I thought I should delete that last post, Trevor, but then I wouldn't be able to make this point.The off topic taunting and bickering going on on this thread is getting too much. Please restrain from making litigious assertions about each other and maintain an intelligent discussion on the topic at hand.
That goes for everyone on this thread!
I have never completely measured the JC-1, myself. I have 2 here, but I have better things to do than to precisely measure an existing design. However, the distortion at virtually any level can be approximated by simply looking at the measurements done by 'Stereophile' and extrapolating. One should presume that 3rd harmonic should reduce with 1/V (squared) and the 2nd will reduce directly with the inverse of the output voltage. Therefore, if you have a graph of the harmonics at 635W, it should be easy enough to estimate the distortion at any power level between idle and clipping.
Since the unit has a class A output of about 29W in this example, it can be presumed that the distortion will be even better than predicted in this extrapolation, below 14W into 4ohms. or 29W into 8 ohms and considered almost nothing, especially the higher orders above 4th.
Above the class A region the graph given by 'Stereophile' and 12.5 V into either 8 or 4 ohms over frequency should imply that 50 Hz might be slightly worse, but there is no specific reason for this as the thermal stability is higher than average, the heatsink thermal capacitance is high, and the Class A to Class B transition is carefully controlled, by the criteria established by Barney Oliver back in the early '70's. 'The Selected Papers of Barney Oliver' pp. 311-317
A good approximation of the difference between 4 and 8 ohms is about a factor of 2 in distortion.
Since the unit has a class A output of about 29W in this example, it can be presumed that the distortion will be even better than predicted in this extrapolation, below 14W into 4ohms. or 29W into 8 ohms and considered almost nothing, especially the higher orders above 4th.
Above the class A region the graph given by 'Stereophile' and 12.5 V into either 8 or 4 ohms over frequency should imply that 50 Hz might be slightly worse, but there is no specific reason for this as the thermal stability is higher than average, the heatsink thermal capacitance is high, and the Class A to Class B transition is carefully controlled, by the criteria established by Barney Oliver back in the early '70's. 'The Selected Papers of Barney Oliver' pp. 311-317
A good approximation of the difference between 4 and 8 ohms is about a factor of 2 in distortion.
Trevor White said:
Trevor,
To make the issue technical and not personal (this is a thread about Otala’s technical contributions after all), I feel the key issue rests at the results of 35yr ago. By that I mean an amplifier design philosophy of high open loop bandwidth and low feedback vs. “the op-amp on steroids” approach. If you look back many contributions by some smart people, like Bob Cordell, jcx, and andy_c to name a few, have basically been dismissed or ignored by some. I’m afraid things are moving toward a situation where even an inseparable amount of so called PIM is “bad” even in the presence of large amounts of other distortions. This will create a technical impass that will block further progress IMO.
scott wurcer said:
I agree with you Scott. I think when Home Theater took over and Stereo died in the proverbial then lots of great engineers got out of the game and development of stereo then stagnated. Now stagnation has been replaced with snake oil and the art hasn't really progressed since then. More's the pity 🙁
regards
Trev
Iain McNeill said:
The off topic taunting and bickering going on on this thread is getting too much. Please restrain from making litigious assertions about each other and maintain an intelligent discussion on the topic at hand.
That goes for everyone on this thread!
Hello Iain
You can delete this if you want but my point was that when you have vested interests in a certain commercial product then it is hard to be objective about it. Is that a fair assumption ??
regards
trev
My concern is that when behind a screen, people communicate without respect. Things are said for shock effect rather than to elicit knowledge. If we were all around a table, face-to-face, the discussion would be very different. This behavior, in my mind, devalues the fora.
In truth, there is 'Snake Oil' in the JC-1 and no 'Snake Oil' in the HCA 3500. Both are essentially the SAME circuit. Both have been reviewed by 'Stereophile'.
Only the JC-1 has ever been listed by 'Stereophile', and it has an 'A' rating and is used as a back-up and a 'reference comparison amp' by 'Stereophile' on frequent occasions. Why, and where is the difference that counts? Can you, Trevor, find it? Is it Voo Doo? No it is not, but I can't 'prove' why the differences make so much difference, subjectively.
Only the JC-1 has ever been listed by 'Stereophile', and it has an 'A' rating and is used as a back-up and a 'reference comparison amp' by 'Stereophile' on frequent occasions. Why, and where is the difference that counts? Can you, Trevor, find it? Is it Voo Doo? No it is not, but I can't 'prove' why the differences make so much difference, subjectively.
Those who never tried and never compared would also speak about snake oil. They would ask for evidence in common averaging measurements, like THD+N. They would say that high order harmonics at -110dB do not matter. They would ask for double blind tests. The last thing they made would be their own hard work, own comparisons, own ears and brains used for evaluation, and listening tests with invited audience, or lending components to someone for comparison.
john curl said:
I stand corrected on Matti's definition of DIM. In any case, you are exactly right in that DIM-30 and DIM 100 measure EVERYTHING. A big DIM number does not tell you if it is TIM or crossover distortion or any other HF distortion.
