Re: Muting/limiting
I suspect they use it just for DC offset protection...Didn't Self object to this on the grounds that the triac would have to be replaced, and that the output devices might well be destroyed as the rail fuse could not be relied upon to blow in time to save them when the triac shorts the output to ground?
Netlist said:
I suspect they use it just for DC offset protection...Didn't Self object to this on the grounds that the triac would have to be replaced, and that the output devices might well be destroyed as the rail fuse could not be relied upon to blow in time to save them when the triac shorts the output to ground?
Hi Mike,
Yes, this is SOA limiting as we are familiar with. Each polarity having it's own limiting and detection device.
What I don't care for is one device to sense over current for both polarities. Either across one emitter resistor or both. This usually means that current limiting will not occur in the normal fashion, but rather more globally.
In my view, brief over current events should cause current starving or some kind of "clipping" rather than a global shut down. Opening a set of relay contacts pretty much ensures poor sound quality to follow. This is unnecessary and avoidable. Classic base clamping methods achieve this goal with no lasting sonic cost.
-Chris
Yes, this is SOA limiting as we are familiar with. Each polarity having it's own limiting and detection device.
What I don't care for is one device to sense over current for both polarities. Either across one emitter resistor or both. This usually means that current limiting will not occur in the normal fashion, but rather more globally.
In my view, brief over current events should cause current starving or some kind of "clipping" rather than a global shut down. Opening a set of relay contacts pretty much ensures poor sound quality to follow. This is unnecessary and avoidable. Classic base clamping methods achieve this goal with no lasting sonic cost.
-Chris
Hi,
based on the dynamic headroom data posted by Mike, then Sony's best are pretty terrible.
A dynamic headroom of 1.8db for an 8ohm load is attrocious, when they should be striving for 0.5db or better in their best amps and possibly as bad as 1db in their good amps.
This appears to be a Japanese manufacturer using advertising script writers to make a bad feature/specification look as if it were attractive.
based on the dynamic headroom data posted by Mike, then Sony's best are pretty terrible.
A dynamic headroom of 1.8db for an 8ohm load is attrocious, when they should be striving for 0.5db or better in their best amps and possibly as bad as 1db in their good amps.
This appears to be a Japanese manufacturer using advertising script writers to make a bad feature/specification look as if it were attractive.
AndrewT said:
Um, why should anyone strive for LESS dynamic headroom? Last I looked, this was a figure determining how much short term power you can get from an amp OVER it's rated power output. The on;y problem I can find with this is that no-one tends to define the 'short' in 'short term'...
Hi,
although I have never had the oportunity to read the IHF standard, it defines how the test should be done.
A high dynamic headroom indicates a poor ability of the amplifier to supply sustained power. The higher the IHF headroom the worse the amplifier.
An ability to maintain output current at low frequency requires the opposite to dynamic headroom.
Have you been conned by the
although I have never had the oportunity to read the IHF standard, it defines how the test should be done.
A high dynamic headroom indicates a poor ability of the amplifier to supply sustained power. The higher the IHF headroom the worse the amplifier.
An ability to maintain output current at low frequency requires the opposite to dynamic headroom.
Have you been conned by the
Hi Andrew,
I believe the headroom spec on my Marantz 300DC was 3 dB, or close to it. It makes a more efficient amplifier as I don't play test tones. I don't think the Marantz is a terrible amplifier.
I get what you are saying, but I don't buy it. A Krell would do well in this test, but I don't want such a heavy, hot object in my living room. An amp that can deliver peaks of this magnitude is impressive given the 8 R RMS rating is 150W / ch. A 20W amp. that collapses down to a 10 W amp. on sustained notes would be pretty wimpy.
-Chris
I believe the headroom spec on my Marantz 300DC was 3 dB, or close to it. It makes a more efficient amplifier as I don't play test tones. I don't think the Marantz is a terrible amplifier.
I get what you are saying, but I don't buy it. A Krell would do well in this test, but I don't want such a heavy, hot object in my living room. An amp that can deliver peaks of this magnitude is impressive given the 8 R RMS rating is 150W / ch. A 20W amp. that collapses down to a 10 W amp. on sustained notes would be pretty wimpy.
-Chris
Hi Anatech,
I built a pair of low bias ClassAB amp modules (BJT quasi) into stereo power amp.
It had a single pair of 150W output devices running, almost cold, on +-50.5Vdc.
It put out 110W +110W into 8r and 200W into 4r. The supply rails only fell about 2.5V from the quiescent condition. It sounded so good I built two more. This was all about 30years ago before I knew much, if anything, about analogue electronics.
