Entech
@ Bob Cordell
I found one of those devices that RNMarsh mentioned here - Entech: Pro Audio Equipment | eBay - Entech Wideband Powerline EMI Noise Analyzer Monster Power Dr Power | eBay = $59.00
I don't know how much they are/were new, but that seems like a good deal to me
Whether or not the "Monster" is the same as in Monster Cables etc ? If so, "possibly" RNMarsh could have designed it ?
If you buy it, it'd be nice to hear your thoughts etc on it
@ Bob Cordell
I found one of those devices that RNMarsh mentioned here - Entech: Pro Audio Equipment | eBay - Entech Wideband Powerline EMI Noise Analyzer Monster Power Dr Power | eBay = $59.00
I don't know how much they are/were new, but that seems like a good deal to me
Whether or not the "Monster" is the same as in Monster Cables etc ? If so, "possibly" RNMarsh could have designed it ?
If you buy it, it'd be nice to hear your thoughts etc on it
I love the 'healthy' definition. So accurate.
And i think of some measurements of capacitors and resistances distortions, as an example, made because some people were pretending to hear differences between different components, who had revealed behaviors we don't knew (or be aware of) before.
(I believe Richard Marsh was part of this quest).
As an example of 'healthy', i think to the 'cables' sound story ;-)
We can hear differences, most of the time. Not healthy is to presume they are due to the metal or the isolating material (and some virgins). Healthy is to understand lumped elements and impedance variation of the charge with frequencies in order to understand the phenomena the right way.
But, for sure, when we understand the real reasons, it is easier to address them the right way than to try various combinations blind.
The problem with the subjectivism's pinch of salt is it requires some modesty from the objectivists we all were at the beginning of our professional life ;-)
Would an older example of that be crossover distortion in the early transistor amps?
That is, to preserve their longevity, some had woefully underbiased output stages. Depending upon the measurement that you took, they still looked "good", but they didn't sound good.
Once people figured out what was going on, it was fairly straightforward to measure distortion and harmonics at the appropriate "first watt" kind of amplitudes, and pinpoint the problem.
So, might this be a case where ears led us to an objective test?
That is, to preserve their longevity, some had woefully underbiased output stages. Depending upon the measurement that you took, they still looked "good", but they didn't sound good.
Once people figured out what was going on, it was fairly straightforward to measure distortion and harmonics at the appropriate "first watt" kind of amplitudes, and pinpoint the problem.
So, might this be a case where ears led us to an objective test?
proabably not - thd vs output power sweeps would have been in any competent engineer's tool kit long before - push-pull tube amps need bias too, can exhibit crossover distortion
more likely is that marketing siezed on a number that could be gamed to look good vs tube amps % full power distortion # in glossy ad copy
more likely is that marketing siezed on a number that could be gamed to look good vs tube amps % full power distortion # in glossy ad copy
maybe yes, also...
remember the times...people expected that the single number THD would rise at low levels, as it typically was a THD+N measurement. It's only if you looked with a spectrum analyzer that you'd see the rise in THD+N wasn't just from the hum and noise in the signal, but from many harmonics in the presence of crossover distortion.
Perhaps it hinges on how commonly available spectrum analyzers were to the audio design community in the early 60s?
remember the times...people expected that the single number THD would rise at low levels, as it typically was a THD+N measurement. It's only if you looked with a spectrum analyzer that you'd see the rise in THD+N wasn't just from the hum and noise in the signal, but from many harmonics in the presence of crossover distortion.
Perhaps it hinges on how commonly available spectrum analyzers were to the audio design community in the early 60s?