By the exact same token, TIM shows up clearly as IM sidebands in the 19+20 kHz CCIF test (the peak rate of change of voltage of that test is a bit smaller than DIM-30 for the same p-p amplitude, however).
Again by the same token, TIM will show up as increased THD-20.
Cheers,
Bob
Re: Re: Re: 50Hz THD
These are very good observations. I have studied and used DC servos quite a bit, and in some cases the amplitude coming out of the servo integrator at low frequencies under conditions of high power can get quite large; sometimes it can even clip. This is doubly true if the servo is given very little control range and at the same time is being called upon to correct a significan DC offset.
John seems to be pretty astute with DC servos, so I am a bit less inclined to believe this is going on in the JC-1.
If this was the source of the problem in an amplifier, one might expect that the resulting LF distortion might not be reduced much when the load was increased from 4 ohms to a very light load.
BTW, it follows that any amplifier that employs a DC servo should be tested very rigorously at very low frequencies (e.g., <5 Hz) at full-amplitude output (even if the amplifier is looking into a very light load).
Cheers,
Bob
Wavebourn said:
These are very good observations. I have studied and used DC servos quite a bit, and in some cases the amplitude coming out of the servo integrator at low frequencies under conditions of high power can get quite large; sometimes it can even clip. This is doubly true if the servo is given very little control range and at the same time is being called upon to correct a significan DC offset.
John seems to be pretty astute with DC servos, so I am a bit less inclined to believe this is going on in the JC-1.
If this was the source of the problem in an amplifier, one might expect that the resulting LF distortion might not be reduced much when the load was increased from 4 ohms to a very light load.
BTW, it follows that any amplifier that employs a DC servo should be tested very rigorously at very low frequencies (e.g., <5 Hz) at full-amplitude output (even if the amplifier is looking into a very light load).
Cheers,
Bob
Trevor White said:
The bitter truth is, manufacturing went abroad and we can't compete if to build here. The good truth is, home theaters did not replace stereo. And they use audio amplifiers as well, it is one more good truth. Far the more, I am currently working on the prototype of a microphone array with analog matrix and optical anti-clipping compressors. It is equally good both for stereo and for surround sound.
And the best truth is, discussing matters here on the forum we help more and more people to understand basics of audio design, in order to design better sounding gear knowing what to do, and how to design, instead of blindly following myths and other rules of thumbs that may work in some certain conditions, but don't work everywhere.
PMA said:
Stereophile did.
Just look at Figure 7 in the review, where THD 1 is shown as a function of power into different loads.
At 600W into 4 ohms, THD-1 for the JC-1 is only about 0.03%, about 10 dB less than at 50 Hz for the same power and load.
Once again, bear in mind that at these high power levels the test setup can sometimes come into the picture and give odd results. For example, a nonlinearity in the test load resistor or the connections to it can induce measured distortion, and it is not inconceivable that such measured distortion might be a function of frequency, especially if local thermals are involved.
Cheers,
Bob
Bob Cordell said:
Most probably you hit the nail, Bob! What did they use for 600W load, is the question. No wonder that light bulbs would show non-zero output resistance of an amp under the question on very low frequencies.
The designer of the amp neither agrees or disagrees with the opinions put forth on this thread by others. I just might have to go and measure it, but what a waste of time.
Bob Cordell said:
That's correct, metal screws, resistance (dirt) of connecting joints. Stereophile already experienced some problems in one of the Halcro measurements.
john curl said:
Surely you would want to measure your own equipment before you submitted it for review just in case it was faulty and/or the equipment reviewer made a mistake in their measurements ??
regards
trev
Bob Cordell said:
I don't think that is the problem Bob. The following are measurements done by Stereophile for a Boulder 2050 Monoblock amp.
http://stereophile.com/solidpoweramps/1004halcro/index5.html
J.A. says: "To facilitate the hooking-up of amplifiers in my test lab, I use 6' lengths of 14-gauge multistrand cable fitted with stackable dual banana (4mm) plugs and short adapter cables fitted with a dual banana plug at one end and, at the other, single bananas, spade lugs, or bare ends. It turned out that the connection of one of the dual bananas to one of the 6' cables was not as conductive as it should have been, due to oxidation of the cable conductors under the banana plug's grub screw. Mystery solved—though I don't have a clue why the nonlinearity manifested itself only at high currents."
J.A. says: "To facilitate the hooking-up of amplifiers in my test lab, I use 6' lengths of 14-gauge multistrand cable fitted with stackable dual banana (4mm) plugs and short adapter cables fitted with a dual banana plug at one end and, at the other, single bananas, spade lugs, or bare ends. It turned out that the connection of one of the dual bananas to one of the 6' cables was not as conductive as it should have been, due to oxidation of the cable conductors under the banana plug's grub screw. Mystery solved—though I don't have a clue why the nonlinearity manifested itself only at high currents."
john curl said:
I would expect power supply (transformer, cap values, etc. ) differences, is that the case?
Pete B.