Both of you kidding right ..well i know Frank is serious , but c'mon Brax ...
Frank, Rick (Abrax) is right. Even in case of preamps with two sets of output RCA jacks.
Input impedances of power amps vary wildly. The lowest I am aware of is 8.2k, Monrio of Italy, but Norway's Electrocompaniet does over 100 k. Most are somewhere beteeen 18k and 30k. My Marantz 170DC is 82k, but my H/K Citation is 22k.
In my particular case, this is of no importance, because all of my possible preamps, H/K Citaion, Luxman c-03 and my own headphone amp/preamp use discrete componentry throughout, and especially in the output stage. H/K has a pair of small signal SEPP outputs, Luxman uses some 10W trannies and my model uses a pair of 50W trannies. The Citation is not very sensitive to dual output drive, but some differences, albeit subtle, can be detected - and if they can be detected, they are obviously not subtle enough. With the Luxman, no differences I can hear, unless i try to run my headphones in parallel with output drive, when they do appear.
Only my own baby doesn't care if you run a power amp in paralell with the headphones, zero audible change. I tried adding an RCA output splitter so I can drive two power amp and the cans - still no difference. That's the BJT version, the tube version is incomparably more sensitive to such frolicks.
So, Rick is very much right, it's far from being all the same.
Yes, input impedances vary a great deal, but that should mean nothing to the the pre-amp. A rubbishy, cheap as chips opamp would be treated as a bit of a joke if it couldn't drive, say, a 2K load with close to zero impact on its performance. Yet we're going to treat precious, expensive pre-amps as spoilt brats, that will misbehave if forced to face something as daunting as a 10K input - give us a break!!
Much more likely scenario is that the power amp is spewing loads of gargage into the power lines, which feeds back into the pre-amp's "feeble" power supply - and bang goes your sound ...
Much more likely scenario is that the power amp is spewing loads of gargage into the power lines, which feeds back into the pre-amp's "feeble" power supply - and bang goes your sound ...
Zero chance of that with me, Frank, remember, I have separate power line filters for each and every device I have. Attenuation is 26 dB at 20 kHz and over 64 dB at 80 kHz (last frequency of any mentionable harmonic I have ever measured on my power lines). And please remember that my filters are completely symmetrical, no set in and out, which means that they also filter both ways. So, for any junk from the power amp to reach the preamp, that junk would have to pass through its own filter first, and then through the preamp's filter. Nothing even emotely woth mentioning can survive two series filters with max attenuation of 67 dB each.
I would look more at the input stages of preamps and power amps, the sometimes very complex RC networks inside.
Also, much as I like and respect op amps, I said it before and I say it again - never had any one of them, from anybody at any price, which didn't work way better when a discrete current booster was added. Basically, we are talking current here, even if the values in play APPEAR to be ridiculously low. Appearances can be misleading.
Last edited:
Good to see serious efforts for mains filtering - one of the "last frontiers", I've done a number of rounds on this, and still am not 100% able to clean all the muck coming through, to null all audible effects. Vastly improved on what I was achieving originally, but stick a nasty noise source into the house power and I can still hear the subtle variation in quality, downward of course. It appears that incorporating a number of layers of protection is needed, or to make the components extremely resistant to interference - no idle exercise ...
With regard to input stages, a power amp designer would get rapped over the knuckles if it appeared as anything but purely resistive to a driving signal - though in the crazy world of high end audio I guess people will do almost anything, ...
With regard to input stages, a power amp designer would get rapped over the knuckles if it appeared as anything but purely resistive to a driving signal - though in the crazy world of high end audio I guess people will do almost anything,
...Good to see serious efforts for mains filtering - one of the "last frontiers", I've done a number of rounds on this, and still am not 100% able to clean all the muck coming through, to null all audible effects. Vastly improved on what I was achieving originally, but stick a nasty noise source into the house power and I can still hear the subtle variation in quality, downward of course. It appears that incorporating a number of layers of protection is needed, or to make the components extremely resistant to interference - no idle exercise ...
With regard to input stages, a power amp designer would get rapped over the knuckles if it appeared as anything but purely resistive to a driving signal - though in the crazy world of high end audio I guess people will do almost anything,
Jesus, don't get me started on this, Frank, I've been making and selling them for the last 11 years.
The reason why you still hear some disturbance here and there is very probably that some of it passes through the filter (no such thing as a perfect filter) and does get into the amp, but since the amp is AFTER the filter, it can still pass it on to the preamp, unless you use individual filters for each device. My 15 years of dealing with the problem tell me that the best we can do is to use indisvidual filters for each input.
You will often hear laments regarding dynamics. This is indeed the bugbear of all filters, not because it has to be so, but because people do not think the whole thing through. Plus preconcepts, the "high end" audiofools figure that if any signal has to pass through anything, there must be distortion - I can agree with that, but if this distortion and/or limiting sinks down to insignificant, purely statistical values, then so what if it's there, all the more so if the mains borne junks is not there any more?
You can't seriously tell me a line filter capable of passing 48 Amps (sic!) befor showing first signs of core saturation is going to limit anything, short of some wild behemoth delivering literally kilowatts into low impedance loads?
Frank, after the last 11 years of active manufacture and sales to 28 countries around the world (yours is not one of them, but New Zealand is), I can tell you only this - before you buy anything, first try it at home, or, failing that, don't buy without an iron clad money back guarantee. THIS is the key problem of one and all power line filters, you never actually KNOW how it will or will not fit into your system until you actually try it for yourself. The results can be anything from rather slim to unbelievable, no other way to know but to try it out.
Very few things are general truths. Actually, I know of only one, my filter will not be good for owners of Jadis equipment, so I always ask what the prospective customer has, and if it's Jadis, I talk them out of buying so as not to waste money on shipping back.
