Amplifier Clinic [Monday epilogue added] [Friday post epilogue added]

My recent audio concentration areas are (1) the tonearm in the main system (now, based around the Hafler 9300), and (2) the surround sound system in the kitchen.

It was the latter that inspried me to get an ATI 1502 amplifier, to better handle my current hungry Revel M20 speakers than my Yamaha HTR-5790 receiver (which was getting awfully hot, and heating up my 4x4 matrix switch above it terrible, both situations dramatically improved by a 1/4 inch foam board I installed very securely just below the matrix switch...the suspicion that the foam board is even making the Yamaha cooler has me thinking there was kind of a reinforcing "oven" effect with the bottom metal panel of the matrix switch).  The Revels are 6 ohm speakers, and the Yamaha nominally can only handle that by flipping a special power-reducing switch--which I don't believe I've done as some experts believe that unless you're out of control levelwise, you get better performance leaving the switch in the 8 ohm position.  I didn't tell you that.

I figured the ATI to be the "perfect" AV amp, very solidly made, low noise and distortion, etc.  And at 150W, sufficient power for the Revels if not the Acoustats (which I was not, really, buying it for).

One thing, however, is that the ATI does NOT have DC trigger.  That led me to think maybe I'd move the ATI into the bedroom system (which has no trigger, and I keep the amp running all the time for several reasons) and move the Parasound HCA-1500A in the bedroom to the kitchen.

And I've long admired Morris Kessler for many reasons: he's time tested in Audio having designed great amplifiers with other contributors (like James Bongiorno, for awhile at SAE) and by himself (the SAE XXXXII was apparently his own design, and one loved most by J Gordon Holt of Stereophile).  He believes in making reliable products--and right here in the USA in my home state of California.  My hero Sigfried Linkwitz was a reliable user of ATI amplifiers; he loved the multichannel ones for his multi amplified speaker systems and highly recommended them to others.

I visited a friend's friend who was selling an ATI amp about 10 years ago, and it sounded good to my friend George and I might almost have bought it except for the flickering light on the power switch...which the unit I have now also suffers from, now not surprising in a 25 year old amplifier.

But, does Kessler make amplifiers equal to the designs of John Curl, whose used amplifiers like the 1500A are often even available (though at slightly higher prices) than the "equivalent" ATI amplifiers.  Curl certainly has more mindshare among high end audiophiles, while Kessler may be better known on the Home Theater circuit nowadays.

I decided to have an "Amplifier Clinic" both to test my new amplifier and ensure it is up to snuff, and to see how, objectively, it compares with the Parasound HCA-1500A.  The 1500A is, btw, a design favored by Curl himself over the original cheaper models (like the HCA-1000A) and which is very similar to the Halo and JC amplifiers in basic design, with JFET inputs and MOSFET drivers and bipolar outputs.  It's also almost contemporaneous with the ATI 1502, both coming from the early to mid 1990's (I think the ATI was a few years earlier, maybe).  I believe the ATI may be an all-bipolar design, Kessler designs being like that.  Curl uses bipolar outputs, but FETs in other locations.

I'm very surprised at how little power this amp uses at idle, 40.3 watts after a few hours "warmup" (if you can call it that) idling.  It started at 38.9W, showing incredibly stable bias, but also very low bias, almost "Class B" operation, something I'm predisposed against.

It also made some chassis hum, but about the same as the Parasound HCA-1500A in a different location, as far as I could tell.

For some reason, my first day of measurements using my (personally calibrated) Sound Technologies 1700B and my Hantek scope and my recently purchased 500W-2000W 2 ohm to 16 ohm loads went badly awry.  I wrongly measured the max output on the ATI which seemed to be only about 27V into 8 ohms (100W), and from then on every measurement I made seemed to confirm the amplfier wasn't that good.

Then I measured the Parasound, which I had often worried about because I leave it running all the time, and it measured basically fine.

Then I compared the numbers...and in many cases they were identical.  Something was wrong here.

I determined several errors in my measurement technique.  Including that I cannot simply turn up the analyzer voltage leaving it on some lower scale.  Somehow that pre-distorts the output.  I discovered to better measure amplifier clipping on the ATI like most amplifiers I need to use the 100V measurement scale on the Sound Technology.  Then it was clear the ATI was clipping at 41V, equivalent to 210W, only a little less than the Parasound which went to 45V or 253W.  The Hafler 9300 clips at about 40.5V.

After that, the distortion and noise numbers are generally so similar that anyone would be hard pressed to say they make a definitive audible difference, despite the distinctly different circuit designs, parts, and philosophies.  Generally the Parasound measures slightly lower distortion, and slightly more so at higher frequencies, but so similar it could be measurement error (I've often made much larger errors--such as my first time around measuring the ATI!).  Curiously, on my very consistent Meguro noise meter, the wideband AND "A Weighted" noise of the two amplifiers is almost identical but slightly in favor, strangely enough,  of the lower bias ATI.

