Sound Quality Vs. Measurements

[bcarso]

Originally Posted by john curl

Tvr, this 120 is a sad sort of amp, and I am sure that even EB thinks so, compared to the others that he produced for Hafler, and his later power amp. It is like a 3 cylinder auto, when 4,6,8,or 12 cylinder autos are available. This is because of the input stage, not the output stage.

Finally seeing the schematic, that unballasted current mirror with the peculiar diode definitely got my attention There must have been a reason, although I guess EB has sworn off the forum.

[john curl]

If I make an 'unfair' criticism of EB, I hear from him, soon enough. So far, nothing.

[john curl]

For the record, EB and I have known each other for about 37 years. Back in 1975, we used to have lunch in Geneva, Suisse, to have lunch looking at the lake. There we discussed circuit design at a slightly higher level than we are discussing here. This is where 'breakthroughs' are made and exchanged, and the rest is history.

[BFNY]

Ah yes, watching "le jet" while eating the local delicacy, small fish from lake Geneva is a great experience. Last week was in Boulder sipping cask conditioned IPA at Whitney's discussing how to age moon rocks via laser excited MS-TOF and how to best implement averaging methods.

But about Hafler amps..

The best one before the transnova stuff was by far the XL-180, IMHO. The front end with dual complimentary cascoded 2sk/2sj Jfets and 6 2sk134/2sj49 mosfet outputs per channel was pretty decent for a $600. amp. It also had separate rectifiers and filter caps per channel, unlike earlier models, for a quasi dual mono supply.
To sound best, the outputs need to be biased hotter than spec'ed in the manual. With a decent efficiency speaker, when used in the class A range, the sound is pretty good.
I think Richard Marsh had something to do with the design, which makes sense based on his AA articles on cascoding JFET fronts ends from about that same time frame. And of course, the EB designs, which I also built and still own, had a very similar front end.
manual w/ schematic
http://www.hafler.com/techsupport/pd...80_amp_man.pdf
review
Hafler XL-280 power amplifier | Stereophile.com

[anatech]

Actually, capacitors do have mechanisms that do reduce performance over time. Even the manner of installation can have profound effects on serviceable life. One common example are radial capacitors installed without proper lead forming. This creates stress in the lead to rubber seal. It is telling that this issue has been addressed by sealing the bottom of the capacitor with what looks like epoxy. Those capacitors tend to survive the effects of time much better than untreated capacitors. The rubber is clearly the weak link in those capacitors

Look at carbon composition resistors. Aside from being noisy and being sensitive to temperature, moisture applied voltage, stress created by improper lead forming can act as a catalyst to normal aging effects. Just about any part you can think of has possible performance reduction due to time.

I do agree that some parts will check good compared to the normal performance for that part. However, the smaller the capacitor (in this case), the more quickly it loses fluid. ESR and DA almost certainly increase. So for small electrolytic caps, replacement is often a prudent move. The same thing can be said for older resistors where they are found near the beginning of a chain of amplifying stages and in the audio signal path. Of course, that constitutes a modification, or improvement. Most of the filter capacitors I check are performing well. I test them in circuit and under load. If they are pulled, I check them out of circuit.

I think that a very basic principle pretty much explains why older equipment needs work before any assessment begins. It is completely pointless to compare equipment with unknown defects. In fact, comparison can only be valid if the equipment in question is working it's original specifications. Wouldn't you agree (anyone?)

Hi tvrgeek,
Sometimes those older designs are still current. This happens more often than you might think. Older component design may not be a detriment. Take the case of the chassis mounted twist-lock capacitor. Those are a really good design. They dissipate a lot of heat safely into the chassis and also air. Too bad many equipment designers mount dropping resistors directly to the terminals. That heats the capacitor beyond where it would be on it's own. Great package though. Understand that old part design was carried out by real engineers who knew what they were doing. They were not unintelligent people at all. There are enough poorly designed new components around, so maybe throwing stones at glass houses isn't wise.

-Chris

[anatech]

Actually, capacitors do have mechanisms that do reduce performance over time. Even the manner of installation can have profound effects on serviceable life. One common example are radial capacitors installed without proper lead forming. This creates stress in the lead to rubber seal. It is telling that this issue has been addressed by sealing the bottom of the capacitor with what looks like epoxy. Those capacitors tend to survive the effects of time much better than untreated capacitors. The rubber is clearly the weak link in those capacitors

Look at carbon composition resistors. Aside from being noisy and being sensitive to temperature, moisture applied voltage, stress created by improper lead forming can act as a catalyst to normal aging effects. Just about any part you can think of has possible performance reduction due to time.

