Buckeye 3 Channel Purifi Amp Review

[Rick Sykora]

Don’t get me started on the audiophool myth that even-order distortion somehow magically sounds better than odd-order distortion.

The amount of distortion inflicted on the signal by the amplifier circuitry (whether it be a tube amp, Class A amp, or anything else) is almost never high enough in level to be audible, unless you are pushing an amplifier close to clipping. What the audiophile press calls the “warm” tube amp sound is largely a function of frequency roll off at the high and low extremes, which can be audible. But even-order harmonic distortion? Doubtful.

Maybe you should, but my distortion comment was done in the context of the the jest that preceded it.

As I stated previously, even with the steel part, the distortion is likely inaudible.

Rock guitarists drive amps to distort all the time and is considered audibly desirable. As for the even/odd aspect, would not want to condone audiophoolery, but in speaker measurements, we often consider HD2 more desirable than HD3. In any case, is off topic and subject for another thread.

 

[DonH56]

Off-topic: This thread goes into distortion a bit, including why we might perceive IMD as more objectionable than HD. Even-order HD generates IMD at frequencies lower and higher than the tones (f1-f2 and f1+f2) whilst odd-order HD generates IMD terms very close to the fundamental frequencies. I keep looking back at these posts, some now pushing 20 years old and originally pulled from posts far earlier, and thinking I should rewrite them and clean them up, but life is short and I am too lazy.

HTH - Don

Harmonic and Intermodulation Distortion

This post provides a quick comparison of how a distortion percentage relates to dynamic range in dB. That is, if an amplifier has 1% distortion, what does that mean? How soft (or loud) is 1%? Figure 1 shows a 1 kHz voltage signal with 1% second harmonic distortion (HD2) added. The spectral...

www.audiosciencereview.com

 

[pma]

Off-topic: This thread goes into distortion a bit, including why we might perceive IMD as more objectionable than HD. Even-order HD generates IMD at frequencies lower and higher than the tones (f1-f2 and f1+f2) whilst

That’s correct, in case that f1 and f2 are at the upper end of the audio band. But, in case of lower frequency of f1 and f2, like 400 and 500Hz, you get not only the 100Hz difference tone, but also for example 2200Hz as a 5th order component 3f1+2f2, the odd one. Or 900Hz as a 2nd order f1+f2. Thus, assumptions on pleasing even order distortion and annoying odd order distortion are incorrect.

 

[DonH56]

That’s correct, in case that f1 and f2 are at the upper end of the audio band. But, in case of lower frequency of f1 and f2, like 400 and 500Hz, you get not only the 100Hz difference tone, but also for example 2200Hz as a 5th order component 3f1+2f2, the odd one. Or 900Hz as a 2nd order f1+f2. Thus, assumptions on pleasing even order distortion and annoying odd order distortion are incorrect.

Yes, the gist of the earlier thread was that IMD (any order) is more objectionable than HD since the distortion creates "dissonance" tones. It sort of makes sense when only harmonic distortion is considered, as a triangle wave (even order harmonics) sounds somewhat more pleasant than a square wave (odd order harmonics), but all that goes out the window when you look at IMD and realize that any time you have HD and more than one tone you also get IMD.

 

[Billy Budapest]

Maybe you should, but my distortion comment was done in the context of the the jest that preceded it.

Got it!

 

[Billy Budapest]

Yes, the gist of the earlier thread was that IMD (any order) is more objectionable than HD since the distortion creates "dissonance" tones. It sort of makes sense when only harmonic distortion is considered, as a triangle wave (even order harmonics) sounds somewhat more pleasant than a square wave (odd order harmonics)

That assumes the distortion is even audible. And that’s why the “pleasing even order distortion” of tube amps (or any other component) is a red herring.

 

[Buckeye Amps]

All Purifi orders that are fulfilled starting today will have new tabs installed (along with all other non-Purifi orders as well obviously). Just an FYI for those reading who have pending orders.

 

[DonH56]

That assumes the distortion is even audible. And that’s why the “pleasing even order distortion” of tube amps (or any other component) is a red herring.

At some point it is always audible, but near clipping, so either you need a bigger (or better) amp or to turn down the volume. A long time ago I posted some test tones with varying amounts of distortion. For a single tone, 1% (-40 dB) was detectable by most folk, but it could take several percent using multitones. IMD was detectable at a somewhat lower level, perhaps 0.5%, but still... Almost nobody (may have been nobody, it was a while back and I do not recall) could detect 0.1% (-60 dB) HD though now and then someone would just break the error margin for IMD. This was through earphones or speakers which added their own distortion but still anything below -80 dB (0.01%) seems unlikely to cause problems. Always possible I and the other people who tried it have ears of clay, of course, that is the usual reaction (your hearing/DAC/speakers/whatever aren't "resolving" enough).

 

[pma]

0.1% H2 of 1kHz sine wave is audible with good headphones but the necessary condition is a proper volume level. It must be set below the point where the intrinsic ear nonlinear distortion prevails. Then, it is audible in the abx test.

 

[Billy Budapest]

At some point it is always audible, but near clipping,

I agree with that, but I am assuming “normal” listening levels where peaks do not go into clipping. I also think it’s far more likely that speaker drivers (especially at bass frequencies) will begin to distort at output levels lower than where the distortion associated with the amplifier becomes audible.

 

[Billy Budapest]

0.1% H2 of 1kHz sine wave is audible with good headphones but the necessary condition is a proper volume level. It must be set below the point where the intrinsic ear nonlinear distortion prevails. Then, it is audible in the abx test.