Update: By the following weekend, I had obtained my second Hafler 9300.  This was supposedly "upgraded" but reading the fine print, it's more like just being refurbished.  The original 20,000uF caps per channel per side were replaced with pairs of 10,000uF caps.  Even doing this replacement is difficult in the small Hafler chassis.  I measured this refurbished 9300 and also my  original "museum quality" NOS 9300.  They measured identically, which is evidence the refurb (which also included the replacement of some small transistors) was done well.  I put the refurbished 9300 into my system, and decided I might like the sound better, having more punch from the new capacitors.

Wideband Noise:
Parasound  (Channel not clear maybe both) 0.12mV wideband
ATI Ch1 0.052mV Ch2 0.056mV
Hafler 0.028mV (both channels)
Hafler Wins, then ATI
(Hafler also has widest bandwidth...so this is a surprise)

A Weighted Noise
Parasound (Channel not written down) 0.039mV
ATI CH1 0.030mV CH2 0.035mV
Hafler 0.0026mV
Hafler Wins, then ATI

Distortion at 0.01V output
Parasound 0.075%
ATI 0.046%
Hafler 0.08%
ATI Wins!  (looks like error, though, Hafler should be best because lowest noise)

Distortion at 0.03V
Parasound was 0.075%, curiously identical to 0.01V, measurement error???
Hafler 0.023%
ATI not measured.
Hafler Wins!

Distortion at 0.3V
ATI 0.05%, slightly higher than 0.01V (higher???)
Hafler 0.023%
Hafler Wins!


Distortion at 1V (which, btw, corresponds to 0.125 watts)
Parasound: 0.025%
ATI: 0.055%
Hafler: 0.026%
Parasound wins!  Finally it seems the lower static noise of the ATI has been exceeded by some kind of dynamic noise in the ATI that starts appearing just below 1 watt--crossover notch???  Parasound
has also pulled ahead of the Hafler, they are but neck and neck.

Distortion at 3V (about 1.25 watt)
Parasound: 0.008%
ATI: 0.023%
Hafler 0.0094%
Parasound wins by a significant margin over ATI, but ATI still only 0.023% distortion, below
the rated level.  This, nevertheless, is the comparatively "worst" point for the ATI, probably a result of crossover notch distortion.  At higher levels, the ATI catches up again.  The Hafler is still almost as good as the Parasound, just slightly worse.

Distortion at 10V (12W)
Parasound: 0.0049%
ATI: 0.01%
Hafler: 0.0068%

Distortion at 30V (125W)
Parasound: 0.0045%
ATI: 0.0055%
Hafler: 0.0073%

By the time you get to 125W, the ATI has nearly caught up with the Parasound in THD, and the Hafler has fallen further behind--but these are tiny levels way below rated THD and close to my measuring residual and greater differences can be caused by my 1700B during warmup, so the apparent differences here might be 100% measurement error and I wouldn't take them too seriously.  The Parasound and ATI give their best measurements at 30V but the Hafler shows its best measurement at 10V curiously.  But the distortion numbers are consistently small, showing that all 3 amplifiers are very well designed transistorized Class AB amplifiers performing like new despite average age of 25+ years old (!!!).

At 10kHz, near ATI full power, the superiority of the Parasound is most clear:

Distortion at 30V (125W) at 10kHz
Parasound 0.023%
ATI 0.13%

But, 0.13% distortion still probably isn't audible (and certainly not at 10kHz) and 125W at 10kHz is pretty extreme (though I use such measurement as proxy for probing more complex things)...  At more realistic levels the amps are again virtually indistinguishable:

Distortion at 1V, 10kHz
Parasound 0.026%

Distortion at 3V, 10kHz
ATI 0.023%

I should also point out this is getting very close to the rated power of the ATI, while it's quite far from the rated power of the Parasound.  As one gets to rated power at high frequencies, distortion in most amplifiers goes way up.  (And in very many cases, such as amps I looked at this week in Stereophile, distortion at 10kHz goes far higher than 0.1% at lower levels than 125 watts).  To be completely fair, the Parasound ought be compared, say, at 175W, which is relatively the same to rated power (and also, to actual hard clipping).


Into 4 ohm loads, the Parasound tended to clip at 40V instead of 45V, and distortion numbers generally higher above 1V, but still below 0.016% at 1kHz:

Parasound into 4 ohms, 1kHz
30V  0.016%
10V  0.006%
3V    0.013%
1V    0.027%
.1V  0.075%
.01V  0.075%

My sighted listening suggests the ATI a tad brighter sounding on the M20's, but that might suit the bedroom acoustic anyway, with the Parasound better on the M20's in the bright sounding kitchen.

I don't at all believe these amps would be distinguishable in level matched blind tests.

Update 3/18: I put shorting plugs into the ATI in the bedroom, and the output is dead silent.  I had to double check the power switch.  I've taped all lights except the clip indicators with electrical tape--crude but effective.  So the front face, now pulled fully forward in my bedroom rack is "out there" providing limited warmth (not much) to my cat, who was disappointed not to find his warm Parasound last night, only an amp deep in the rack bottom which was covered in dust.

