|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
50.150.232.230
In Reply to: RE: I serious recommend you read up. posted by Ralph on September 18, 2015 at 09:34:45
OTL's are load intolerant , actually all amps to some degree are, so discussing them in a vacuum is partially irrelevant as none will achieve greatness if not matched to the correct driven load.Best 8 ohm amp design
Best 4 ohm amp design
Best 2 ohm amp design
and so it goes ....
Regards
Edits: 09/18/15Follow Ups:
In the case of all amplifiers, tube or transistor, the 4 ohm distortion spec will always be higher than that of the 8 or 16 ohm spec.
IOW, if you are really serious about getting the best out of your amplifier dollar, it will be best served by a speaker that is 8 ohms or more. Four ohms helps you get more power if your amp can double power into 4 ohms but that is not an argument for quality, its an argument for quantity and a poor one (3db) at best.
Of course, many people say something like ' I have four ohm speakers and they sound great', and they probably do, but if you could change them to 8 ohms without changing anything else, they would sound even better. Its not a good idea to make any amplifier work hard; the result will be harsher sound with less detail due to increased distortion.
FWIW, our M-60 amplifier makes 60 watts into 8 ohms. It does not do so well into 4 (although we have 3 larger amps that do) but if you just make sure that the speaker is 8 ohms in the bass the amp can do quite well. IOW its not really that hard to find speakers that work with it; its our most popular amp and we are still here after 39 years...
So now the Distortion spec counts for sound quality ..:)
The best systems i have ever heard , had multiple big amps and large multi driver speakers , of course nothing that could be driven with an OTL nor an SET amplifier and they were'nt anywhere near 16 ohms. Power is very much necessary for realism, regardless of sensitivity, do the math and the scale is the same low- high. Speaker size is also very necessary for realism, The percussive energy required to sound real (to replicate real instruments ) doesn't come from little speakers running on small amps.
So , yes distortion increases with low -Z, but how much is too much and only because the class-A bias and feedback used is diminished as you lower drive impedance , PSU noise not withstanding.
IMO,
It's Very important when pursuing SOTA dsigns to have a load tolerant Amplfier, Since the typical dynamic speaker is not a resistor and will have low-Z values below it's nominal, Which Brings us to your original Picked bone.
There is no such thing as a 16 ohm speaker sounding better, just because it happens to represent a Nominal impedance of 16 ohm, a multiple of said speaker operating at 4 ohm, will sound better if not amplfier limited, as the lower thd and increased sensitivity achieved by doubling Driver surface area, will overcome the ususal slight increase in THD from the amplfier, unless inadequate (Halcro) or not designed to do so (OTL).
Regards.
My speakers at home are not only 16 ohms, they are also 98 db 1 watt/1 meter.
I agree completely that the amplifier power is necessary for realism- I can get to 110-112 db in my room at the listening chair without clipping the amps. They make 140 watts into 8 ohms...
The speakers go to 20Hz and have no breakups in the audio passband thanks to careful crossover work and beryllium diaphragms in the midrange unit, which does the lion's share of the work.
I've yet to go to a show or a customer's house and hear a more convincing system, although as we both know that's no definitive measure. Regarding the distortion issue: when you run the distortion up on any amplifier by decreasing the load impedance, the distortion component will be of the most audible type (IOW higher orders), so yes to "So now the Distortion spec counts for sound quality ..:)", as the higher orders are readily audible even in trace amounts as brightness and harshness.
Regarding the 'no such thing' comment, actually there is a way to demonstrate that via the use a transformer to convert the load impedance. The fact that you went from 4 to 16 ohms is readily audible as a smoother sound with greater detail, this on a solid state amp that has no troubles doubling its power as the load is halved. The autoformer used for that test BTW has bandwidth to nearly 2MHz due to the turns ratio being so low. So it does not appear that the autoformer itself is the variable. In interviewing a number of manufacturers at shows, a common comment regarding this topic was (paraphrasing) 'yes, just because an amplifier can double its power is not the same as saying its sounding it best and in all likelihood it isn't'.
