Batteries ....! Dead sound ..........
Batteries, like everything else, have their pros and cons. On the one hand, they are as near to perfect DC as we can come, which is a definite plus. But they are slow, much slower than most electronic ciruits, and are thus not well suited to anything that has sudden large voltage/current swings, such as power amplifiers.
Stax had a 100 wrms out-of-this-world power amp, which was on a cart, and another cart moved around its impressive set of batteries, weighing in at over 100 kilos ( > 220 lbs). The inital series was fed straight from the batteries (don't ask me how) and returned dead as a doornail sound. This was quickly rectified by adding an electronic regulation circuit between the batteries and the amp, after which it turned into a legend in Japan. Unfortunately, as all things Stax, while meade in state of the art mode, its price was sheer astronomy, around Jupiter, I think, so few were made.
These days, batteries are used to power Nagra's tube preamp, but not the power amp. Both cost an arm and a leg.
I should have the service schematic of the above Stax amp, if anyone is interested, I can either post it, or e-mail it.
@dvv
" I hope there is no-one here who believes that powwer transformers are not stressed by line hash. When they start humming is only the last stage, by that time, you have a catastrophe on your hands."
What on earth is that all about??
You must be drinking something very different to me
edit: didn't see Abraxalito's comment
As is, a transformer hums when power is applied, the sound is reasonable but strictly 2D. Add a filter, no humming, and the soundstage acquires some distinct 3D features.
Measure S/N ratios as is, add a power filter, and measure again - better.
For what's in the grid, simply connect an oscilloscope and take a peek. You WILL be surprised, although I must add that this does vary quite a bit depending on your location. If the local power transformer is next to your home, the benfits will be smaller, but increaseing the further away you are.
Especially effective in densly populated areas, such as big cities.
Especially effective for tube audio, which seems to benefit the most. As for why, that you need to ask the designers.
A 50:50 proposition for amps with power regulation. Works very well with older Naim gear, but not impressive with new Naim gear; I have heard it literally transform two all Krell systems from mundane to outstanding, and big Krell FPB amps have several regulation circuits inside. You never know, the only way to really know is to try it and see/hear.
THAT'S the bugbear of all power line electronics - manufacturers advertise as if they were selling a ticket to paradise, which can happen, but is by no means guaranteed to be so. Again - the only way to KNOW is to try them.
Hi,
Correction, a switched mode supply powers the Nagra Tube Pre. The Batteries power that.
I have used battery powered solid state before. It does have it's distinct advantages, but practicality is low. Since I have investigated power supplies more to get closer to batteries without needing them.
Ciao T
Correction, a switched mode supply powers the Nagra Tube Pre. The Batteries power that.
I have used battery powered solid state before. It does have it's distinct advantages, but practicality is low. Since I have investigated power supplies more to get closer to batteries without needing them.
Ciao T
Hi,
The problem in this case is DC offset on the mains. All gear of mine (at home) and all I design includes as an ABSOLUTE MINIMUM the DC blocker and a snubber across the mains to kill the mains inductance.
I normally prefer to build the filtering into the transformer, where I cannot do that sufficiently, I used heavily overrated LC filters with corner frequencies as low as I can find. The NEC-Tokin Filter I picked for my poweramp has a 30KHz corner and the size of a big electrolytic cap... It will be combined with 68,000uF cap as DC blocker...
Ciao T
The problem in this case is DC offset on the mains. All gear of mine (at home) and all I design includes as an ABSOLUTE MINIMUM the DC blocker and a snubber across the mains to kill the mains inductance.
I normally prefer to build the filtering into the transformer, where I cannot do that sufficiently, I used heavily overrated LC filters with corner frequencies as low as I can find. The NEC-Tokin Filter I picked for my poweramp has a 30KHz corner and the size of a big electrolytic cap... It will be combined with 68,000uF cap as DC blocker...
Ciao T
In my view, about 98% of what should not be there but is lies in the 20 - 80,000 Hz range. Mostly fundamentals from various devices and their harmonics.
Manufacturers often quote cell and wireless phones, radio and TV - while they all do have their share, what is often too lightly passed over is that their relative POWER ratio is very low, especially compared to such devices as washing machines, hot water boilers, etc. I am not saying they are not a problem, I am saying they are much less of problem than they are made out to be.