The Parasound stated idling at 31 watts but quickly started rising at first, then slowly after an hour or more to 59 watts, pretty much my predicted "+50%" from the ATI.

Update 3/22: Later did reach 60 watts.

Others DO say the ATI is on the brighter side of neutral, fwiw, but also beautifully transparent, revealing instrument tonaties as well as any amp.  I have not measured damping factor vs frequency which would be the most revealing in these regards.  Frequency response either but we can assume it's quite flat to 20kHz for both amps (and from what I remember from the specs about 100kHz in both cases, or maybe 60kHz).

I am beginning to think both Parasound and ATI amps have too much chassis hum for continuous operation in the bedroom.  In the deepest night, the hum can be heard.  I believe it was that way  before, I always blamed the neighbors A/C compressor, which is less than 10 feet from the bedroom exterior wall.

But finally I did turn the ATI, now pressed into the cabinet a bit for more help, totally off.  And the late night silence was staggeringly different.

I am now beginning to believe what Benchmark says about switching power supplies (not to be confused with digital class amplifiers).  Switching power supplies are quieter and induce less magnetic effects from leakage.  (And, they're regulated too!)  In principle, and the good ones moreso.

But I'm sure both amplifiers would benefit from capacitor replacement, as the old capacitors probably leak far more than brand new ones, and that leakage requires more current through the transformer, and hence more physical vibrations.

Epilogue

For 24 hours after the second weekend of amplifier clinic, I tried to integrate the ATI 1502 into my bedroom system.

The first challenge was the physical hum level.  Intial impressions were wrong, it does appear this unit has more physical hum vibration than my other amps.  But since it is fairly thin front-to-back, it could be pushed all the way back in my wood rack-cabinet.  It can also be angled various ways.  I tried both, and it always seemed I was making progress, but after a few hours it was the same old.

Even when it doesn't seem like you can hear the hum directly from the unit, it's energizing the room some other way, and you can still hear it.

This motivated me to finally add a remote Insteon on/off module to control the amplifier.  I found a spare on/off unit in the laboratory.  I figured I could add the amplifier to the same button on the bedside insteon controller that controls the Meural picture.   I would only turn either on when I was not trying to fall asleep.

That problem being worked-around, I focussed on the very slight buzz I could hear from the right speaker.  I thought I had heard a little of that previously with the Parasound, but not as much.

I first tried making sure the two digital EQ's, which are grounded, were plugged into the same outlet strip as the amplifier.  Well, the amp was plugged directly into the Brickwall surge protector, so I managed to plug the EQ's in near it, though I couldn't plug both into the same duplex outlet on the Brickwall as the amplifier itself.

That didn't seem to help much.  I also tried different interconnects from the EQ's to the amplifier, including a new pair of very well shielded Blue Jeans interconnects, which didn't help either. I was beginning to think the hum was in the output of the Behringer EQ's themselves.  So I did an upgrade I'd been planning to do for years.  I wired in one of the two Denon DVD 5000's as the dac driving the amplifier for the main speakers.  The 5000's use dual differential 1704's and digital inputs, and I purchased 3 of them so I could have triamplification using all 1704's before I decided the 5000's just didn't sound wonderful on the living room system.  So then I put two of them in a pile in the bedroom, thinking they'd be good enough for the bedroom amp and subwoofer.  There they sat for two years until now.

This took some work, but was successful in reducing the electronic hum coming from the left speaker significantly.  Not total elimination, but close enough I thought.

Then I listened for awhile to the living room FM tuner through Sonos on the bedroom stereo.  Perhaps it's largely because I've gotten used to the much better sounding kitchen tuner (which now has an outdoor antenna), I was not liking the sound.  It was too noisy and also bright and edgy, almost unpleasant to listen too.  Well this was a combination of the slight edginess I had always thought the 5000's have (along with a crystaline clarity that often makes you overlook the edginess) and what appears to be the slight added brightness of the ATI 1502, a match made in hell.  But there was one other thing.

One of the things I had overlooked in the amplifier replacement was the 12dB Harrison Labs attenuators.  I just put those aside, assuming incorrectly I had not been using them correctly.  But it turns out these attenuators, or something like them,  are ESSENTIAL in the bedroom system, otherwise the DACs are only using a fraction of their dynamic range, with the signal strength meters on the Behringers only blinking the bottom light.  Unless corrected (which I didn't discover I needed to do for years when I first started using the Behringers) the sound becomes very coarse and grainy.

I first measured the Harrison Labs attenuators, to ensure they are designed for plugging into the destination for a signal.  They are: the top of the attenuation network is connected to the female RCA hot connection where you plug the signal bearing cable in.  It would not be good to plug them into an output signal RCA jack, AND they would have been worse had they been designed to be used that way because then a resistor would be in series with driving the cable.

The Harrison Labs attenuators greatly improved the sound by restoring the digital resolution available from the DAC.  And also reduced the buzz even more.  But I was still not liking the sound.