Last edited:
Yes, sorry, dvv, it's almost bedtime here - I had forgotten that sideline of yours ...
I wouldn't buy a commercial one, my latest creation is completely original, and is designed around, complements the power supply, which is also modified, of the unit it's driving. It is very "light" in its action, no brute force handling of heavy current flow with huge transformers or chokes - the amps flow freely ...
I wouldn't buy a commercial one, my latest creation is completely original, and is designed around, complements the power supply, which is also modified, of the unit it's driving. It is very "light" in its action, no brute force handling of heavy current flow with huge transformers or chokes - the amps flow freely ...
Yes, sorry, dvv, it's almost bedtime here - I had forgotten that sideline of yours ...
I wouldn't buy a commercial one, my latest creation is completely original, and is designed around, complements the power supply, which is also modified, of the unit it's driving. It is very "light" in its action, no brute force handling of heavy current flow with huge transformers or chokes - the amps flow freely ...
... but it obviously leaves something to be desired if you hear subtle changes when something else at home comes on line.
Yes, that's very true. Theoretically, the attenuation over a huge range of frequencies should mean all is OK, but I haven't the equipment to test the reality - it certainly works on extreme noise injection, I've mentioned my simple test of creating a mini arc welder by deliberately half plugging in a strong current draw into the adjoining power socket: on a normal system the speakers become a fireworks soundtrack, but this filter kills that interference to stony silence.
The thing is, the interference may be bypassing the filter in some subtle way, there are so many pathways for unwanted electrical effects to disturb things - it becomes a ceaseless activity of trying one thing after the other, and at times you just get sick of doing this ... just leave it be, if there is a simple alternative: switch off the offending device ...
Sure - 'should' mean nothing to the preamp. But in practice it does - plenty of listening (of course anecdotal, not rigorous tests) reports say that higher impedance load sounds better. I was looking a couple of months ago at a valve preamp and they were selling an external PSU for it as an upgrade and said that it improves dynamics. I must admit I was highly skeptical at the time I first discovered it but I now see no reason to doubt what they say, based on my own experiments and listening.
Measured performance, sure no question. My AD605s are driving lower than this (about 400 ohms, but the supply volts are lower, 5V single-ended) and putting a CCS on the output made a notable difference to the dynamics. Adding more caps beyond the 6000uF already on the rails makes a more subtle difference to rhythm and pace.
So its not the driving chip (or discrete amp), rather its the power supply. Loading it down reduces dynamics. Can't for the life of me think why the sound should be so sensitive to the power supply, but there it is.
Too anthropomorphic for me Frank.
They perform best when the load is lightest, no question about it. I reckon this is the reason why trafo-based volume controls are considered the most transparent. Even when they're autotransformers and not providing any isolation. Over on the 'Beyond Ariel' thread the guys that listen say an autotrafo sounds better than a resistive attenuator.
My pet topic at the moment so I have a question here. When you used your discrete current booster was it fed from precisely the same rails as the preceding opamp? I'd just like to have a go at teasing out how much of the subjective effect is PSU related and how much might be opamp related.
I've read somewhere that two exact same amps from the same manufacturer sound and measured different!
Yes, it's got a lot to do with this: designers design PS's, or they design circuits - and never the twain shall meet ...
Every time I've delved into truly modelling what goes inside a circuit, the total circuit starting with the tatty mains coming in, and adding in some decent, real world parasitics, it's pretty obvious why one has problems: there are spurious glitches everywhere, and one ends up marvelling that the sound does come out as clean as it sometimes does!
The impedance loading thing is still a joke, nearly all circuits are based on stages, where the previous sub-circuit is fed into the next, loaded effectively by somewhere between 10K and 100K. By this reasoning all circuits are defective, because they trying to drive too many low impedances, internally! If you chopped off the side of the pre-amp, and a side of the power, and hardwired output to input, direct, with hookup wire would this interface still be so flimsy ...?
My DIY chip amp had very low input impedance, 2K if I remember correctly, the old Yamaha battleship CD player had no problem directly driving this, gave me sound as good as any I've had ...
I have yet to hear a pot type device which is not clearly audible, with an especially negative impact. First thing I would do is rip every single one out, and substitute with an effective alternative ...
Last edited:
Yes, well done actives should easily rule the roost, Bob, since the electronics would be optimised for the particular configuration of drivers ...Is active or passive better? ...Personally I prefer active (3D glasses).
No - looks like you jumped to an unwarranted conclusion there. Since this seems to be about power supply loading, then the form of classA which maintains a constant draw on the supply would be totally immune. Its my view (hypothesis) that its this effect that leads dvv to swear by loading his opamp outputs with a CCS.
In my current experiments with AD605s, adding current sources to the two outputs (running balanced, anti-phase) did produce a significant improvement. This puts the output stage into CCS-loaded classA - so long as the current taken from one stage returns to the other then there's no dynamic load on the PSU. However I'll admit I haven't gone to lengths to ensure this criterion is met. There must be other effects internally to the supply too since adding caps helps.
Since the AD605 is internally an AFA (see diagram on my blog) which is an architecture which has been intriguing me for many months now, I have studied DSs for other parts in the family - AD8129 is one. This one's DS has a plot of supply current vs differential input voltage, and there is indeed a significant variation of several mA, perhaps a characteristic that's the achilles heel of the AFA. So I figure as the AD605 is internally very similar to the AD8129 its internal stages aren't working in pure classA.
Yes, well done actives should easily rule the roost, Bob, since the electronics would be optimised for the particular configuration of drivers ...
Anyone seen 'well done' actives? I haven't yet, still looking. They'd need passive line-level XO for a start, totally separate power supplies optimized for the individual amps, amps optimized for each of the drivers' loads ...
- Status
- Not open for further replies.