Testing to find all the Acoustat intermittent shorting faults continues. This now 6 month process has often revealed my tendency to jump to solid beliefs without sufficient evidence, as one belief after another was demolished by further testing.
There is no evidence the Krell FPB 300 amplifier, fully serviced this year, is at fault, except for apparently detecting a highly intermittent "short" problem which simpler amplifiers, like the Aragon and the Transnova (my Hafler 9300) brush off as long as they don't persist too long. It plays another low efficiency speaker set, the Revel M20's, at as high output as I can tolerate. It just doesn't tolerate shorted outputs for more than microseconds, which is apparently what the Acoustats are/were doing.
The Krell now also seems to play the Left (modified) Acoustat speaker OK. I tested that for 3 weeks including the FM tuner (which seems to cause more shutdowns than anything else) without shutdown, now that I have reset the LF transformer to the correct "Red" wiring which has the highest step-up ratio and matching for a 2 panel system. That seems to have fixed or at least bypassed the fault, which also appears to have no fault in its HF circuit (much to my surprise, as I had long figured the problem to be intermittant contacts in the HF variable resistor...but those are now very well cleaned and treated with Deoxit). The Left HF circuit by itself did not fail in weeks of testing. Then, I disconnected the HF and reconnected the LF correctly to the Red wire. That did not fail in 3 weeks of testing, so I then tested the speaker full range for a week, and it did not exhibit any problem. However, as I describe below, there has been one failure which *might* have been caused by the Left speaker...there is still room for doubt...in fact any test of an intermittant problem in a limited time span could be wrong...there is simply no absolutely perfect test as long as the problem is intermittant, and the ones I've been looking at now have become the kind of thing which *might* happen once a month, making it very difficult to track down.
The Right Acoustat seems to have a problem in the HF circuit. I'm unsure if there is a problem in it's LF circuit, which I might have also changed (since I didn't write down what it was before disconnection). I've had the HF circuit disconnected for more than a month now just to determine if there is ALSO a problem with the LF circuit.
Setup like that, I have had exactly one shutdown when playing both the R and L speakers, but with only the bass transformer connected on the Right side. So does that indicate a fault with the L or R ??? Two weeks of followup testing on the Right by itself have not revealed a further issue. If none appears for another week or so, I'm going to write off that incident as a freak event, perhaps even power-related.
Meanwhile, I tried a super high power test, which yielded further ambiguous results, possibly caused damage.
I have a polarity test app which came with a wave file featuring a full level impulse sustained to a quite low frequency, possibly below 1 Hz. And it's a positive only impulse. This is a potentially destructive signal. And it repeats many times per minute on the wav file which lasts for 30 minutes.
But I can play it continuously for 30 minutes or more on both Acoustats, at 0dB (DAC is set to output 2V, which means the 600W potential power at 2.35V is cut to 255W peak rms), with the 100Hz crossover.
The crossover is very important. If I disable the 24dB/octave crossover at 100W, the LF transformers heat up greatly. I never see that heating up anymore with the 100 Hz crossover.
I ran full power (255W peak impulses with extended LF) into the right Acoustat for over 30 minutes. I did not notice the heating up. Then the fuse blew. Then I discovered the interface box had heated up to 120F right above the LF transformer.
I have not repeated that test. On the right speaker, with both transformers now connected, the 3A fuse blew immediately when switching in the signal. So, technically, this signal is not in the safe range of operation.
I'm suspecting the fuse blew immediately on the right side because both transformers were connected, creating a much greater HF load. However, it could also be because...the R channel transformer still has an operational fault, even using the correct wiring. If so, I'm not even sure how important that operational fault is, considering it's out of the safe operating range anyway.
Subsequent to testing both speakers to the full range signal, now known to be out-of-safe-range, I tested both with the 100 Hz crossover re-engaged, for at least 30 minutes, without failure (and the interface box seems to have negligible thermal rise, 2 degrees above ambient or less). This actually applies the same peak voltage, equivalent to 255W, into the speakers. And I have played the speakers for awhile. So it does not appear I actually damaged them with this test. But it raises further doubt to the belief that when the recent amplifier shutdown playing both speakers occurred, it was not because of the right LF transformer but the left.
BTW I've also shown that the Behringer DEQ does not switch out LC filters when you turn them off in the PEQ menu. You have to restart the DEQ to get this change (potentially unsafe...) to take effect. However, for a quick and possibly destructive test, the PEQ's can be completely bypassed in the bypass menu, and that takes immediate effect).
Engaging LC filters, however, does seem to take immediate effect.