So finally I decided on a different plan.  I returned the Parasound HCA-1500A to the bedroom, and decided to use what  used-to-be my living room amplifier, my original NOS "minty" Hafler 9300 in the kitchen, and the new refurbed 9300 (which has more punch, it seems) in the living room.  This repurposes the ATI 1502 as nothing more than an emergency spare for now.  Something I didn't really need to buy (though...perhaps it is the magical "black swan" amplifier which measures fine but sounds bad enough to be different on blind tests...?).

This decision was not at all based on scientific proof that there was anything audibly wrong with the 1502, except for the chassis hum, and also when I plugged the Parasound into the now improved bedroom system (with the upgrades I had done to try to make the 1502 acceptible) the Parasound had virtually no electronic buzz compared to the trace of buzz still remaining with the 1502.

I did feel the sound was more pleasant on the noisy FM radio, using the Parasound.  Later as the evening slipped by, there were patches of greater noise on the radio, and I wondered just how true it was that the Parasound made the FM noise sound less objectionable.  Maybe it was just a larger than average patch of noise that happened just as I had hooked up the ATI.  Perhaps it had just been bad luck for the ATI.  But I felt very happy about my new amplifier lineup.  Also I had made several long postponed upgrades to the bedroom system:

1) Using 1704 based DAC instead of Behringer DEQ output for the main speakers.
2) High end teflon cable between DAC and amplifier.
3) Remote On/Off switch on the amplifier.
4) Verified correct use of the Harrison Labs attenuator.
5) DAC and amplifier plugged into same duplex outlet on the surge supressor.

It might be better yet using the Integra Research DVD player instead of the Denon 5000.  The Integra Research also uses 1704's, but is somehow better sounding I've felt, and I was using it on the living room system for over a year.  A Denon 9000 might be better sounding still, but I feel they run too hot for being left on all the time.

Post Epilogue

Several days later I got motivated to check the time alignment between subwoofers and monitors.  I had only guessed the latency added to the monitor side by putting the DVD-5000 in the signal path.  So I started playing the dirac pulse from the polarity checker app on the bedroom system.  I noticed that the  6 inch woofers on the Revel M20's were really popping out.  That wasn't supposed to happen on this system, the M20's are crossed over at 60Hz.  Looking at the Behringer RTA for the monitors, once actually set to the digital output (as going to the 5000), I could see there was no rolloff in the bass.

After the Bedroom Tact preamp died, I had figured out how to get along without it.  I'm running the digital signal directly from Sonos into one of the two Behringers, and then, while that Behringer is putting out the high passed or low passed analog signal, it is also putting out a flat digital signal for the other Behringer.

With this arrangement, it's clear that the "last" Behringer is the only one that can put out a high passed or low passed signal.  If the signal goes through the monitor Behringer first, then it will be putting out a flat digital signal and not a highpassed one.

As it turned out, I was putting the digital signal from Sonos into the subwoofer Behringer first, then taking the flat digital from there for the monitor Behringer.  But, not remembering that I had done it this way, and thinking I was putting the Sonos signal into the monitor Behringer first, I had set the digital output to flat on both Behringers.  And when you set the digital output for the coax output on the Behringers, it applies to the optical output as well.

So, after trying other things then figuring this out, I fixed the problem merely by changing the digital output on the monitor Behringer to the highpassed output.  I also plugged the toslink cable into THAT Behringer, instead of the subwoofer one.  These were the mistakes I made when first setting up the DVD-5000 so I could get less hum out of the ATI amplifier.

And this could explain why I thought the sound of the ATI amplifier wasn't right.  It was not getting the highpassed signal, and the noisy FM radio signal has lots of bass modulation noise which is causing much more woofer excursion than the monitor speakers should have had.

Then, after switching back to the Parasound, I had expected the situation to improve, and it seemed to at first, but later not so much, both in the quality of the sound (a bit rougher than expected) and the chassis and electrical hum levels, which seemed to increase significantly after the Parasound had been turned on for a few hours.

Thus, my judgement about the ATI not sounding quite as good is now fully called into question.  Not only was it a useless non-blind snap judgement listening to a highly variable source, but I had things set up incorrectly in such a way as to cause the slightly off sound I thought I was hearing.

This sort of mistake happens all the time to me, and I suspect other audiophiles make mistakes also.

The still unadjusted lack of time alignment could be a factor also.

TIM and all that

Subjectophiles don't want to believe measurements, particularly THD measurements, but there is no good reason for this.  (Of course, you can't just go by specs, which may be lies, but actual measurements at different power levels and frequencies, etc.)

Some of the rampant distrust of measurements began with audio engineers who did a lot of them.  Matti Otala was a designer for Harman Kardon who designed the great Citation XX and many other good amplifiers.  He determined that conventional THD measurements were incomplete, and a new kind of measurement was needed he called TIM, and the way to achieve TIM was by eliminating or reducing global feedback.

This whole theory is still believed by many, but is wrong.

