leok

kajh, Interesting point. I hadn't realized that the autoformer outputs on McIntosh ss amps had impedance taps. As with tube amps, changing the tap will change the impedance "seen" by the output drivers (transistors in your case) as well as the amp's damping factor or output impedance seen by the speaker. It will also change, as was noted above, the output amplitude. But one assumes you have output amplitude to spare so that's not an issue. I'd try the 4 Ohm and even the 2 Ohm taps. Two things that I can think of may be improved or made worse by the change: 1 is the amp's crossover distortion, which would be harshness in instruments like violins or exaggerated, unnatural sounding sibilances. This would be as the impedance seen by the output transistors moves toward or away from an ideal impedance at which they are the most linear in the crossover region. 2 is control over bass: lower Ohm output taps will generally provide tighter, better controled bass. Try them all for a few days each and see which is best for your speakers. Leo

Shawn, Essentially, I agree with all or most of your comments. I do think the impact on what people hear is more benign (thus my comments about this issue being hyped) than simple numbers indicate. In that the upper frequency response is flat for some nominal tweeter impedance means that the deviation for a particular tweeter is only a function of that tweeter's impedance deviation from the amp's target nominal. People may simply not know this. Also, the pwm immunity to impedance swings at mid and low frequencies is beneficial with many of the older Klipsch 3-ways which have a significant impedance peak in the midrange, making them sound thinner with designs less tolerant of impedance swings in that area. PWM also handles the often encountered impedance dips in the bass region. For those who are concerned, the pwm output filters can be tuned to maximize flatness in the upper region. For most, this is probably not an issue, and not a source of such objectionable types of distortion that are perceived as, say, harshness. And, in the end, remember: this is a $130 stereo amplifier we're talking about. Leo

Shawn, 2 Watts may be way too high. With 2 Watts acoustic program material, power to the tweeter is probably 0.2Watts. This may be part of what is confusing here. People are reporting that the high end sounds detailed and extended (and that's what I hear too) while measurements are claiming the high end is reduced by as much as 50% or more. Can you try the output impedance measurement at a fraction of a Watt? On the datasheet, the load used showing 1dB down at 20KHz was 4 Ohms. Power is 1 Watt. Value of the inductors is most likely 10uF. My experience with datasheets has been, when reality doesn't match the spec., the measurement conditions are different. For example: I don't believe the 0.03% distortion spec. is met with a capacitor on the input, which is why, in my application, I re-biased the circuit so my input is DC coupled. I think there is something going on with the amp's performance at top frequencies, but I doubt it is really cutting in half (or even a quarter) the power going to the tweeter, and we have yet to learn how to effectively measure what is going on. To paraphrase a certain statement: "If is sounds good and measures bad, you're measuring the wrong thing." OK, I just re-checked the datasheet: The frequency dependency you mention is shown (Frequency Response, P7). That is measured at 1W. I suspect the "output impedance" is lower at lower power. Also, the filter in this case appears to be tuned so that a tweeter with 6 Ohms impedance will produce a flat response to 20 KHz. So, even if the "output impedance" is high, this doesn't translate into rolled off highs, and may translate into exaggerated highs if tweeter impedance is over 6 Ohms. One might check the impedance of the tweeter in use and compare it with the Tripath Frequency Response chart. The filter in my amps is the one measured by tripath to produce this chart. I don't know what values were used in the various amps that use Tripath chips. Leo

Shawn, I just checked the TA2024 datasheet. If the output impedance was really 4 Ohms at 20 KHz, the output should be down by 3dB at that frequency. According to the datasheet, the output is down by only 1dB at 20KHz. I think there's more to this, and for some reason, the impedance measurements are misleading. Leo

Shawn, What was the output power during impedance measurement? PWM, due to output filtering, may be slew rate limited making it appear to have a higher (than actual) impedance at upper frequencies. The power handling capacity of the tweeter of a KHorn, or any other speaker is a small fraction of the power handling of the speaker overall. Slew rate limiting would limit the power available at high frequencies while not really reducing power available (or increasing impedance) for realistic acoustic requirements or demands at those frequencies. Leo

Shawn, I'll take a look at that. The way pwm works I'm a bit skeptical .. both numbers look high. Also, I'm not sure what the lower frequency "impedance" measurements really mean in that the output straddles the target output voltage at all points, meaning the output can't loose any amplitude due to "impedance." I'll take a closer look and see if I can add anything helpful. PWM tends to be tough to measure using standard methods. Leo

pauln, The statements about pwm output impedance being a function of frequency is more hype than reality, especially with the Tripath amps that use modulation frequencies around 1MHz (which is high enough so that there is little impedance change even at 20KHz). PWM is, in fact, remarkably immune to impedances of the speakers it is driving: real problems being excess current which could fry the output FETs (although there is a pretty good protection scheme in place for that) and opens. These problems could damage the amp (as they might damage other ss or tube amps). Treat the amp with the kind of care you would any other amp. If you hear it clip, turn it down .. and if it clips you'll hear it. Resist the temtation to see how hard you can drive it until in completely collapses .. you might have an answer, but no amp (as might be the case with any number of amps treated that way). Leo

Plug the cd player directly into the 10.1a inputs. Connect the speakers to the 10.1a outputs (observe 10.1a speaker terminal channel ID and polarity). For phono, use the NAD's tape outputs. If you must include the NAD for other (line level) sources, use the NAD's tape outputs also. All of the above minimizes influence of the NAD on the sound (esp the first example, which doesn't use the NAD at all. That way, if there is a difference between the NAD and 10.1a, you will be able to hear it. Leo