many audio tube power amps of the 50's, 60's were push-pull - I maintained one that was a legacy in a study space I frequented in college, rebiasing, replacing tubes as they aged - there was plenty of opportunity to experience crossover distortion with tube amps
the cult of the SET is a modern audiophile phenom: Valve Amps
the boys at Bell Labs understood harmonic multiplication with negative feedback, IMD ect. - hundreds of audio band amps in series as repeaters were required for the 1st few generations of long distance service, later even better performance was demanded over wider frequency as FDM multiplied channels and made IMD a bigger signal integrity threat - almost all audiophile concerns in feedback amplifiers were identified, effectively addressed by their engineering teams well before the 60's
if you want to know a big part of why all those dielectrics, constructions of caps were there for Marsh to make up his smorgasboard, look to Ma Bell's requirements
the cult of the SET is a modern audiophile phenom: Valve Amps
the boys at Bell Labs understood harmonic multiplication with negative feedback, IMD ect. - hundreds of audio band amps in series as repeaters were required for the 1st few generations of long distance service, later even better performance was demanded over wider frequency as FDM multiplied channels and made IMD a bigger signal integrity threat - almost all audiophile concerns in feedback amplifiers were identified, effectively addressed by their engineering teams well before the 60's
if you want to know a big part of why all those dielectrics, constructions of caps were there for Marsh to make up his smorgasboard, look to Ma Bell's requirements
These are all good points. We also must always be aware of "different" versus better or worse. A tone control makes the sound different, and this no-longer-neutral preamplifier may sound better or worse to some.
Regarding speaker cables, some are probably snake oil, but some are probably based on science, especially when certain interactions between certain cables and amplifiers may be involved. This can be true both for straight-wire speaker cables (the vast majority) and speaker cables that employ some passive components, such as Zobel networks in the line or at the end of the line. I believe that the MIT cables, for example, include a Zobel near the speaker end and a 2uH series inductor (this is where John Curl pukes) somewhere along the line. Whether this makes the MIT cables better or worse, and under what conditions, is unclear.
Cheers,
Bob
It is possible that the SQ effects of clipping were not looked at closely enough until subjective observations were made. This might also be the case with overly-aggressive protection circuits used frequently in the days when output transistor SOA was rather limited.
Cheers,
Bob
Practically all THD measurements were THD+N in dem days but the values achieved were far from the noise floor.
The Sugden articles http://www.audiomisc.co.uk/HFN/Sugden/Sugden.html#_N1were significant on several fronts.
- one of the first to investigate xover dist. & its effect on sound
- use of a log X-axis to show THD with power level. Practically all reviewers used a linear scale which led to a graph like Fig 1 of the class problem at the top of the page under the pic of the Richard Allan amp.
- probably the first to advocate Class A operation of transistors for better sound.
The SOTA THD instruments of the day were the Radford LDO & THD meter. Sugden's, who were originally instrument makers, designed & sold a somewhat cheaper but similar test set to help people design amps properly.
Wave Analyzers were VERY expensive and probably limited to labs doing Defense work.
But looking at the residual from a THD instrument, its obvious when you have xover. These are the 20kHz waveforms I show in http://www.diyaudio.com/forums/solid-state/235188-tpc-vs-tmc-vs-pure-cherry.html
Bob, I wonder if there would be space in your new edition for a chapter asking
What if the MOST IMPORTANT 'measurements' used in the design of amps were DBLTs?
Such amps would be very different from present amps but would sound better.
I can name a number of features from such an exercsie which would fill both the pseudo Golden Pinnae AND the objective merchants with horror
- severely limited bandwidth - eg brickwall at 20kHz or even lower
- probably soft clipping .. perhaps Klever Klipper
Note I said MOST IMPORTANT and not ONLY. There's no reason why the amp shouldn't have reasonable eg <0.01% THD20k (or rather its IMD equivalent) or very high slew rate. (some thought needed to combine this with Klever Klipper) But it would also be
- very simple cos 1ppm THD & zillion V/us slew weren't primary objectives
- unconditionally stable with load unlike many (most?) Golden Pinnae amps
- Serious attention to good overload & recovery behaviour
- probably an intelligent PSU to help the above
- be very large cos we dun really wun it to clip, current limit or overload though we provide all dis bandaids.
Contrary to both da deaf Golden Pinnae AND da objectivists, there HAS been a fair amount of DBLTs on important amp 'features'. Its just that the results go so far against the ingrained prejudices of both sets of fanatics that these tests are immediately forgotten.
I successfully used DBLTs as the most important speaker 'measurement' for nearly 2 decades. I don't see why we shouldn't do the same for amps.
But if you think the latest AP is expensive, just wait till you try & do DBLTs properly.
Last edited:
I dont understand if someone want to prove high slew rate is not important on an amplifier why he/she do Double Blind Test with unlimited input bandwidth compare to limited input bandwidth using SAME amplifier?
He test an amplifier input or an amplifier output?
I think we must explore all possibility. If we sure about something and want to prove it, than the test is bias.
He test an amplifier input or an amplifier output?