It has been determined that TIM is nothing unique, and especially that it does not require a particular kind of design (low global feedback) to achieve it.  It has been proven many times over that high feedback amps can have vanishingly low TIM as well as other kinds of distortion, simply by being designed for low THD under all conditions.

One Huge Gap in Surround

Problem with the 7.1 analog inputs on my Yamaha HTR-5790 is that there is no way to make them do any processing, they only go straight through.  Therefore, I am stuck with whatever functions my Oppo BDP-95 has, and those are limited to volume, delay, and mixdown.  The Oppo offers no functions to convert 5.1 to 7.1 using Dolby Pro Logic II x, for example.  With an AC3 connection, I should be able to select whatever processing I like, but then I have a compressed signal, and therefore am losing the full resolution of DVD-Audio and SACD recordings.  I can see why the receiver is like this, it must only have 2 ADC's inside, and it would need 6 or 8 to handle processing of analog multichannel inputs.  Are there any receivers that DO handle such processing???

This makes one of the fundamental issues in Surround, the inherent incompatibility of 5.1 and 7.1, much worse.  To explain, requires a history lesson.

I was becoming an audiophile just as the original Quadraphonic wars were running.  Despite their differences in implementation, all Quadraphonic systems simply assumed a 4 channel speaker layout, 2 speakers in front, and 2 in back as mirror image.

It turns out that this 4 speaker configuration is fairly worthless.  Humans do not have very good front-to-back auditory sensitivity.  Things in back are somewhat muted, but it's hard to have any idea of where they are unless you turn your head.

When this was all re-thought for Home Theater, it was fundamentally changed.  Instead of Quadraphonic (because more is better or something) we got Surround, in which an ambient soundfield is created arournd the listener, starting with SIDE speakers.  This actually has some perceptual utility, maybe (though I think that often multichannel listeners don't appreciate the forward depth of stereo systems, which is often superior to multichannel in practice for various reasons, one of them being the confounding effects of the center channel, which IMO should be eliminated).

But somehow, the sense that "back" was needed still  seemed important to some people.  So, the 5.1 specification fudges the side speaker position so that they are also somewhat behind the listener.   Often the specification is something like 90 degrees to 110 degrees, but most people, especially the relatively uniformed, put their side surrounds much further back that 110 degrees.

When you get to 7.1, finally it is done correctly, with both sides and back speakers.  But now, the sides are supposed to be at 90 degrees, as close as possible, right on the sides of the listener.

So if you set up your system for 7.1, as I have finally done in my kitchen system (sans center channel, which is being automatically added back into the front stereo pair), it will not play 5.1 recordings, which are the vast majority, in the way that most 5.1 systems will play them.  There is no back signal at all.

Effectively you need some kind of processing like Dolby Pro Logic II x (which, btw, my Yamaha receiver has, it says so right on the front, but it is ONLY applied to digital inputs).  Although Dolby Pro Logic is mostly known for generating fake surround from stereo, it also features a matrix for converting 5.1 to 7.1.  Although you could simply duplicate the side signals in the backs, and some schemes actually do that, it's best handled through a matrix.  What I've heard described in a matrix where the sameness between the sides is subtracted from the backs.  It might also benefit from some kind of steering which leaves things alone if they appear to be concentrated in one speaker.

Now that we have endless named proprietary schemes for doing matrixing we don't have any plain vanilla matrix boxes to fool around with as in the good old days.  Stuck with the fact I can't actually use the Pro Logic II x on my Yamaha, or find a simple 5.1 to 7.1 outboard converter I could stick in between my Oppo and Yamaha, I now dream of building such matrix boxes myself.

Update:  I may already have the answer to this problem in an unexpected place.

Two years ago I started buying Integra Research RDC-7 surround preamps simply as stereo DACs because they have dual differential 1704's, something I was pursuing at the time.  For various reasons, I decided not to use this unit, and the low 1.0V voltage for I/O would be problematic in my high end living room stereo.  I had reached the point where I was thinking of recycling at least one of these beasts (the beat up looking one) and possibly both because of no perceived need.

Sadly, the manual for the RDC-7 is not actually available, what one finds online is the RDC-7.1, a different beast.

Reading the manual for the RDC-7.1 it is clear that Dolby Pro Logic II x can be applied to the multichannel input (it lists the various options for multichannel inputs, and that is one of them).

I suspect either this or something similar is available on my RDC-7, especially because it has an "Upgraded" tag on the front.  It appears the THX Ultra 2 has a "Cinema" mode which also expands 5.1 to 7.1 surround.  In any case, this $4500 surround controller is not limited to "straight through" or "pure" for the multichannel inputs as the Yamaha HTR-5790 is.  Perhaps that's the deal, to do the cool stuff you need to get the actual preamp/controller.
  