Now, looking back, I suspect the peak voltage on short term signals might be just 2V higher than the sustained peak voltage. This implies an approximate headroom of 0.4db (44Vpk into 8r vs 42Vpk into 8r). Even a 3V gain in Vpk will still result in a dynamic headroom of 0.6db. Dynamic headroom figures of 2db to 3db are terrible.
I could skimp on the smoothing caps and get a lower continuous power. I could skimp on the transformer and get a further lowering of continuous power. Doing both will definitely achieve a greater headroom but the result will be an amplifier that would sound less good.
I would call the high headroom ampilfier flawed but significantly cheaper than the "balanced" design.
I would certainly NOT call it "efficient".
"Cheap and nasty" seems a more appropriate term.
I built a pair of low bias ClassAB amp modules (BJT quasi) into stereo power amp.
It had a single pair of 150W output devices running, almost cold, on +-50.5Vdc.
It put out 110W +110W into 8r and 200W into 4r. The supply rails only fell about 2.5V from the quiescent condition. It sounded so good I built two more. This was all about 30years ago before I knew much, if anything, about analogue electronics.
Now, looking back, I suspect the peak voltage on short term signals might be just 2V higher than the sustained peak voltage. This implies an approximate headroom of 0.4db (44Vpk into 8r vs 42Vpk into 8r). Even a 3V gain in Vpk will still result in a dynamic headroom of 0.6db. Dynamic headroom figures of 2db to 3db are terrible.
I could skimp on the smoothing caps and get a lower continuous power. I could skimp on the transformer and get a further lowering of continuous power. Doing both will definitely achieve a greater headroom but the result will be an amplifier that would sound less good.
I would call the high headroom ampilfier flawed but significantly cheaper than the "balanced" design.
I would certainly NOT call it "efficient".
"Cheap and nasty" seems a more appropriate term.
Hi Andrew,
I haven't measured this, but the design of the 300DC is anything but cheap. Each power section has it's own winding, rectifiers and filtering caps. Each voltage amp section has it's own winding, rectifiers, filtering and regulators.
It could be Marantz's way of understating the normal RMS power output compared to the real peak. The output rails are around 70 VDC I think. The main filter caps are 10,000 uF things for each channel.
I guess I'm tring to point out the specs are not always arrived at the way you might assume. Marantz has always had a reputation of understating the capacity of their products.
-Chris
I haven't measured this, but the design of the 300DC is anything but cheap. Each power section has it's own winding, rectifiers and filtering caps. Each voltage amp section has it's own winding, rectifiers, filtering and regulators.
It could be Marantz's way of understating the normal RMS power output compared to the real peak. The output rails are around 70 VDC I think. The main filter caps are 10,000 uF things for each channel.
I guess I'm tring to point out the specs are not always arrived at the way you might assume. Marantz has always had a reputation of understating the capacity of their products.
-Chris
Hi Mike,
Yes, I was thinking of the previous discussions when I first replied. The second link where it refers to SS devices between the amp and speaker are out.
Power supply shut down is fine as long as it works properly. You would need some capacitance after the parts to bypass them. I doubt they would sound any good without.
However, for SOA protection, wouldn't it be less disruptive to limit the current in a multislope fashion. You could kick the power off if SOA limiting occurs for a time period. Just simply shutting everything down right off is drastic, don't you think?
-Chris
Yes, I was thinking of the previous discussions when I first replied. The second link where it refers to SS devices between the amp and speaker are out.
Power supply shut down is fine as long as it works properly. You would need some capacitance after the parts to bypass them. I doubt they would sound any good without.
However, for SOA protection, wouldn't it be less disruptive to limit the current in a multislope fashion. You could kick the power off if SOA limiting occurs for a time period. Just simply shutting everything down right off is drastic, don't you think?
-Chris
anatech said:
Yes........i think you're right.
http://www.diyaudio.com/forums/showthread.php?postid=629604#post629604
Re: Examples:
Sony TAN-R1:
http://www.thevintageknob.org/SONY/sonyesprit/TANR1/TANR1.html
mikeks said:Sony's 'best':
http://www.audio-circuit.dk/Schematics/Sony_TA-NR1.pdf
http://www.audio-circuit.dk/Schematics/Sony_TA-N80ES.pdf
http://www.audio-circuit.dk/Schematics/Sony_TA-E90ES.pdf
These are excellent designs in many respects, but, alas, are ruined by this relay SOA 'protection' arrangement.
Moreover, these designs compound the problem by detecting SOA conditions from only one polarity, which is a false economy at best as speech, for instance, is frequently offset to one polarity.
Sony TAN-R1:
http://www.thevintageknob.org/SONY/sonyesprit/TANR1/TANR1.html
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.