******
But setting aside all the endless mind-numbing testing with ambiguous results, always incomplete, and on to what it has also brought to attention.
The Acoustat interface system is ingenius, but unlike the Transnova amplifier I don't see why it need to be considered a perfect design.
Listing to just the LF transformer alone, correctly set to the Red winding, yields fabulous sound. Perhaps better in some way than with both transformers connected. It has a greater coherent wholeness, and the speaker sounds far less up-tilted.
As I write this, however, I haven't spent much time listening to the newly corrected Right speaker, which now has greater output from the LF transformer with the Red connection. That was a huge and unquestionable improvement, perhaps the biggest upgrade in speaker performance since I got the acoustats. Never mind all the amplifier changes, etc, running the speaker on the correct transformer winding is huge, and huger than huge when it's all part of a dual frequency band network, as it off balances the two parts.
In fact, I could say the entire "lost 6 months" so far, since a speaker damaging accident was rewarded by this one discovery, so important is it, that I had the speaker(s) on the wrong LF transformer winding.
But now that I've measured the fairly decent HF performance of the LF transformer by itself...essentially flat to 1Khz, it makes little sense to me why the HF transformer should be crossed in starting around 350Hz. And I'm thinking the 1-transformer impulse looks nicer.
But these are also things I need to test better.
I've enjoyed thinking about the possibility of bi-amplifying the acoustat interface LF and HF circuits, but it would probably not be safe, as both amplifiers would share the same ultimate load on the secondary side of the transformers.
Still, I like the idea of removing the HF network, or at least the capacitor part, from the box for better free wheeling experimentation. I'd be interested in only rolling in the HF transformer above 7kHz, or finding some way to get phase perfect impulses.
But the issue might not even be in the dual transformer arrangement either, but in the very idea of a need for extended high frequency power response. Perhaps rolled off highs have a way of just sounding better. Lots of classic systems have that sound. Maybe this is even especially true for the FM source I listen to nearly always in background, and maybe less true for more 'clean' sources.
So perhaps the "solution" could just as efficiently provided with digital EQ, applied to the sources that need it.
Anyway, the speakers own internal engineering is now exposed, it can be questioned with and tinkered with, and great impacts and questions arise. Differences that hugely affect measurement and acoustic performance. Much greater than the imaginary wiring and amplifier differences.
A friend thinks I'm beginning to question the virtue of 'flat frequency response' above other things, which is good, to his mind.
There is no evidence the Krell FPB 300 amplifier, fully serviced this year, is at fault, except for apparently detecting a highly intermittent "short" problem which simpler amplifiers, like the Aragon and the Transnova (my Hafler 9300) brush off as long as they don't persist too long. It plays another low efficiency speaker set, the Revel M20's, at as high output as I can tolerate. It just doesn't tolerate shorted outputs for more than microseconds, which is apparently what the Acoustats are/were doing.
The Krell now also seems to play the Left (modified) Acoustat speaker OK. I tested that for 3 weeks including the FM tuner (which seems to cause more shutdowns than anything else) without shutdown, now that I have reset the LF transformer to the correct "Red" wiring which has the highest step-up ratio and matching for a 2 panel system. That seems to have fixed or at least bypassed the fault, which also appears to have no fault in its HF circuit (much to my surprise, as I had long figured the problem to be intermittant contacts in the HF variable resistor...but those are now very well cleaned and treated with Deoxit). The Left HF circuit by itself did not fail in weeks of testing. Then, I disconnected the HF and reconnected the LF correctly to the Red wire. That did not fail in 3 weeks of testing, so I then tested the speaker full range for a week, and it did not exhibit any problem. However, as I describe below, there has been one failure which *might* have been caused by the Left speaker...there is still room for doubt...in fact any test of an intermittant problem in a limited time span could be wrong...there is simply no absolutely perfect test as long as the problem is intermittant, and the ones I've been looking at now have become the kind of thing which *might* happen once a month, making it very difficult to track down.
The Right Acoustat seems to have a problem in the HF circuit. I'm unsure if there is a problem in it's LF circuit, which I might have also changed (since I didn't write down what it was before disconnection). I've had the HF circuit disconnected for more than a month now just to determine if there is ALSO a problem with the LF circuit.
Setup like that, I have had exactly one shutdown when playing both the R and L speakers, but with only the bass transformer connected on the Right side. So does that indicate a fault with the L or R ??? Two weeks of followup testing on the Right by itself have not revealed a further issue. If none appears for another week or so, I'm going to write off that incident as a freak event, perhaps even power-related.
Meanwhile, I tried a super high power test, which yielded further ambiguous results, possibly caused damage.