Since the time this was made and now, generally all Home Theater Receivers and Surround Processors have been eliminating analog multichannel inputs, which  I think is a bad idea.  In addition the the compressed AC3 "Dolby Digital" signal, newer products transfer multichannel through HDMI, or at least HDMI connectors.  This may have the advantage that a digital signal is being transferred, and a layer of DAC and ADC may therefore be avoided.  I think that issue is extremely overrated--surprisingly digital coversion, especially to 24/96 and above, is essentially lossless and about as good as a pure analog interface--the kind high end audio has been built upon from the beginning.  Plus, analog signals do not have complex clocking issues or cause RFI from high speed pulses.  I am also not sure you will always get a full resolution signal via HDMI from all kinds of discs.  I have heard HDMI clocking may be problematic (but Oppo fixed this in the BDP-205) also. 

However, if it turns out I can't use the Integra Research to do my 5.1 to 7.1 expansion, any HTR with HDMI multichannel inputs and Dolby Pro Logic II x (or perhaps THX Ultra 2) will do.

I will first see if my RDC-7 *can* convert 5.1 to 7.1 through the multichannel inputs, and if so, begin working on incorporating it into my kitchen system.   My Yamaha HTR would then do nicely as a multichannel amplifier, possibly supplemented by a high end stereo amp for the front channels.

Addendum: Chesky put forward an alternative use for 6 channels in which the back surrounds are removed (they aren't THAT big a deal anyway) and replaced with front speakers at the widest possible positioning.  (Stereo can't use the widest positioning because that would destroy the phantom center image.)  Yahama and others used the extra two channels as "Presence" speakers which are above the stereo speakers.  I didn't mean to necessarily disparage these ideas, I have just found it more convenient as well as more typical and supported to have surround back speakers.

Dolby Pro Logic II x is not the only process to convert 5.1 to 7.1.  I already made that clear, but in the minds of many, all that older stuff has now been replaced by Dolby Atmos.  I think thats way over the top for home theater, and particularly for 5.1 to 7.1 conversion.  I would stay away from any process that has mandatory automatic room correction, though I'm not sure Atomos actually mandates that.

By all acounts (though none might be informed) the 5.1 to 7.1 conversion in Dolby Pro Logic II x is a matrix applied to the surrounds ONLY.  It does not reference the front or center speakers.  So, in principle, you could do this or a similar conversion with two inputs and three outputs.  This is a relatively simple thing, and for a 7.1 system, perhaps just what is needed.  As I said, Atomos looks way over the top.

In addition to "surround" formats, there are also some special formats like Amiphonics and Ambisonics.  These may require stereo back speakers to work.  These formats are somewhat controversial and I still wonder how well they actually work.  However, I purchased a few recordings of this kind, and that is another use for back speakers.  In this case, replugging connections may be the best way of handling it, plugging the 5.1 surround outputs into the back.

Measuring the FPB 300 damping factor [Updated Jan 13 2019]

With sadness but hope, I've decided to mothball the Krell FPB 300.  Though it has been very entertaining to try to work out the bugs, after 3 perfect repairs, and 3 years of experiments to get it to work properly, I've decided to hang it up until I have the spare time to disassemble and repair, rebuild, or re-engineer myself.  Which may never happen, but at least I'll have time someday to take it apart which would be worth something to me.  I can't bring myself to sell it after all this time and money invested.  As they say, as long as there's money, there's hope.  I also have hope that eventually suitable replacement transistors will be available.  I remember the way this went with tubes, which have never been more available than today.

Meanwhile, I have measured the distortion performance in both channels at 4 and 8 ohms, and the numbers are excellent in the good channel, at all power levels, basically at spect at 8 ohms and slightly worse than 8 ohm spec at 4 (which is not specified).  The numbers are not quite as good in the "bad" channel, but 0.04% THD at 8 ohms taint that bad.  I actually measured 0.09% THD at full power at 10kHz after a full warmup in that "bad" channel, beating the spec IIRC.  (Now I can't seem to warmup anymore...)

The Hafler might measure slightly better than the Krell good channel into 8 ohms, I hadn't measured 4, up to significantly above rated power.  It seemed to have peak output of about 60V, which surprised me.

The Aragon hasn't measured quite as well, however, I haven't done a full set of measurments only a quick check while tuning the bias, and distortion at final bias was never checked, and actually the final bias may not be the final final bias either.  Still, I got the Aragon down to 0.07% at 8 ohms at modest levels at 1khz, in bias fiddling a couple of years ago.

Tonight I was able to measure the Krell at 220 Hz at 8, 4, and 2 ohms, and unloaded.  (This is the "bad" right channel.)  I put the load on one set of outputs and a 6 foot cable to DVM on the other output.

1 watt nominal level into 8 ohms, about 2.88V

220 Hz

8 ohms  2.797V (warmer)

4 ohms  2.785V (coldest)

2 ohms  2.762V (cold)

noload   2.812-2.810(warmer)

DF(220hz,8ohms,1w) = Eunl / (Eunl-Eload) = 2.810 / (2.810-2.797) = 216
DF(220hz,4ohms,1w) = 2.812/(2.812 - 2.785) = 104
DF(220hz,2ohms,1w) = 2.811/(2.811 - 2.762) = 57.4

(Note: the DF should scale as half for each halving of load.  8 ohm damping factor is the one normally specified.)