Which explains why simple, IEC socket installed, power line filters do next to nothing - they have their -3 dB cutoff point at around 600 kHz, by which time 98% of the hash has already gone inside.
Attach a temperature proble to a power transformer. Measure temp under normal operating conditions as is, then insert a filter, and measure again. Lower, not astonishingly so (as some would have you believe), but 3-5 degrees could happen easily.
As for voltages, with any luck, you will see peaks in the range of 600-900 Volts, all very short of course, typically less than 1 mS.
Try some of what I pointed out, and we can discuss it. I am used to people not believeing what's going on in the grid, so until you see it for yourself, there's no point in discussing it. Or, if you do wish to discuss the point further, move it to the private message region, else I'll be accused of misusing the forum to sell.
Point taken, but in real world terms, it's a small difference. The original power source is still a battery, a near perfect DC source. Whatever you do to it afterwards will not improve it, just change it.
Agreed on the disasterously low practicality factor. Simply put, they are cumbersome, expensive and fidgety. You're much better off investing the money and real estate inside a device on classic supplies, but done right.
Thorsten, from our discussions way back, I know you pay much attention to your power supplies. You have realized that power is to a circuit what our blood is to us - you can't expect a man with heart problems to run a marathon.
What you have to understand that you are THE minority. It often surprises me how flippant about power sources the mainstream industry is. The logic seems to be - what the hell, just throw in two medium sized caps which don't cost too much, and that's it. What? We have 5 100 wrms channels inside? OK, then don't use a pair of 6,800 uF caps, use a pair of 8,200 uF caps.
End result - if you want some serious oomph from your subwoofer, you have to get a self-powered one.
If anything has distinctly DEGRADED in the mainstream audio since the 70ies, those would be the power supplies. Just take a look at the Pioneers, Kenwoods, Sonys and the rest of them in the 70ies and compare them with similar products today.
In those days, products like Sansui's 2x85 wrms amps had at least 80,000 uF in front of them, often from separate left and right power transformers of rather good quality; today, if you find 30,000 uF in a similar product, you are one lucky man.
Some, notably the Brits (Naim, Cyrus, etc), will offer you add-on power supplies which will improve your sound - but by the time you have your end price, you discover you would have been better off buying a say Accuphase and having it all inside just one case. As I see it, they are charging you extra for what should have been there in the first place.
One of my two gripes about H/K gear is what I feel are their undersized filter caps. They get the numer right, they get the power transformer size right, but
they somewhow always undersize them by at least one step, usually two or three. Where they put in 8,200 uF, I would have put in at least 10,000 uF, more probably 12,000 or even 15,000 uF.
But, of course, they have price limits they must satisfy, and these by deafult come before the quality consideratons.
Hi,
Well, what is done in this case is to turn into a near perfect squarewave, send it through a stepup transformer/inductor and then rectify it. To me this would suggest defeating all the key benefits of the battery...
Oh no, I understand this all too well. When AMR eventually gets around to it, we will re-package all the mains side technology we build into our 77 series Products and make them into a standalone conditioner... But we are currently working on other stuff...
We still have to work out the unique angle's and selling points (that is the marketing side).
The fun part is that practically every dealer and distributor has found that AMR gear most of the time sound best straight into the wall, no filters, expensive distribution etc...
I remember one very upset customer when I told him what degraded the sound of his system (AMR Phono, CD-Player and Monblocked Amp's) was his very expensive power conditioner. In the end the only that sounded better plugged into that conditioner rather than straight into the wall was his Turntable!
I suspect that if every manufacturer did what AMR do on the powersupply side (the actual cost is fairly notional, at least in a high end product) you'd be out of the mains conditioner business...
Ciao T
Well, what is done in this case is to turn into a near perfect squarewave, send it through a stepup transformer/inductor and then rectify it. To me this would suggest defeating all the key benefits of the battery...
Oh no, I understand this all too well. When AMR eventually gets around to it, we will re-package all the mains side technology we build into our 77 series Products and make them into a standalone conditioner... But we are currently working on other stuff...
We still have to work out the unique angle's and selling points (that is the marketing side).
