My 2 cts worth (as if anybody cares...):
It really doesn't matter if you built 1, 10 or 37 amps. What matters is that when you start to sell them, you have no control over the weird things people do to them. No control over the speaker cables, the xover networks, the speakers. No control over whatever they shove down the input. But there is one thing a user hates most: a blown amp. Or, for that matter , an amp going into Spike protection.
It is simply a matter of responsible behaviour to try to avoid that type of problem for the unknown, unsuspection user. That is the real issue here, not whether PD's of FD's or mine amp works fine for me. I run my amps without fuses, because I think I know what I'm doing. But I wouldn't dream selling amps without fuses.
The responsible thing to do if you start selling these things is to put in a Zobel. And a fuse.
Jan Didden
It really doesn't matter if you built 1, 10 or 37 amps. What matters is that when you start to sell them, you have no control over the weird things people do to them. No control over the speaker cables, the xover networks, the speakers. No control over whatever they shove down the input. But there is one thing a user hates most: a blown amp. Or, for that matter , an amp going into Spike protection.
It is simply a matter of responsible behaviour to try to avoid that type of problem for the unknown, unsuspection user. That is the real issue here, not whether PD's of FD's or mine amp works fine for me. I run my amps without fuses, because I think I know what I'm doing. But I wouldn't dream selling amps without fuses.
The responsible thing to do if you start selling these things is to put in a Zobel. And a fuse.
Jan Didden
fdegrove said:
You're probably talking very high capacitance cables now.
Otherwise, I know at least one case where low capacitance
is deliberately sacrificed in order to get low inductance, with
the motivation that a little extra capacitance doesn't matter
for speaker cables, but if anything matters at all, it is to have
low inductance. I wouldn't think those cables have so high
capacitance they would put you in jeopardy with normal
amps though. OTOH, normal amps have Zobel filters.
I totally agree, but if it turns out that 1000 people use PD's amp and none of them are having trouble, then by statistics PD may be right but this is a gamble and I would not do anything like that if I wasn't very, very sure.janneman said:
But the facts is that this Zobel thing isn't a major issue but I would recoommend Peter to connect a 10 m cable to a speaker and see what happens, and of course using an oscilloscope.
If Peter happens to sell readymade units I advice you to take a closer look to valid standards for safety and RFI/EMI with emphasis on safety. How should the power supply look like to fulfill standards?
The whole audio chain should be regarded as one system. Cables and speakers definitely belong to it.
If the amp speakers impedance is corrected and the cable's properties allow it then you might run the amp without zobel. If not, use one.
If no one knows what will be connected to an amp in advance (most cases) a zobel AND a damped series inductor might be your life insurance.
Simple as that.
Regards
Charles
If the amp speakers impedance is corrected and the cable's properties allow it then you might run the amp without zobel. If not, use one.
If no one knows what will be connected to an amp in advance (most cases) a zobel AND a damped series inductor might be your life insurance.
Simple as that.
Regards
Charles
Planet10
That isn't true, all Naim amps have a Zobel network, it's the series inductor at the output they lack, having a small series R and stipulating a speaker cable of moderate inductance to fulfil the role instead.
There's a clear stipulation in the user manual not to use cables other than Naim's own, without advice. The LCR specs of their cable are public knowledge, if one wants to try alternatives.
So even the most hard nosed use a Zobel, I wonder why?
Andy.
That isn't true, all Naim amps have a Zobel network, it's the series inductor at the output they lack, having a small series R and stipulating a speaker cable of moderate inductance to fulfil the role instead.
There's a clear stipulation in the user manual not to use cables other than Naim's own, without advice. The LCR specs of their cable are public knowledge, if one wants to try alternatives.
So even the most hard nosed use a Zobel, I wonder why?
Andy.
janneman said:
Jan, would you admit that "high end" amps often have different design goals, practices, and expectations compared with the stuff sold to the mass market? The latter is indeed driven by cost-consciousness. The former is to a much lesser extent- the important thing there is the story you're telling, not really the quality of the unit you're selling. Amps blowing up are absolutely unacceptable in the mass market. They're excused in the high end because <sarcasm> everyone knows that you have to sacrifice to make good-sounding amps, they're going to be tweakier and twitchier than the crap sold in mass market outlets, you have to expect that for high performance equipment</sarcasm>.
The analogy I think of is cars- a Ford Escort had better run flawlessly on any kinds of roads. A Ferrari, OTOH, is forgiven when the coils burn out or the valves go flying through the top of the engine or the user finds that they can only be driven on very smooth roads. A Pioneer receiver ($200) had better work all the time with any kind of lamp cord. A Jadis ($10,000) can blow up when you look at it sideways or fail to speak French in its presence. Their markets have different expectations.
That said, every solid state amp I use or would use has an output inductor and an RC network. But my system or my hearing probably sucks since I can't seem to hear so many of the wonderful things people tout.
... one thing though: LM3875/3886 are very good power amp IC's (the best?), better than many discrete amps, therefore the necessity is less important. This is something also National claims, and PD/BrianGT agree with, not exactly outspoken but still...janneman said:
Then agian I have both zobel and L//R in my design. If it turns out they are unneccassary, just remove them.
Sorry P-A, but that's bull.
It has nothing to do with 'the best' whatever that is. If you think that you clearly have no idea *why* those zobels and L are needed.
It has to do with loop gain, phase shift and excess gain. Those IC amps are no different than discretes in this respect, possibly worse, what with the relatively low bandtwidth output devices.
Jan Didden
It has nothing to do with 'the best' whatever that is. If you think that you clearly have no idea *why* those zobels and L are needed.
It has to do with loop gain, phase shift and excess gain. Those IC amps are no different than discretes in this respect, possibly worse, what with the relatively low bandtwidth output devices.
Jan Didden
As far as I understand, the RC network has two advantages. It makes us possible to control frequency response and it prevents unwanted oscillations. If someone says that I do not need it, he does not need it, tho.
Really, SY? But, I do not think that hi-end audio makers pursue something special at cost of safety or reliability. If the reliability problem occurs from it, the maker’s reputation hits the bottom. The hi-end audio market is a small world.
Really, SY? But, I do not think that hi-end audio makers pursue something special at cost of safety or reliability. If the reliability problem occurs from it, the maker’s reputation hits the bottom. The hi-end audio market is a small world.
P-A,
These chip amps are a lot slower than most discrete amps. The output transistors are no match for those fast Toshiba's and Sanken's that everybody raves about. In fact, it is clear if you look at the plot of the open loop gain. These amps (especially of course the output devices) are so slow, giving rise to large phase shifts, that they roll off the open loop gain 3dB already below 100Hz. (I think they are too embarrased at National to show the real roll-of point, because the curve stops at 100Hz).
In this respect, they are to discrete amps what the old LM741 is to modern. fast opamps. And, not coincidentally, the 741 also rolls of below 100Hz, IIRC already at 10Hz.
Jan Didden
These chip amps are a lot slower than most discrete amps. The output transistors are no match for those fast Toshiba's and Sanken's that everybody raves about. In fact, it is clear if you look at the plot of the open loop gain. These amps (especially of course the output devices) are so slow, giving rise to large phase shifts, that they roll off the open loop gain 3dB already below 100Hz. (I think they are too embarrased at National to show the real roll-of point, because the curve stops at 100Hz).
In this respect, they are to discrete amps what the old LM741 is to modern. fast opamps. And, not coincidentally, the 741 also rolls of below 100Hz, IIRC already at 10Hz.
Jan Didden
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.