The amp did not run long enough to actually get very warm, but even during the 10 minutes or so of operation there was improvement from 28.12 to 28.10 in the unloaded output voltage.  I could not go back and remeasure 2 and 4 ohms because the amp shut down twice when doing these measurements, and I didn't want to push it any more than that.

I was also able to measure 10kHz at 8 ohms:

10kHz

8 ohms 2.201

noload  2.213

DF(10kHz,8ohms,1w) = 2.213/(2.213/2.201) = 184

Very little change in damping factor from 220Hz to 10khz.

The apparent HF loss of -4.5dB compared with 220 Hz is probably mostly being caused by my crossover and EQ....more than I expected but it achieves the desired result here (flat response).  Also a dB or so HF loss in the Fluke 8060 meter I am using at 10khz.  I know from previous measurements, specs, and reports, that the Krell is virtually flat well beyond 10kHz, but that is not the question here, so I'm not going to turn off the crossover, bring out a better meter, etc., to measure this better. Turning off the crossover requires a reboot of the DEQ's and has messed me up in the past (when I forgot to turn it back on, etc).

****

In the "good" channel, that runs cooler with no suspicious hotspot and seemingly does not trigger shutdowns, I measured slightly better low frequency damping factor, but more than just slight improvement in the change in damping factor at 10kHz.

DF 8 ohms, 1w, 220 Hz, 222 (equivalent to 0.036 ohms)
DF 8 ohms, 1w, 880 Hz, 207 (equivalent to 0.039 ohms)
DF 8 ohms, 1w, 10kHz, 204 (equivalent to 0.039 ohms)

This time I normalized all measurements to 1 watt by adjusting the volume to 2.88v in all cases, overridding what are mainly affects from my crossover on the frequency response.

The improvement is consistent  with a theory that that feedback and servo are keeping the bad channel in line, but also reducing the ability to cope with changes in load at high frequencies, if only by a small degree, since it's "resources" are already being used to fight the faulty transistor(s).

Though perhaps there was some improvement due to change in measurement strategy.

And you might imagine if there really were a transistor in very bad shape, the result would be much worse.  Differences like this might occur from differential aging as well as manufacturing and repair histories and usage histories.

I was hoping to see a bigger difference, so that the good channel might have damping factor like 500 or something, and there being something "obviously" wrong with the bad channel, more than just a slight variation, even if that variation is consistent with the theory that the bad channel is slightly faulty.  The distortion measurements showed the same thing, with the good channel meeting spec and the bad channel being slightly off spec.

The damping factor for the FPB 300 is not specified, actually.

Unobtanium Transistors

The principal problem with maintaining amplifiers like my Krell FPB 300 is that the power output transistors, MJ1302 and MJ3281, have long been unavailable.

Motorola discontinued these transistors while Krell was still making amplifiers with them, so Krell bought the entire remaining inventory in the mid 1990's.  They used them in many different amplifier designs, and according to legend, Krell freely installed new ones during amp servicing.  That was wonderful while it lasted, until about 2009 or so.

Now, to keep the amplifiers running with authentic parts, it is necessary to scavenge transistors from parts unit amplifiers, similar amplifiers that might be even more valuable than the one you are trying to fix.  There is no public place where such scavenged transistors can be safely purchased, maybe you know someone you trust (and, do you really trust them).

MJ1302A and MJ3281A are sold on Ebay and through Asian distributors are almost certainly fake parts.  I bought some of these in the hopes they may be OK now that Chinese fabs have been making them for well over a decade.

Similar silicon is still available in the TO-247 case in the MJW1302AG by ON Semiconductor, the Motorola spin-off which still manufactures basic transistors for linear applications.

The problem is that the FPB amplifier heatsinks were designed for TO-3 transistors.  To use TO-247 would require complete re-engineering of the chassis and heatsinks.  One might as well buy a D'Agostino amp.

Meanwhile, ON Semiconductor does still make TO-3 power transistors, including the MJ15024G, which others now use in power amplifiers.

What defines Audiophile?

Struggle.  It's not High End audio if it doesn't take some work.  Contrary to what audio store mavens would tell you, you can't simply buy musical enjoyment.  It's not "all about the music."  It's about getting the sound that makes it sing, understanding how to get there, and traveling at least some small part on your own feet.  As Euclid said about Geometry, there is no royal road to being an Audiophile.  Not only is it not possible to get their without your own work (which could be just...actually serious listening...if you have totally obsequious dealer you trust and tireless elves to help daily) it's not the same thing.  To be an audiophile is to be endlessly deciding how to do audio, but hopefully not obsessing too much over these decisions, but hopefully making better ones as time goes on, that lead to more rather than less enjoyment overall.

To be an audiophile is to at least nearly always be on the edge of Audiophilia Nervosa.  That is the dark side, which we must stay as far away from as possible, I believe.  Some guidelines for avoiding that:

1) There is no one right way to do things, starting from the fact there is no one right set of goals to be aspiring to, including the nature and extent of one's unpaid personal labor and buying in audio, and when and what to listen to and how.