Another caveat—I am talking about audibility of distortion of speaker amplifiers. Headphones and headphone amps are a whole other world.

 

[DonH56]

I agree with that, but I am assuming “normal” listening levels where peaks do not go into clipping. I also think it’s far more likely that speaker drivers (especially at bass frequencies) will begin to distort at output levels lower than where the distortion associated with the amplifier becomes audible.

Agree except it depends upon the amplifier. A lot of AVRs only put out 100 W or less driving even just the front three, and with so many low-sensitivity speakers around clipping may be more common than I think (not something I have thought much about, though). I ditched a near-TOTL Sony AVR a few years ago because it was obviously clipping driving my Maggies, then found out in a review that the output dropped from rated 100+ W/ch to about 50 W/ch with three channels driven and 35 W/ch (!) if all five were driven, all into 8 ohms. It did not like my 4-ohm Maggies. I do not listen all that loudly, usually 65~70 dB or so at the MLP. It is doing OK upstairs driving somewhat more sensitive smaller Revels.

 

[Urvile]

Agree except it depends upon the amplifier. A lot of AVRs only put out 100 W or less driving even just the front three, and with so many low-sensitivity speakers around clipping may be more common than I think.

I whole heartedly agree -- I also think that manufactures are... less than honest? speaker manufactures about the power needs of their speakers, which compounds the problem. You have to look at recommended power range / Minimum Impedance / Sensitivity, to guide you to a realistic power need. And AVR manufactures who only give power ratings with one speaker driven at 8 Ohms..

 

[Feyire]

And the culprit is...

View attachment 263237

The binding post connection tab.

Unknown to @Buckeye Amps when he bought the binding posts, the tab is primarily made of steel. Steel is a relatively poor conductor and like a loose or oxidized connection can cause distortion. As with some others, I thought the culprit was the input board, but fortunately it was not. Honorable mention goes to @pma. Within his first few posts in this thread, he mentioned problems with ferrous materials in the signal path. We just had to find it.

Fantastic investigative work by all involved .

@Buckeye Amps , I'd perhaps suggest getting some small neodymium magnets, which you can use to easily verify whether or not any of the other materials used where the amplified audio signal passes through, is ferromagnetic or not. You'd be surprised for example, how common it is for the nuts on cheap bindings posts to just be gold covered steel. The same goes for a lot of speaker cable terminations - possible sources of added distortions everywhere!

As I've often shared in these forums, any ferromagnetic material in the amplified audio signal path will cause distortions, and these increase significantly with higher frequency due to hysteresis. This thread is a textbook example with empirical evidence.

If it was for a random guy saying that the binding post would make a difference lots of people here would put him in the snake oil territory.

That's very true. I'm glad that various professionals were here to assist in the diagnosis and also provide the before and after measurements.

 

[charlielaub]

Why use FASTONs at all? How about soldering the wires on to a binding post meant for that sort of connection? Like this:

 

[kevinsullivan]

Fantastic investigative work by all involved .

@Buckeye Amps , I'd perhaps suggest getting some small neodymium magnets, which you can use to easily verify whether or not any of the other materials used where the amplified audio signal passes through, is ferromagnetic or not. You'd be surprised for example, how common it is for the nuts on cheap bindings posts to just be gold covered steel. The same goes for a lot of speaker cable terminations - possible sources of added distortions everywhere!

As I've often shared in these forums, any ferromagnetic material in the amplified audio signal path will cause distortions, and these increase significantly with higher frequency due to hysteresis. This thread is a textbook example with empirical evidence.


That's very true. I'm glad that various professionals were here to assist in the diagnosis and also provide the before and after measurements.

It's great news if that's the correct and final diagnosis. A few pennies worth of parts to drop in sure beats having to manufacture new boards. In the end it will have been a scary but ultimately good lesson on the need for up-front assured adequacy of testing procedures. The magnet test is also a great and cheap solution to fix the gap in testing that seems to have produced this issue. Good news for Buckeye Amps, which also showed throughout a commitment to quality, transparency, and customer support and satisfaction. A great little engineering story to tell grandkids! (Phew!)

 

[Too Many Notes]

Why use FASTONs at all? How about soldering the wires on to a binding post meant for that sort of connection? Like this:

I am learning as I go here, so newbie question... Do fastons not make it easier for the end user to swap out a bad module for a replacement themselves, as opposed to needing to ship the unit back to Buckeye for repair? Personally, I have zero interest in learning how to solder. One of the reasons I went with Buckeye Amps is because these amps seem easier to self-service.

 

[Buckeye Amps]

I am learning as I go here, so newbie question... Do fastons not make it easier for the end user to swap out a bad module for a replacement themselves, as opposed to needing to ship the unit back to Buckeye for repair? Personally, I have zero interest in learning how to solder. One of the reasons I went with Buckeye Amps is because these amps seem easier to self-service.

This was one of the main reasons when I first set out designing the lineup back in 2020 I made sure no soldering was involved: the ability for the end user to either “repair/replace” parts easily if necessary. Along with the ability to upgrade if wanted (such as to a Rack case, etc) without needing to send back to me.

The Fastons are a great, secure connection. Closing in on 1000 total sales orders and never had one slip off during shipment or use. Aside from the steel tab issue, their performance is no different from soldering.

 

[restorer-john]

Closing in on 1000 total sales orders

What an awesome achievement. Congratulations.

 

[Buckeye Amps]

What an awesome achievement. Congratulations.

Thank you.

This never would have been possible without this community from the start. So a huge thank you to the ASR community.