I do agree that some parts will check good compared to the normal performance for that part. However, the smaller the capacitor (in this case), the more quickly it loses fluid. ESR and DA almost certainly increase. So for small electrolytic caps, replacement is often a prudent move. The same thing can be said for older resistors where they are found near the beginning of a chain of amplifying stages and in the audio signal path. Of course, that constitutes a modification, or improvement. Most of the filter capacitors I check are performing well. I test them in circuit and under load. If they are pulled, I check them out of circuit.

I think that a very basic principle pretty much explains why older equipment needs work before any assessment begins. It is completely pointless to compare equipment with unknown defects. In fact, comparison can only be valid if the equipment in question is working it's original specifications. Wouldn't you agree (anyone?)

Hi tvrgeek,
Sometimes those older designs are still current. This happens more often than you might think. Older component design may not be a detriment. Take the case of the chassis mounted twist-lock capacitor. Those are a really good design. They dissipate a lot of heat safely into the chassis and also air. Too bad many equipment designers mount dropping resistors directly to the terminals. That heats the capacitor beyond where it would be on it's own. Great package though. Understand that old part design was carried out by real engineers who knew what they were doing. They were not unintelligent people at all. There are enough poorly designed new components around, so maybe throwing stones at glass houses isn't wise.

-Chris

[bcarso]

Originally Posted by john curl

For the record, EB and I have known each other for about 37 years. Back in 1975, we used to have lunch in Geneva, Suisse, to have lunch looking at the lake. There we discussed circuit design at a slightly higher level than we are discussing here. This is where 'breakthroughs' are made and exchanged, and the rest is history.

Sorry to indulge in off-topic, but if it please the Moderators:

I have great respect for Erno. Unfortunately, once a while back in a fit of general pique (those times that a good friend should advise you to sleep on something for a few days before release) I sent a letter to a magazine and criticized three articles in one issue all at once

One was a very nice article about JFET circuits from Erno, one a error-ridden piece by the editor, and one by Norm Thagard (who merely got a highpass filter number wrong). I wan't really that negative, but the EDitor printed the letter and didn't himself otherwise respond (and reprinted the same errors later), Norm responded diplomatically and we became thereafter friends, and Erno was a bit p-ohed. But I expressed genuine disappointment, because I had hoped to hear about some new parts or new topologies regarding JFETs. And as we all know, the economic exigencies have at best left us all with fewer and fewer good parts (despite the much-appreciated existence of Interfet and LIS).

So I apologize very belatedly.

One of my prized possessions is a map from about 1640 of Lac Léman.

Brad Wood

[bcarso]

Originally Posted by anatech

Actually, capacitors do have mechanisms that do reduce performance over time. Even the manner of installation can have profound effects on serviceable life. One common example are radial capacitors installed without proper lead forming. This creates stress in the lead to rubber seal. It is telling that this issue has been addressed by sealing the bottom of the capacitor with what looks like epoxy. Those capacitors tend to survive the effects of time much better than untreated capacitors. The rubber is clearly the weak link in those capacitors

Look at carbon composition resistors. Aside from being noisy and being sensitive to temperature, moisture applied voltage, stress created by improper lead forming can act as a catalyst to normal aging effects. Just about any part you can think of has possible performance reduction due to time.

I do agree that some parts will check good compared to the normal performance for that part. However, the smaller the capacitor (in this case), the more quickly it loses fluid. ESR and DA almost certainly increase. So for small electrolytic caps, replacement is often a prudent move. The same thing can be said for older resistors where they are found near the beginning of a chain of amplifying stages and in the audio signal path. Of course, that constitutes a modification, or improvement. Most of the filter capacitors I check are performing well. I test them in circuit and under load. If they are pulled, I check them out of circuit.

I think that a very basic principle pretty much explains why older equipment needs work before any assessment begins. It is completely pointless to compare equipment with unknown defects. In fact, comparison can only be valid if the equipment in question is working it's original specifications. Wouldn't you agree (anyone?)

...
-Chris

Someone recently posted in another thread about his NAD3020 developing hum. A few replies blamed the bulk caps, and the initiator said that they had visibly leaked. But from his description it sounded like the epoxy that the factory applied around the bases to affix the parts to the board. As I have one of those amps I had noticed this, and as well had had a recent failure with a similar manifestation, which turned out to be in the auxillary supplies, and was traced to a nearly open-circuited 'lytic for one of the aux rails.

No complaints about the 3020 though --- the best bargain I think I've ever gotten in an audio component. Thanks again to KOJ for the recommendation so many years ago!

Brad

[john curl]

I know how you feel, bcarso, I once criticized an EB design on the internet and got a frosty response.

[Wavebourn]

What's the point in criticizing someone's design if nobody asked?