1a) The enjoyment people seem to be getting from their audio systems often seems inversely proportional to how good they are by various objective standards.  Why can't we who take audio seriously have it both ways?  Have high standards when we're tightening connections, and then, just enjoy?

1b) I believe 1a also includes that...it doesn't take much audio capability to make audio enjoyable.  The main thing is in fact...the frame of mind about it all.  To be able to accept things as they are now.

1c) Once one has banished unquestionably horrid sound, most audio systems are capable of communicating sufficient information for a fully expansive experience each time.  In fact, this doesn't take much.  The truly great systems have even more information to choose from, but both are sufficient.

So, avoiding obsession includes accepting the fact that the present will never meet the highest standard, or even sometimes ones that were recently being met, but still the envelope of expending experiences must continue.

Antenna Pointers

Ground Loop from Outside Antenna feeds (which themselves have proper outside bonded grounds and suppressors)

Despite ground bonding, I was getting significant ground loop hum from the tuner connected to my outside antenna.  I tried various ways to fix this.  Two things "worked":

1) Running antenna lead to F connector surge protector on my Monster power bar, then to tuner.  This reduced ground loop hum a lot, but not entirely.  I think having suppressor outside and inside as well on antenna feeds is good practice for safety reasons.  I did not notice signal deterioration.

2) Running antenna feed through "Magic" isolator.  I got Magic isolators for TV use almost 20 years ago.  I was worried they might not work for FM, but the case says nothing about having an FM trap, so I went ahead, and noticed no signal reduction.  In combination with #1, this eliminated ground loop hum entirely.  The 60 hz hum at the input to my Behringer DEQ 2496 connected to the Sub Out line signal from my Yahama THX receiver dropped to -140dB.   Before doing #1, the level was worse than -90dB.  As I did this after #1, I did not try it without #1.

Ground loop reduction became the order of the day when I started running my subwoofer signal through the Behringer DEQ.  Isolating the signal between the receiver and the DEQ with a Jensen Isomax transformer actually made matters worse.  The Behringer has 3 wire AC input whereas the Yamaha is only 2 wire.  The ground of the Yamaha is therefore determined by what it plugged into it, or not.

However, until I discovered antenna grounding and isolating, I was forced to use a Jensen Isomax transformer on the audio output of the tuner.  Listening to radio without that, hum was terrible, like -50dB, and other sources not much better than -60dB.  Once I applied the antenna grounding and isolating steps, the isolation transformer was not only no longer needed...it was introducing a little hum by itself on the radio signal and was better removed.

NB: Ground loops did not appear until I started using the DEQ because nothing else in my kitchen audio system is grounded.

Before doing any of this, I tried measuring the ground voltage and current running through the coax.  It was extremely tiny.  But that's all it takes to induce -60dB hum.  I decided in itself it was "safe," not representing a ground failure, simply a ground loop.

Folded Dipole Antennas made from Twin Lead must have Soldered Ends

My biggest discovery of the year, the lack of soldered ends explained recurrent issues I had been having since 2002 when I made a special antenna tuned for TV Channel 9 which I am still using.  In the end, the folded dipole with unsoldered ends became a Channel 9 blocker for other antennas, so I couldn't get Channel 9 on anything.  Soldering the ends fixed everything, and I get perfect reception now.  Bare copper is prone to getting semi-conducting oxidation which does very weird things.

UHF Antenna location is extremely important, down to millimeters.

As I hoped, I was able to find a location on the south side of the kitchen for a UHF TV antenna that picks up all local stations perfectly.

The problematic multipath-prone TV signals were coming from the south.   Previously I had both UHF and VHF antennas on the north wall, but the UHF signals for Channel 4 (which actually broadcasts on Channel 48) and Channel 5 (which uses another UHF channel) had multipath issues with recurrent pixelization multiple times per minute.

I first figured the south wall would be better because I could position the antenna above the bathroom mirror on the other side of the wall which was probably not helpful.  It would also then have a clear path to a southern originating UHF transmission, instead of being blocked by HVAC and other metal objects around kitchen.

But even on the south wall, the precise position proved to be extremely important.  Too far one way and I would get dropouts on channel 29, which comes from the north!  Too far the other way, and I was still getting multipath issues on channels 4 and 5.  There was a happy spot right in the middle so I stuck my Clearstream Eclipse antenna with "permanent" adhesive strip right there and enjoyed perfect UHF TV all night on local channels.

Overnight the Eclipse had unstuck from the wall and slipped down just over an inch.  The performance was now abysmal again.  I now stapled the antenna back into place.  But making some error, I stapled it about 2mm lower than before (fortunately, I had marked the original position).

This was not as good, I determined that afternoon.  I was not getting as many dropouts, but I was still getting them.

Back on the wall, I found that positions above the original position I had found were even better.  In fact, the signal strength and freedom from multipath kept increasing as I raised the antenna all the way to the ceiling, so I stapled it there.

Now reception is perfect on all local UHF channels, with no dropouts.