The fun part is that practically every dealer and distributor has found that AMR gear most of the time sound best straight into the wall, no filters, expensive distribution etc...
I remember one very upset customer when I told him what degraded the sound of his system (AMR Phono, CD-Player and Monblocked Amp's) was his very expensive power conditioner. In the end the only that sounded better plugged into that conditioner rather than straight into the wall was his Turntable!
I suspect that if every manufacturer did what AMR do on the powersupply side (the actual cost is fairly notional, at least in a high end product) you'd be out of the mains conditioner business...
Ciao T
I, personally, am not aware of any of the similar calibre.
Btw, if anyone knows decent cutting house in Europe, please let me know or PM me with contact details, as I'm looking for one.
By "decent" I mean one, that has its signal chain free of unnecessary cabling/electronics and what else and that don't use A-820/A-80VU/A-80RC machines with their stock playback electronics. C-37 also doesn't count due to mechanical reasons. Plus they should be specialists in their field as well.
Hi,
Well, what is done in this case is to turn into a near perfect squarewave, send it through a stepup transformer/inductor and then rectify it. To me this would suggest defeating all the key benefits of the battery...
Well, when you put it like that ... I did wonder what was the point of using batteries coupled to another set of redundant electronics before you pass it on to the audio circuits ... but that's just me.
Oh no, I understand this all too well. When AMR eventually gets around to it, we will re-package all the mains side technology we build into our 77 series Products and make them into a standalone conditioner... But we are currently working on other stuff...
We still have to work out the unique angle's and selling points (that is the marketing side).
The fun part is that practically every dealer and distributor has found that AMR gear most of the time sound best straight into the wall, no filters, expensive distribution etc...
I remember one very upset customer when I told him what degraded the sound of his system (AMR Phono, CD-Player and Monblocked Amp's) was his very expensive power conditioner. In the end the only that sounded better plugged into that conditioner rather than straight into the wall was his Turntable!
I suspect that if every manufacturer did what AMR do on the powersupply side (the actual cost is fairly notional, at least in a high end product) you'd be out of the mains conditioner business...
Ciao T
No problem at all, I plan to retire at 65, which is just 6 years and 3 months away.
I'll probably sell the company lock, stock and barrel, and devote myself to doing what I've been wanting to do for many years, but never had the time. And - hopefully - to grandchildren.
Heck, I might even buy a dog, I've wanted one since I was a child in nappies. But, being a city slicker, and living on the 8th floor, I never really could.
Well most if not all amps have more than a couple inches of rail , yet I cannot recall ever seeing one with an decoupling cap at every output device ..
Hi,
Well, what is done in this case is to turn into a near perfect squarewave, send it through a stepup transformer/inductor and then rectify it. To me this would suggest defeating all the key benefits of the battery...
Oh no, I understand this all too well. When AMR eventually gets around to it, we will re-package all the mains side technology we build into our 77 series Products and make them into a standalone conditioner... But we are currently working on other stuff...
We still have to work out the unique angle's and selling points (that is the marketing side).
The fun part is that practically every dealer and distributor has found that AMR gear most of the time sound best straight into the wall, no filters, expensive distribution etc...
I remember one very upset customer when I told him what degraded the sound of his system (AMR Phono, CD-Player and Monblocked Amp's) was his very expensive power conditioner. In the end the only that sounded better plugged into that conditioner rather than straight into the wall was his Turntable!
I suspect that if every manufacturer did what AMR do on the powersupply side (the actual cost is fairly notional, at least in a high end product) you'd be out of the mains conditioner business...
Ciao T
All power amplifiers sound best when connected straight to the wall, power filters work best when connected to ancillary equipment, of course there can be exceptions as DVV pointed out with one particular brand and if supply is really bad , I'm sure an filter would save the day...
In such a case a fully regulated PSU may just be the best choice.....
Last edited:
Hi,
McCormack DNA series did, IIRC, in the AMR Amplifiers I put a local 1,000uF Elnal Silmic directly next to each device (I know, 1,000uF are chickenfeed by your standards).
The Amp I'm re-working has the big PSU Cap's right on the power transistor board, maximum distance to a transistor is maybe 2" or 3". I will be putting Electrocube 15uF/100V MKP Cap's for each output device and one more at the driver...
Ciao T
McCormack DNA series did, IIRC, in the AMR Amplifiers I put a local 1,000uF Elnal Silmic directly next to each device (I know, 1,000uF are chickenfeed by your standards).
The Amp I'm re-working has the big PSU Cap's right on the power transistor board, maximum distance to a transistor is maybe 2" or 3". I will be putting Electrocube 15uF/100V MKP Cap's for each output device and one more at the driver...
Ciao T
Actually, how they will sound directly depends on the quality (not the quantity!) of their power supplies.
The range I have seen and heard ranged from hardly audible at all to a completely new ball game. That a filter will reform a so-so product doesn't surprise me at all, but that it should make such a difference with products like the Krell abd Audio Research, that does surprise me.
And makes me repeat, over and over again - you don't know until you actually try it.
Not to be impertinent, Wayne, but you could be in for a BIG surprise with your gear. Most owners of massive power amps think they have it made, until ...
Actually, what makes me the odd man out is that long ago, I realized that you can't skimp on power capacity of any filter, and experience has taught me that a number of otherwise very good quality power line filters manage to fail in that department. So unnecessary, so stupid. Which is why I make mine capable of handling 2,400 VA (10A at 240 VAC) or 4,800 VA (20A at 240 VAC). Which is probably quite a bit more than your power amp could ever draw from the mains by a factor of at least 4 (I'm guessing here).
How much does your power amp draw from the wall socket at full power?
Italy's manufacturers used to put 4,700 uF next to each output device, over and above of the central power supply line caps (Galactron), way back in the late 70ies.
Thorsten is probably limited to 1,000 uF because Lady Chancelor Merkel has decreed that Germans will save.
I am not so limited, so I'll be putting 2,200 uF caps by the output devices.
Hi,
Nope the evil force that makes me use 1,000uF is Elna, they simply have nothing bigger in Silmic II at the voltages needed.
An AMR AM-77 has 8pcs of these per channel, so an extra 8,000uF, main PSU has 100,000uF per channel, for a 180W hybrid Amp that does NOT double down...
Ciao T
Thorsten is probably limited to 1,000 uF because Lady Chancelor Merkel has decreed that Germans will save.
Nope the evil force that makes me use 1,000uF is Elna, they simply have nothing bigger in Silmic II at the voltages needed.
An AMR AM-77 has 8pcs of these per channel, so an extra 8,000uF, main PSU has 100,000uF per channel, for a 180W hybrid Amp that does NOT double down...
Ciao T
Thorsten and dvv, Ive been seeing some interesting posts about design here but there are some factors not looked into.
dvv claims that using 2 to92 parts (mpsa92) gives better THD figures. I would agree but this only happens at low frequency below about 1Khz. This is because the parallel combination has a higher breakdown rating so early voltage is higher. But what happens at high frequency is that Cob gets doubled causing doubled high frequency distortion, seems ones chooses his poison.
The case is similar for Thorsten. As early voltage is mainly responsible for the degree of low order harmonics when this is increased, mainly 2nd order harmonics will lower by a large amount when paralelled so it wont have such a nice looking structure any longer. If you want much higher 2nd harmonics and so called hiraga distortion the way to go is by using transistors with low early voltage, or one could say low breakdown but then one has to keep in mind supply voltages and overall distortion will be higher. Here too high frequency distortion will roughly be doubled at about 20 khz because doubling of the Cob.
The best way Ive found to keep "nice" distortion spectrum and low Thd is by using cascoding. Now a low early voltage part can be used, cascoding not only takes care of the early voltage part but also miller effect. Also this way very high beta parts or even small signal darlington like Mpsa13 can be used and if they have low Cob even better. For top high frequency performance Rf transistors work nicely. I usually use mpsa18 as it has very high beta with the benefits it brings + very low parasitic capacitances.
If you look closely at Samuel Groners paper and the distortion graphs youll find that it collaborates with his findings. Im not sure if this shows in sims but real world is diffirent matter.
dvv claims that using 2 to92 parts (mpsa92) gives better THD figures. I would agree but this only happens at low frequency below about 1Khz. This is because the parallel combination has a higher breakdown rating so early voltage is higher. But what happens at high frequency is that Cob gets doubled causing doubled high frequency distortion, seems ones chooses his poison.
The case is similar for Thorsten. As early voltage is mainly responsible for the degree of low order harmonics when this is increased, mainly 2nd order harmonics will lower by a large amount when paralelled so it wont have such a nice looking structure any longer. If you want much higher 2nd harmonics and so called hiraga distortion the way to go is by using transistors with low early voltage, or one could say low breakdown but then one has to keep in mind supply voltages and overall distortion will be higher. Here too high frequency distortion will roughly be doubled at about 20 khz because doubling of the Cob.
The best way Ive found to keep "nice" distortion spectrum and low Thd is by using cascoding. Now a low early voltage part can be used, cascoding not only takes care of the early voltage part but also miller effect. Also this way very high beta parts or even small signal darlington like Mpsa13 can be used and if they have low Cob even better. For top high frequency performance Rf transistors work nicely. I usually use mpsa18 as it has very high beta with the benefits it brings + very low parasitic capacitances.
If you look closely at Samuel Groners paper and the distortion graphs youll find that it collaborates with his findings. Im not sure if this shows in sims but real world is diffirent matter.
Hi,
Or maybe they are and there are reasons why things are done differently to what you expect?
The argument is for two or more common low power parts vs. a single higher current part. If you look closely, several low power parts can provide the same power dissipation, but much better Beta, Beta Linearity, Ft and Cbo than a single higher power part, of course correct selection is a must.
Just using several parts in parallel is not what is suggested...
I would always argue that all else equal using one part is the better choice, however, things are not equal.
In my case, for my Amplifier project using the 2SA970 and 2SC2240 in parallel in the VAS position is simply a conclusion taken from the fact that they are already in place (originally as differential VAS and Current Mirror), with emitter resistors. So I need to make minimal changes.
Plus I get an pair of 0.6W Transistors with around 5pF Cbo each under small signal conditions in the Amp, > 100MHz Ft and a Beta of around 300. Much of this matters little as each halve of the VAS has a J-Fet Driver that has around 400 Ohm output impedance, so the effects of Cbo and Ft/Fbeta as well of non-linear beta are quite reduced already, but why not take more.
Another way to look at this is to say that distortion is reduced AND as we have in my case a PP VAS with truely MASSIVE degeneration (which effectively further increases aparent early voltage), the VAS should contribute very little distortion of it's own, which was my design aim. Others may have different ones.
I do not necessarily disagree on the subject of cascoding, but in my case this is much larger to implement than what I actually have decided upon. And as said, given the amount of degeneration how much a cascode can help and how much it lowers the VAS HD next to the output stage is a matter of debate.
Ciao T
Or maybe they are and there are reasons why things are done differently to what you expect?
The argument is for two or more common low power parts vs. a single higher current part. If you look closely, several low power parts can provide the same power dissipation, but much better Beta, Beta Linearity, Ft and Cbo than a single higher power part, of course correct selection is a must.
Just using several parts in parallel is not what is suggested...
I would always argue that all else equal using one part is the better choice, however, things are not equal.
In my case, for my Amplifier project using the 2SA970 and 2SC2240 in parallel in the VAS position is simply a conclusion taken from the fact that they are already in place (originally as differential VAS and Current Mirror), with emitter resistors. So I need to make minimal changes.
Plus I get an pair of 0.6W Transistors with around 5pF Cbo each under small signal conditions in the Amp, > 100MHz Ft and a Beta of around 300. Much of this matters little as each halve of the VAS has a J-Fet Driver that has around 400 Ohm output impedance, so the effects of Cbo and Ft/Fbeta as well of non-linear beta are quite reduced already, but why not take more.
Another way to look at this is to say that distortion is reduced AND as we have in my case a PP VAS with truely MASSIVE degeneration (which effectively further increases aparent early voltage), the VAS should contribute very little distortion of it's own, which was my design aim. Others may have different ones.
I do not necessarily disagree on the subject of cascoding, but in my case this is much larger to implement than what I actually have decided upon. And as said, given the amount of degeneration how much a cascode can help and how much it lowers the VAS HD next to the output stage is a matter of debate.
Ciao T
- Status
- Not open for further replies.