'cause that all sounds like more subjective opinion, not fact.
try it! you know you want to!
I've yet to see anyone that was objective- they all express subjective opinion sooner or later...
But actually its been well-known since the 1930s that the 7th harmonic even in trace amounts (0.01% for example) is quite audible and contributes to brightness and harshness. If this is news to you then you need to read up.
Our good friend John Curl has been trying to pound that into my thick skull. :) He's been a student of the study of distortion types, their audibility, causes and solutions for darn near his entire career, as far as I know.There isn't anything "magic" about the 7th harmonic, it's just that that's approximately where it starts to become noticeably dissonant and irritating. There are good reasons for this, as we all know.
:)
Edits: 09/24/15 09/24/15
and that is why it was detected as a harshness 'way back in the 1930s...
John Curl is spot on BTW...
Yes but the odd (and even) harmonics above 7th are all dissonant and increasingly so. To be avoided (at least at low power) if at all possible. You become more tolerant of increasing distortion and higher harmonics as the SPL goes up...this is why Cheever's metric is SPL dependent.
Well the class-D im playing with now is very tolerant at low levels , suprisingly so , makes you want to go up in levels and then headache reach for the volume at higher levels .
So no silver bullet yet .....
Is it an ncore? Or something "worse"?
I polled my friend now on his opinion from our Wilson/Devialet adventure and he too was underwhelmed. He said it was all head and no heart...it just didn't pull you in and make you want to listen in particular...a 50K background-while-you-work system...not hold your attention at all.
All WA speakers does that to me , exception in the past , was Watt/puppy on spectral Gear. Most recently i have heard them on Dan D'Agostino's new amp and VTL, both very under whelming, not my sound.
Edits: 09/30/15
The original X1-Grand SLAMM with KR Audio was pretty awesome...I repeated this awesomeness with an X1-MK3 in a shop in Switzerland.
The WP7 sounded quite good with Lamm ML1.1 monos as well and not bad with a big VTL rig...all other demos I have heard were far from great.
I must amend my earlier post in this sub-thread, regarding the 7th harmonic. While in the midst of creating a table which would show why natural harmonics don't line up with equal temperament tuning, somebody wisely sent me a table which had already been created (Thank you!). In fact, although many harmonics are dissonant with equal temperament, the 7th really sticks out like a sore thumb, so, there IS something magic about it.
You can view the table here: http://www.phy.mtu.edu/~suits/badnote.html
But be careful - you could end up spending the next hour or two reading all of the linked pages!
:)
Thanks for the info! I will look into your link.
Very few amplifiers today have the PSU to do the low-Z dance, so I'm not surprised by your comment.
Our MA-1 power supply can produce about 50Amps using the same measure that many solid state amplifier manufacturers use.
The issue here is finesse. If you want to create realism, the last thing you want to do is burden the amp in such a way that it makes more distortion because when it does so, its the kind that is audible.
I think you said it yourself- you have enough power so that there is plenty of power for realism. But you might consider what it is that causes realism in the first place. One thing certainly is a lack of distortion (and by this I am making it clear that I don't like presentations that exhibit a large amount of the 2nd order; while they certainly sound sweet, its hard to regard such as neutral).
But in addition to that power, the distortion components that the ear finds as the most audible (IM, and the higher ordered harmonics) will have to be kept to a minimum. That's hard to do when you have a lower impedance. In a solid state amplifier, the output devices have a non-linear capacitive aspect built into the junctions. This capacitance is magnified by current (BTW, the FM radio in your car is tuned by a device called a varactor diode that takes advantage of this principle). If you want to avoid the non-linear capacitive effects on the circuit, its to your advantage to keep the load impedance higher.
Now you can see this sensitivity to load in the specs of any amplifier, even class D.
