system? If not, that's another easy way to better sound. The power line filter does for the rest of your system what the Black Gates did for your 963. Also, I assume you're using the multi-channel outputs, not the stereo outputs. If not, try the multi-channel outputs. Leo