Thanks Mark, I meant to write "HDCD" and have corrected the offending entry. Leo

mas, The technology. When one records HDCD, or puts an HDCD decoder in a player (and adds the "HDCD" logo, one pays royalty to Microsoft. From Wikipedia on HDCD: "Currently, Microsoft, the owners of HDCD technology, have removed all information about it from their website." Leo

Hello Max, My day job stepped in with a high priority "task" lasting a year. That just ended. I hope to get back to the Hypex in a month or so. I obtained the Hypex parts and sketched the amp and power supply last year. So all I have to do is finalize the designs, purchase parts and build. I hope to have amps within 6 mo. Good to hear from you. Leo

I've used the Tripath TA1101B (previous generation Tripath chip, same power) to drive my RF-7s. The sound is good. They won't get loud enough to damage your hearing or produce thundering bass. But used within its power capability the amp will do well. And, again, within that power capability, It will do fine with the RF-7 impedance dips. If it clips you'll know it. Leo

"You ain't listenin' to bookshelf speakers so don't choose power like you are." No, bookshelf speakers are less efficient and require more power for equivalent amplitude than larger, more efficient speakers. So, go for high quality (low distortion) at low power. If you can get some power with that too, fine, but don't trade away quality at low power. Leo

Kaiser SET Say, Unfortunately, my cd players don't have SACD decoders so, my understanding is, when I play SACDs I get 15 bits of music and the least significant bit is encoded dynamic info that is just noise to a standard 16 bit decoder. I was under the impression that SACD was taking off, Microsoft bought it, and almost everyone stopped putting the decoders in DACs and players and encoded disks virtually stopped. Does anyone know if this is the case? Leo Correction, 20-Apr: I meant to write "HDCD" above, not SACD. sorry for the confusion. Leo

Congratulations, Looks like a great way to go. You'll enjoy it. The HDCD opens up some opportunities for the Reference Recordings. Leo

speakerfritz, By 10 years ago 8x interpolation (oversampling) was the norm. That resolves to 352.8K samples per second. The norm for DAC resolution was 20 bits or more. What has improved mostly is the low level distortion characteristics of the op-amps used as current to Voltage converters and buffers. You can get a 10 year old DAC, install pin-compatible op-amps from Analog Devices or Burr-Brown (TI) and have a very competitive DAC that is inexpensive. The 192Ks/s upsampling converter in my Philips DVD 963SA is just a tad smoother than the 8x interpolation/20 bit system in my Sonic Frontiers "Transdac" in which I have upgraded the op-amps. The "old" DACs got us 90% of where we are now. They can be very good additions to a system. Leo

Dave, I'm not questioning that you can, or anyone can get a more economical, and and even musically "better" sound with a ss DAC. I'm suggesting that the comparison, in this case may be to an op-amp job in a cd player that wasn't done well. I've heard plenty of terrible sounding ss junk, and a lot of even mediocre tube equipment sounds better. So, one could draw the conclusion: tube sounds better. Does tube sound better than well done ss? I don't happen to think so, but one would have to then know which ss to go for. That's all I was, and am saying. Leo

The "tubed" portion is the audio output buffer and sometimes the current to Voltage stage. The tube versions are running class A and have a good chance of sounding better than the sloppy (and cheap) work done with op-amps on many DACs. Leo

Hmmmm .. one of the reasons I prefer the Klipsch 2-ways is I prefer how they handle violins. I suggest you don't dump the KG4s for a 3-way until you've made a comparison. Full bass and treble backed off .. wow. Something's not quite right. Odd that if it's the speakers they would both be broken in exactly the same way. Thebes, if you're following this, wasn't that a pair of KG4s we were listening to on your downstairs system? Any ideas? Maybe I'm thinking of a different speaker. Leo

Does the MV45 have 4 Ohm or 2 Ohm outputs? Try the lowest Ohm output available. If there's an impedance dip in the KG4 in the bass, that will help. Also, can you skip the preamp completely, driving the amp directly from a CD player or Phono stage using a passive volume control or volume control in the player/stage? That's just to determine if the problem is really just the preamp, not the speakers. If you can get your hands on a little Sonic Impact amp, try that. It won't do structural damage to your house (or your ears), but has a reliably neutral sound with low impedance output and is a good cheap reference. Try placing the KG4s in a corner. I believe they have rear firing port or bass radiator. I've heard KG4s driven flat (no tone controls) and I thought they were very nice. I actually prefer the 2 way sound to the three way sound of the cornwalls for much of what I listen to. Good luck Leo

I try for as close to what would be the natural concert level as possible. Since I prefer sitting midway back in a concert hall where the volume is moderate, this works particularly well with recordings that present a similar soundstage (which includes a fair amount of hall reverberation). With recordings that put the listener in the front row (some string quartets are done this way (DG esp) I find the appropriate volume is higher. Since I prefer sound a bit quieter I tend to move further away from the speakers when listening to such recordings. As usual, for me, what's important is how far into the silence the dynamics of the presentation can go, not how loud it can get. Leo

Martin, I don't think it will hurt anything. The volume of the horn will be unmatched (higher) than the cones, however, because for a given Volume Control setting the 8 Ohm Voltage out from the amp is higher than the 4 Ohm Voltage out. Give it a try and tell us what you hear. Leo

Nice amps! I really like the parallel feed (parafeed) design which keeps the bias out of the output transformer. I find it provides significant improvement in high and low extension. I ended up modifying my Moondogs to parafeed output using either Magnaquest or Sowter transformers. I think those Bottlehead amps are the way to go for SET. Does anyone know the design of the Bottlehead input (gain and driver) stages? Just curious. Leo

Thebes, sheesh. how do you think of that stuff? Leo

It's a difficult call. I've tried Jensens and Hovland (and others) in both the pp and set amps. Others were easy to rule out. I finally settled on the Hovland Musicaps in both. They sound more natural and detail is improved. Leo