The test would be to send a noisy signal through the device and compare input to output. If the signal is unchanged and the noise is reduced, that would be interesting.
If you could try to replicate what they´ve done with the broadband noise source and 10 Mhz sine signal combined via a power combiner, it would be interesting.........
@ TNT,
Actually the "lab report" was done by an independent lab (independent means unrelated to the magazine and other companies).
The operator can should be qualified for doing such measurements, but - as said before - who knows? Obviously the guy is very much interested in "high end audio" and is - beside other projects like calibration units and emv test chambers (for example) - doing developments in this field.
@ jneutron,
Isn´t the comparison between cable direct connection and cable equipped with the "small elements" a comparison between "baseline" and "effect" ?
To be fair, we don´t know if he did repetitions or not.
The "Studio magazine" is not exactly a typical "high end audio mag" as they are doing only "studio equipment reviews/reports" aimed almost exclusively for the professional audio segment.
Of course, but reviews of this kind of products isn´t (afair) does not happen often in this mag.
Actually the "lab report" was done by an independent lab (independent means unrelated to the magazine and other companies).
The operator can should be qualified for doing such measurements, but - as said before - who knows? Obviously the guy is very much interested in "high end audio" and is - beside other projects like calibration units and emv test chambers (for example) - doing developments in this field.
@ jneutron,
Isn´t the comparison between cable direct connection and cable equipped with the "small elements" a comparison between "baseline" and "effect" ?
To be fair, we don´t know if he did repetitions or not.
The "Studio magazine" is not exactly a typical "high end audio mag" as they are doing only "studio equipment reviews/reports" aimed almost exclusively for the professional audio segment.
Of course, but reviews of this kind of products isn´t (afair) does not happen often in this mag.
Jakob, as I wrote just above, this is going to be a case where unless someone here does the measurements, and covers the bases needed by using a suite of precision low frequency and RF instruments in a shielded room (plus all the replications, resetting of the test setup to make sure it's not aerials injecting into the measurement, etc., etc.) there's a group that won't be satisfied. On the corollary, there's a group that say "I hear a difference and all your measurements be damned", which no amount of analysis will satisfy.
Just not worth the effort. For all I know, John likes to bring up Bybee's and hurl his regular insult that, "no one is interested in high end or improving the state of the art," solely to get at a few people's goats. Surely he's smart enough to know that it's not going to change opinions.
Just not worth the effort. For all I know, John likes to bring up Bybee's and hurl his regular insult that, "no one is interested in high end or improving the state of the art," solely to get at a few people's goats. Surely he's smart enough to know that it's not going to change opinions.
Without seeing the setup, we can't know if "direct" is the baseline. at 100 dB down, even wiring physical layout can have an effect.
For LOTO, we will check the meter against a known source of the correct voltage, then test the entity we want to look at, and then re-test the meter on a known source. That way we know that the meter works both before and after we checked the widgit. Otherwise, safety can be compromised.
Here, we have no way of knowing if the 1dB is test error, measurement setup, room air conditioning, whatever.. They don't mention going back to "baseline".
We agree, one result can actually be meaningless if it doesn't repeat.
I suspect that the lossy part they noticed and spoke of is probably what's going on. But again, without any details, nobody can validate..
jn
My experience of test labs is that the technicians follow recipes for standard measurements - but step off the script, or need understanding of the process - well, good luck!
JN, you are right, we should have a test point for easily measuring the output transistor power supply voltage. We normally don't put it in, because it really normally doesn't matter, as it is set by the power transformer output voltage.
Unfortunately, the Taiwanese engineers, trying for more power, raised the transformer ratio (without telling me) so I was put in a position of having too much voltage across the power supply caps in this present prototype, and I was curious as to how much. Well, we measured about 102V DC from the collector of one of the output devices, we have 24, and the probe slipped (typical) and we damaged at least one output transistor.
Of course, I threatened my tech, over lunch, with taking it out of his salary, not that I have not done the same thing a number of times in the past. I should have left the measurement for later when it was really necessary, because we could control it with the Variac that we placed in series with the amp, and we were going to adjust it down if the amp overheated with the optimum bias setting. We will, of course, in future, modify for less AC output voltage in the 2200W power transformer, in the final units.
Unfortunately, the Taiwanese engineers, trying for more power, raised the transformer ratio (without telling me) so I was put in a position of having too much voltage across the power supply caps in this present prototype, and I was curious as to how much. Well, we measured about 102V DC from the collector of one of the output devices, we have 24, and the probe slipped (typical) and we damaged at least one output transistor.
Of course, I threatened my tech, over lunch, with taking it out of his salary, not that I have not done the same thing a number of times in the past. I should have left the measurement for later when it was really necessary, because we could control it with the Variac that we placed in series with the amp, and we were going to adjust it down if the amp overheated with the optimum bias setting. We will, of course, in future, modify for less AC output voltage in the 2200W power transformer, in the final units.
Absolutely...
Wait....did we just agree on something???
With the secondary high resistance fb path, if sense opens, the system will continue to operate without blowing up the speaker. So the system would keep on working without any untoward events. It would be better to limp along with 5% distortion of the bass rather than total silence.
You of all people are quite in tune with the reliability requirements.
PMA: The change in feedback path does not change the amplifier requirements, the SPL will be unaffected by my scheme vs normal.
In normal case, the output voltage is raised to the level to meet the acoustic need, mine does the exact same.
In my scheme, if you use a crossover inductor in series, the amp will try to compensate around it to maintain flat acoustic output with zero phase change and in that case it will run out of compliance. However, the max levels of acoustic power available will remain the same with the same amp.
What I'm not sure about is how this scheme would deal with a multi order speaker. Direct radiator closed box, seems a no brainer. But, T-lines, reflex to a point, 4th and 6th order bandpass, all store energy in resonant or pseudo-resonant structures, so I'm not sure how well my scheme would work on those.
Subwoofers tend to be limited frequency bandwidth with lots of orders, so it might actually be that this dvc arrangement would be better with single driver or bi/tri amp systems with low order bass cabs.
I mean, in theory, this scheme will control the vc throughout it's working range. My prediction is that it will flatten the acoustic output response, flatten the phase, and lower distortion.. But it'd be nice to have that confirmed or shot down in flames...
jn
Last edited:
The test description and graphs provided in the article (Bybees) are unclear, that is why I repeatedly spoke about audiophile magazine publication. Independent lab does not help to publish exactness in the article text. This is very different from publication in a scientific magazine. I think Jakob often called for exact expressions, so I would wonder if he was satisfied with that article.
It is this point only where I have to respectfully but violently disagree with you.
The stuff my techs work on start as simple tubes worth maybe a grand or two, but as it builds up, they end up working on hardware worth 2, 3, or 4 million dollars. If my techs thought that in any way they could be penalized for a mistake, they would probably never go near the stuff.
I make it absolutely clear that I am the one with the bullseye on the shirt, not them. And, that if they make any mods on the fly, to let me know even after the fact, as I trust them totally, enough to take full responsibility..
A careless tech on the other hand, with a careless history, yah...yell at em..
jn
During WWII the Germans made synthetic fuel, having lost access to oil fields. This involved very high pressure of explosive gases. Accidents were too common. So over a break the guys in charge replaced the meter scales with ones that dropped a zero. They then told the workers they had modified the process so everything now worked at safer pressures being one tenth of what it used to be. Accidents went way down ascribed to lower worker stress.
Now I have had folks break brand new audio power amplifiers costing a decent bit. They tend to de-stress when the repair is to change a part that usually cost a few dollars.
Quite a bit different than when a unit fails under warranty and a trip to the site may cost a few days, airline ticket, hotel, car rental and meals to swap amplifiers. Cost of the repair part really becomes insignificant.
But yes I do tease the guys who break things occasionally. Fire them if it appears it might be a habit.
Now can we talk about acceptable error rate in say nuclear power plants? Atomic weapons safety precautions?
Now I have had folks break brand new audio power amplifiers costing a decent bit. They tend to de-stress when the repair is to change a part that usually cost a few dollars.
Quite a bit different than when a unit fails under warranty and a trip to the site may cost a few days, airline ticket, hotel, car rental and meals to swap amplifiers. Cost of the repair part really becomes insignificant.
But yes I do tease the guys who break things occasionally. Fire them if it appears it might be a habit.
Now can we talk about acceptable error rate in say nuclear power plants? Atomic weapons safety precautions?
After working in a cal lab, and being tested by our internal quality guy - an audit (passed), we are very good at following directions that differ from what we were used to. One of the tests was to use a procedure that differed from the industry norm. I followed the specified instructions instead of doing "what was right". I honestly don't think you would have any problem with a good calibration lab as long as you specified the tests to make and in which order.
Now, if you send something in without any special instructions, expect a report that agrees with convention as far as the tests that were run. You would still have to furnish the specs for that device / component in order that they can make sure the TUR was at least 4:1, and that they didn't waste time allowing something(and the instruments) to settle down for an extremely accurate reading when one isn't required.
-Chris
Now, if you send something in without any special instructions, expect a report that agrees with convention as far as the tests that were run. You would still have to furnish the specs for that device / component in order that they can make sure the TUR was at least 4:1, and that they didn't waste time allowing something(and the instruments) to settle down for an extremely accurate reading when one isn't required.
-Chris
I agree when I had reports the first thing was not who to blame but what to blame. So a layout person flips the ESD diode cell and shorts all the pins to the supplies. The layout review process, supposedly involving several top engineers from multiple disciplines was broken, fix it. This really happened, even twice.
- Status
- Not open for further replies.