I think we must explore all possibility. If we sure about something and want to prove it, than the test is bias.
the usual "problem" with DBT test is that very few of audio guru's verities can be positively identified
but a handful of results from well estabilished psychoacoustics lets you predict with fair certainty that many of the described circuit effects on audio signals will not be heard
in DBT
but a handful of results from well estabilished psychoacoustics lets you predict with fair certainty that many of the described circuit effects on audio signals will not be heard
in DBT
I'm not interested da DBLT that 'prove' certain 'features' are unimportant or inaudible. Let da Golden Pinnae exhaust these topics ad nauseum while proving to all & sundry that they are deaf.
I'm ONLY interested in the features that DBLT show ARE audible and important.
YES, DBLTs have shown some differences are audible and there are definite preferences.
My 'features' list are those with the clearest and most unambiguous preferences in DBLTs, both in this and the last Millenium.
These preferences are of course taken only from true golden pinnae who give consistent results in DBLTs.
For the rest, if you can't hear the difference, why should you be bothered if I limit the bandwidth to 20kHz?
I'm ONLY interested in the features that DBLT show ARE audible and important.
YES, DBLTs have shown some differences are audible and there are definite preferences.
My 'features' list are those with the clearest and most unambiguous preferences in DBLTs, both in this and the last Millenium.
These preferences are of course taken only from true golden pinnae who give consistent results in DBLTs.
For the rest, if you can't hear the difference, why should you be bothered if I limit the bandwidth to 20kHz?
Last edited:
I agree if you compare the input limited bandwidth and unlimited bandwidth. But you must compare :
1. using SAME amplifier with low slew rate
2. using SAME amplifier with high slew rate
3. using different amplifier with one have low slew rate and other have high slew rate
I want to know if slew rate limiting is not generate another kind of distortion.
I am able to notice the different amp with low slew rate dan high slew rate with input limited bandwidth. Of course I am not sure the different is cause by slew rate or not. I appreciate if someone do double blind test to confirm it.
It only makes sense to do 2.
My list only includes Bandwidth Limitation cos this has been tested multiple times with the same results each time.
AFAIK, zillion V/us slew hasn't been tested to the same rigor so I only mention that an amp designed using DBLTs as its MAIN metric could easily have supa slew too .. eg if the base amp was a CFA .. which would meet some of the other criteria like simplicity.
Brilliant remark.
In this course, it helps, working on an amplifier to know both the source and the transducer. And to know what we are looking for (sound stage, little details reproduction, or/and spacialisation etc.)
And to know if, working on a part of a chain, we are not compensating some defect around, improving the whole chain, but introducing a defect in an other environment.
What helped me must to improve my electronic designs were my sound engineer experience. And vice versa. Both for the knowledge of the sources and the complex electronic assemblies.
Jan, you asked examples, i gave one (caps and resistance qualities). We take great care, now, to chose the best and oversized feedback resistances, to chose (and listen) carefully the reservoir capacitances devices...
I can give-you a lot of other examples. Zobel and 'motional' compensation of speakers, to get a flat impedancese curve... Symmetrical design of speaker's passive filters... Curent-feedback vs voltage feedback ;-)... Sense of the power plugs to minimize leakeages... isolation of ground returns from shields in cabling (pseudo symetrical wiring)...SMPS vs linear ... C core transformers VS toroidal...So many details in digital parts designs(re-clocking, isolation of rails)... chose of OPAs... care to printed board design (where you take you feedback point as an example).
A very interesting example, on my point of view, is all you can do on a DCX2496 to change a poor sounding device to a very transparent and dynamic device for few bucks, with not so much differences in measurements.
An other example: i prefer (with grounding scheme care) a unique power supply to dual mono when it is about spacial stability. I can explain why.
Near all those can be measured (or explained) and make a difference. But sometimes, it makes a difference you cannot measure, but you can hear. No smake oil, no virgins.
All the tricks i believe all designers use without noticing anymore, kind of reflex, witch were learned by listening experiences.
I wonder how some people can design amplifiers with simulators and build-them as it.
THD being an RMS-type of number, even isolating THD from N still wouldn't show up xover distortion. You could only see xover looking at an analyzer's residual on a scope or with an FFT analyzer.
All this assuming of course that not-too-drastic xover distortion can be heard by the rank-and-file audiophile...
Jan
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.