I have a polarity test app which came with a wave file featuring a full level impulse sustained to a quite low frequency, possibly below 1 Hz. And it's a positive only impulse. This is a potentially destructive signal. And it repeats many times per minute on the wav file which lasts for 30 minutes.
But I can play it continuously for 30 minutes or more on both Acoustats, at 0dB (DAC is set to output 2V, which means the 600W potential power at 2.35V is cut to 255W peak rms), with the 100Hz crossover.
The crossover is very important. If I disable the 24dB/octave crossover at 100W, the LF transformers heat up greatly. I never see that heating up anymore with the 100 Hz crossover.
I ran full power (255W peak impulses with extended LF) into the right Acoustat for over 30 minutes. I did not notice the heating up. Then the fuse blew. Then I discovered the interface box had heated up to 120F right above the LF transformer.
I have not repeated that test. On the right speaker, with both transformers now connected, the 3A fuse blew immediately when switching in the signal. So, technically, this signal is not in the safe range of operation.
I'm suspecting the fuse blew immediately on the right side because both transformers were connected, creating a much greater HF load. However, it could also be because...the R channel transformer still has an operational fault, even using the correct wiring. If so, I'm not even sure how important that operational fault is, considering it's out of the safe operating range anyway.
Subsequent to testing both speakers to the full range signal, now known to be out-of-safe-range, I tested both with the 100 Hz crossover re-engaged, for at least 30 minutes, without failure (and the interface box seems to have negligible thermal rise, 2 degrees above ambient or less). This actually applies the same peak voltage, equivalent to 255W, into the speakers. And I have played the speakers for awhile. So it does not appear I actually damaged them with this test. But it raises further doubt to the belief that when the recent amplifier shutdown playing both speakers occurred, it was not because of the right LF transformer but the left.
BTW I've also shown that the Behringer DEQ does not switch out LC filters when you turn them off in the PEQ menu. You have to restart the DEQ to get this change (potentially unsafe...) to take effect. However, for a quick and possibly destructive test, the PEQ's can be completely bypassed in the bypass menu, and that takes immediate effect).
Engaging LC filters, however, does seem to take immediate effect.
******
But setting aside all the endless mind-numbing testing with ambiguous results, always incomplete, and on to what it has also brought to attention.
The Acoustat interface system is ingenius, but unlike the Transnova amplifier I don't see why it need to be considered a perfect design.
Listing to just the LF transformer alone, correctly set to the Red winding, yields fabulous sound. Perhaps better in some way than with both transformers connected. It has a greater coherent wholeness, and the speaker sounds far less up-tilted.
As I write this, however, I haven't spent much time listening to the newly corrected Right speaker, which now has greater output from the LF transformer with the Red connection. That was a huge and unquestionable improvement, perhaps the biggest upgrade in speaker performance since I got the acoustats. Never mind all the amplifier changes, etc, running the speaker on the correct transformer winding is huge, and huger than huge when it's all part of a dual frequency band network, as it off balances the two parts.
In fact, I could say the entire "lost 6 months" so far, since a speaker damaging accident was rewarded by this one discovery, so important is it, that I had the speaker(s) on the wrong LF transformer winding.
But now that I've measured the fairly decent HF performance of the LF transformer by itself...essentially flat to 1Khz, it makes little sense to me why the HF transformer should be crossed in starting around 350Hz. And I'm thinking the 1-transformer impulse looks nicer.
But these are also things I need to test better.
I've enjoyed thinking about the possibility of bi-amplifying the acoustat interface LF and HF circuits, but it would probably not be safe, as both amplifiers would share the same ultimate load on the secondary side of the transformers.
Still, I like the idea of removing the HF network, or at least the capacitor part, from the box for better free wheeling experimentation. I'd be interested in only rolling in the HF transformer above 7kHz, or finding some way to get phase perfect impulses.
But the issue might not even be in the dual transformer arrangement either, but in the very idea of a need for extended high frequency power response. Perhaps rolled off highs have a way of just sounding better. Lots of classic systems have that sound. Maybe this is even especially true for the FM source I listen to nearly always in background, and maybe less true for more 'clean' sources.
So perhaps the "solution" could just as efficiently provided with digital EQ, applied to the sources that need it.
Anyway, the speakers own internal engineering is now exposed, it can be questioned with and tinkered with, and great impacts and questions arise. Differences that hugely affect measurement and acoustic performance. Much greater than the imaginary wiring and amplifier differences.
A friend thinks I'm beginning to question the virtue of 'flat frequency response' above other things, which is good, to his mind.
No comments:
Post a Comment