Another reason 4 ohms is a hard way to go is the speaker cables. Now I am old enough to remember how I was able to go to Ace hardware and get zip cable, and it worked. But back in those days a lot of speakers were higher impedance. If you have a 4 ohm load the cable is just going to be a lot more critical plain and simple. As such it will induce a coloration, so you inherently have to keep it short. This again is easy to demonstrate. RCA published a nomograph showing just how important the speaker cables were back in the 1960s. Essentially, the lower the speaker load impedance, the more the cable works to reduce the damping factor. At 16 ohms its almost a non-issue.
try it! you know you want to!
-
"But in addition to that power, the distortion components that the ear finds as the most audible (IM, and the higher ordered harmonics) will have to be kept to a minimum. That's hard to do when you have a lower impedance. In a solid state amplifier, the output devices have a non-linear capacitive aspect built into the junctions. This capacitance is magnified by current (BTW, the FM radio in your car is tuned by a device called a varactor diode that takes advantage of this principle). If you want to avoid the non-linear capacitive effects on the circuit, its to your advantage to keep the load impedance higher.
Now you can see this sensitivity to load in the specs of any amplifier, even class D."
So I showed you an example of a class d amp that is not sensitive to load.
Similarly, the IM it produces is below audibility.
Some call this "sterile". I call it accurate.
try it! you know you want to!
The sensitivity to load is just as you quoted me above- the lower the load impedance, the higher the distortion, even in class D amps.
Maybe you should check the specs of the ncore because in some regions, lower impedance loads actually produce less distortion....
try it! you know you want to!
Holy smokes I was just joking about you having 85db speakers, as it turns out they are ~83 db? Wow, I didn't know they made them that inefficient.
Looks like you are stuck with the pucks for the time being, no soup for you!
It says you have a 6-pack of pucks? Is that for surround or are you multi-amping those gluttons?
△ᴉʇɐuᴉɯnllI oᴉpn∀△
RalphI think that most of SS/BJT based amplifiers benefits from 16-ohm load(speaker) vs 4-ohm LS primary from reduced beta drop effect ,
MOS-FET-s ,laterals ,verticals and other FET types don`t suffer from this negative effect , but from high input capacitance which is again highly non linear , and this can be `cured` only with higher standing bias current( A class ) .
Further general problem of SS devices is that N and P complementary power devices is not so complementary as is advertised ,
for example power MOS-FET N & P input capacitances , it is so different , same effect is valid for BJT N & P power devices ,
transfer characteristic is different and not so complementary,
to avoid this N vs P type basic differences some SS amps designers use only same type of power devices , usually only N type connected in SEPP power bridge configuration ,and rarely connected in Circlotron PP power bridge ,
for example somewhere above mentioned Einstein hybrid power amp use Circlotron PP (a)/B class configuration ,
except of this couple examples exist so many others real basic problems related to implementation of SS devices for good audio power amps .
__
"Art which does not have the appearance of art is true art."
- Old Roman saying -
Edits: 09/18/15
In a solid state amplifier, the output devices have a non-linear capacitive aspect built into the junctions. This capacitance is magnified by current
- Ralph
You talking BJT and or MOSFETS, High bias Mosfets dont have this issue , just crank the bias up to reduce. Seriously Ralph, junction capacitance is easily controlled. Tube amplifiers also suffer from parasitic capacitance like anything else.As to speaker wires, we have progressed beyond zip cable at Lafayette, RLC circuits (speaker wires) are available to care of any situation, speakers wires should be short anyway, .5M is about the best , 1M is the max regardless if 8/4/2 ohms , not to mention most good SS amps have pretty low output impedances ....
Regards
Edits: 09/18/15 09/18/15
I guess I'm not understanding why you would want more coloration out of the amp (and distortion of any kind is interpreted by the ear as a coloration....).
you should have seen the PSU on my NAT...20Kg power transformer, 100K uF capacitance, 4 chokes...it is a pretty manly PSU for a 100 watt amp (ok, its class A).
THere is more to sound quality than scale you know...
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: