Jocko,
shouting over from the tube world 🙂 .... still have to get adjusted how you say things (all , does the guy wear his teacher's badge outside ? 🙂 ) but i have to agree with you to100%. From listening exprerience. I thoroughly know Allen Wright's differential preamp going far into the MHz BW realm, no loop feedback but hefty FB with most stages. And must admit it is the best super-hifi preamp i ever heard, no sonic flaws. But for listening to music, for getting lost in a Beethoven piano sonata e.g., i prefer ironman Thomas Mayer's design (vinylsavor) who doesn't give a §$%& on bandwidth and i have given up to count the "slow" iron pieces in his tube preamp's signal path. Zero feedback, bandwidth just enough to ensure proper phase for audio frequencies. This simply is it.
I have particiapted on repeated listening to a SS amp having feedback and BW reduced on and on up to 100 and some kHz until the sucker was enjoyable. We had the same experience with a nice lil' PP power amp having 340 kHz BW before and finally sounding best with BW of 120 to140 kHz and everything taken out which could cause the amp to get a high frequency orgasm. Funnily, both amps, SS and tube amp ended up with about the same BW for best sonics.
I doubt the BW hype is it. And a loop FB always is a regulating loop. For getting alive, the error amp needs an error signal. Then, not earlier it can correct the error. Hence the error amp always is late. Too late? Depends. Depends on the design, depends also whether the listener is sensitve to those corrected errors. I admit it is possible to get the error small and bearable.
Klaus,
Thanks for the info. I have built several amps of the op-amp (or Naim) style with pseudo-complementary output stages. This type of output stage is cunning because it enables close matching of the power transistors and use of high speed transistors (since both are npn). Of course, our valve friends automatically benefit from close device matching since they have no choice but to use all N type devices. I would like to see what JLH has to say about D5. I can tell you that if you are using a 1N4148 it is unlikely to be optimum - try a wide variety of diodes and listen to the differences, you may be quite surprised.
Bernhard/Jocko
I agree that THD and other "standard" measures don't correlate well with perceived sound quality. I have also experienced the terrible music killing effects of feedback in some circuits. In fact I used to believe feedback was the work of the Devil and simply bad for sound in any form. However, I am now a lot wiser and I understand a lot more about why sound degrades. I can confidently say that it isn't feedback itself that is the problem. It is rather a widespread ignorance of how feedback performance is affected by non-linearities and thus a lack of due care in the open-loop circuit design. Having an error signal isn't a problem: no-feedback designs also have an error signal and it is usually much greater.
Thanks for the info. I have built several amps of the op-amp (or Naim) style with pseudo-complementary output stages. This type of output stage is cunning because it enables close matching of the power transistors and use of high speed transistors (since both are npn). Of course, our valve friends automatically benefit from close device matching since they have no choice but to use all N type devices. I would like to see what JLH has to say about D5. I can tell you that if you are using a 1N4148 it is unlikely to be optimum - try a wide variety of diodes and listen to the differences, you may be quite surprised.
Bernhard/Jocko
I agree that THD and other "standard" measures don't correlate well with perceived sound quality. I have also experienced the terrible music killing effects of feedback in some circuits. In fact I used to believe feedback was the work of the Devil and simply bad for sound in any form. However, I am now a lot wiser and I understand a lot more about why sound degrades. I can confidently say that it isn't feedback itself that is the problem. It is rather a widespread ignorance of how feedback performance is affected by non-linearities and thus a lack of due care in the open-loop circuit design. Having an error signal isn't a problem: no-feedback designs also have an error signal and it is usually much greater.
BAM,
I'm not sure I follow-
This type of output stage is cunning because it enables close matching of the power transistors and use of high speed transistors (since both are npn).
In a valve pp amp each side is being used exactly the same way. I do not believe this is true in this update of Shaw's improved quasi-complementry. Aren't the tranfer characteristics of both halves quite different, which is not the case in a hollow state amp? I'm not implying this isn't an excellent sounding topology as I've heard some great amps using it, Edge....
Paul
I'm not sure I follow-
This type of output stage is cunning because it enables close matching of the power transistors and use of high speed transistors (since both are npn).
In a valve pp amp each side is being used exactly the same way. I do not believe this is true in this update of Shaw's improved quasi-complementry. Aren't the tranfer characteristics of both halves quite different, which is not the case in a hollow state amp? I'm not implying this isn't an excellent sounding topology as I've heard some great amps using it, Edge....
Paul
Hi Paul,
To my reckoning the two halves are very nearly electrically equivalent if the right components are used. The choice of the diode is crucial and the driver transistors need to be well matched (which is achievable with drivers). Other than that they are practically identical. To be absolutely identical a resistor and diode are needed in the collector of the npn driver, but this will have negligible affect.
BAM
To my reckoning the two halves are very nearly electrically equivalent if the right components are used. The choice of the diode is crucial and the driver transistors need to be well matched (which is achievable with drivers). Other than that they are practically identical. To be absolutely identical a resistor and diode are needed in the collector of the npn driver, but this will have negligible affect.
BAM
For listening purposes specs are almost totally useless.... Aleph's are great amps and once a few of us have had a chance to build the XA we'll will be more qualified to judge it on it's merits. If your looking to build soon i'd go for the Aleph, you can check out the gallery pages @ www.passdiy.com and the "Aleph 2 and others as a project" thread in this forum........ the Aleph requires huge heatsinks because it operates in Single Ended Class A mode which means that a great deal of the power that goes into it becomes heat not sound.
Also just let me remind you that the system is only as good as the weakest link so if the rest of the system is not up to the task, then you may not be pleased with the result. Finally, you may initially find that the Aleph's or other hifi amps dont sound like what you perceive to be flat.... they usually sound a little dark in the midrange and this perception is generally due to the poor quality amps most us were used to listening to before we got involved in diy audio.
Also just let me remind you that the system is only as good as the weakest link so if the rest of the system is not up to the task, then you may not be pleased with the result. Finally, you may initially find that the Aleph's or other hifi amps dont sound like what you perceive to be flat.... they usually sound a little dark in the midrange and this perception is generally due to the poor quality amps most us were used to listening to before we got involved in diy audio.
Hi,
Take a look at 118W ZAPsoluteMK4 ta LC Audio. This is great construction and i think sound will be great, too.
link: http://www.lcaudio.com/mk4dia.pdf
Take a look at 118W ZAPsoluteMK4 ta LC Audio. This is great construction and i think sound will be great, too.
link: http://www.lcaudio.com/mk4dia.pdf
I may as well give my 2cents worth as well .... 🙂
Gabster, you have undoubtedly realised there is no "highest-fi", or not yet anyway. You need to define what you are looking for a little more carefully. Can you accept SE class A, with the large hardware and enormous heat? Or, do you require an amp which is class AB, ie. less power inefficient? Do you really need 100W, or have you just plucked this figure out of the air because it sounds impressive?
If you are using reasonably efficient speakers (>=85dB/W, say) then you are actually unlikely to need this sort of power and your choices will be much wider.
Class A/B the Leach amps are a good choice. I used them for many years and they are honest performers (the one I built in 1978 is still working in my garage). Some of the other suggestions would likewise be suitable.
Class A, I obviously have a bias to Passlabs but there are others out there (JLH for example http://www.gmweb.btinternet.co.uk/jlh1996.htm ). The A40 (http://www.passdiy.com/pdf/a40.pdf) is push-pull class A so not quite as inefficient as the Alephs and is an excellent amplifier. Once you traverse the 100W mark here (Aleph4, 1.2) you are talking really serious hardware and heat. Most of us only listen in the <10W range anyway, so the remainder is just room warming!
If you do really need some serious power, see the 1kW beast on Anthony's site:
http://www.aussieamplifiers.com/poweramp.htm
Rod Elliot makes some good amplifiers:
http://sound.westhost.com/project03.htm
There are also a number of very good articles on his site. (he also has an interesting section on the "Death of Zen" ..... but that's stoning material around here 🙂 )
If you can handle the heat, I would recommend the Alephs above the Leach amps (and virtually all "quasi-class-A" amps I have heard), simply they sound better.
Unless you really do need the additional power I would recommend the Aleph5 (65W - OR, the Volksamp equivalent) .... the hardware is just a little more bearable. Remember the Aleph4 (100W) only gains a couple of dB over this anyway, so if you are saying 65W is not enough you really ought to be going to 200W+ !!
cheers & good luck, mark 😎
Gabster, you have undoubtedly realised there is no "highest-fi", or not yet anyway. You need to define what you are looking for a little more carefully. Can you accept SE class A, with the large hardware and enormous heat? Or, do you require an amp which is class AB, ie. less power inefficient? Do you really need 100W, or have you just plucked this figure out of the air because it sounds impressive?
If you are using reasonably efficient speakers (>=85dB/W, say) then you are actually unlikely to need this sort of power and your choices will be much wider.
Class A/B the Leach amps are a good choice. I used them for many years and they are honest performers (the one I built in 1978 is still working in my garage). Some of the other suggestions would likewise be suitable.
Class A, I obviously have a bias to Passlabs but there are others out there (JLH for example http://www.gmweb.btinternet.co.uk/jlh1996.htm ). The A40 (http://www.passdiy.com/pdf/a40.pdf) is push-pull class A so not quite as inefficient as the Alephs and is an excellent amplifier. Once you traverse the 100W mark here (Aleph4, 1.2) you are talking really serious hardware and heat. Most of us only listen in the <10W range anyway, so the remainder is just room warming!
If you do really need some serious power, see the 1kW beast on Anthony's site:
http://www.aussieamplifiers.com/poweramp.htm
Rod Elliot makes some good amplifiers:
http://sound.westhost.com/project03.htm
There are also a number of very good articles on his site. (he also has an interesting section on the "Death of Zen" ..... but that's stoning material around here 🙂 )
If you can handle the heat, I would recommend the Alephs above the Leach amps (and virtually all "quasi-class-A" amps I have heard), simply they sound better.
Unless you really do need the additional power I would recommend the Aleph5 (65W - OR, the Volksamp equivalent) .... the hardware is just a little more bearable. Remember the Aleph4 (100W) only gains a couple of dB over this anyway, so if you are saying 65W is not enough you really ought to be going to 200W+ !!
cheers & good luck, mark 😎
mefinnis,
I must admit, I'm not the most experienced listener/builder around here and I must rely on people like you to make good choices for my listening experiences. On the other hand, that's why I'm more likely to look at specs before building an amp.
At this point I'm almost set on the Aleph. I would like to build the Aleph 4 with its 100 watts, but if you say that the 5 sounds better I will let my greed away.
Finally, is this common for a class A amplifiers to have 1% distortion (talking about the Aleph). So far, all the amps I've seen perform better? (I'm aware that good sounding amps don't mean low distortion).
thanks
😉
I must admit, I'm not the most experienced listener/builder around here and I must rely on people like you to make good choices for my listening experiences. On the other hand, that's why I'm more likely to look at specs before building an amp.
At this point I'm almost set on the Aleph. I would like to build the Aleph 4 with its 100 watts, but if you say that the 5 sounds better I will let my greed away.
Finally, is this common for a class A amplifiers to have 1% distortion (talking about the Aleph). So far, all the amps I've seen perform better? (I'm aware that good sounding amps don't mean low distortion).
thanks
😉
Nelson states THD specs at maximal output - ie, this is "worst figure" from this amp. Many manufacturers do not do this, they will give you the figure at 1W, which is dramatically better. The THD specs for Nelson's amps in the typical listening range are vastly less than the maximum.
Yes, there are amplifiers with better THD specs, no question but there is more to it than that. What order harmonic the distortion is, xover effects etc are all "different" to the ear and are not strictly comparable by "average percentage".
You really should try to listen to the different amplifier topologies if possible. If you want a class A amp with less THD then Rod Elliot's would be a good bet.
In selecting power you really do need to appreciate that a 100W amplifier does not produce 2xdB of a 50W amplifier. The scale is logarithmic and as I stated eralier you only gain a few dB going from 60W to 100W. The Aleph5 in general listening will (I expect - haven't finished my A5 yet) be "near indistinguishable" from the A4 and it will cost you less to build, be smaller, use less power and generate less heat.
The call is yours. I only built the A4.v.A5 because I found some toroids at this supply voltage cheap and at that stage I was under the false illusion both amps used the same compliment of FETs.
Starting from scratch, now, I would build the A5, or, if I really needed the extra power I would build A1.2 monoblocks (with a reduced FET compliment by choice). The later however will cost lots more money and be 3x the size overall.
Rememeber, there are really only 2 things which matter, 1) the music at the end, and 2) that you have fun in the process 🙂
cheers, mark
Yes, there are amplifiers with better THD specs, no question but there is more to it than that. What order harmonic the distortion is, xover effects etc are all "different" to the ear and are not strictly comparable by "average percentage".
You really should try to listen to the different amplifier topologies if possible. If you want a class A amp with less THD then Rod Elliot's would be a good bet.
In selecting power you really do need to appreciate that a 100W amplifier does not produce 2xdB of a 50W amplifier. The scale is logarithmic and as I stated eralier you only gain a few dB going from 60W to 100W. The Aleph5 in general listening will (I expect - haven't finished my A5 yet) be "near indistinguishable" from the A4 and it will cost you less to build, be smaller, use less power and generate less heat.
The call is yours. I only built the A4.v.A5 because I found some toroids at this supply voltage cheap and at that stage I was under the false illusion both amps used the same compliment of FETs.
Starting from scratch, now, I would build the A5, or, if I really needed the extra power I would build A1.2 monoblocks (with a reduced FET compliment by choice). The later however will cost lots more money and be 3x the size overall.
Rememeber, there are really only 2 things which matter, 1) the music at the end, and 2) that you have fun in the process 🙂
cheers, mark
Thanks for your advice. Very appreciated.
Do you know other websites that
http://members.ozemail.com.au/~dkfinnis/passlabs/index.htm
that talks about the construction of these amps. On this site, the link to the circuit schematics is not working.
I've also search the pass labs forum without luck (the link that was refering to pass' website is down too).
regards
Do you know other websites that
http://members.ozemail.com.au/~dkfinnis/passlabs/index.htm
that talks about the construction of these amps. On this site, the link to the circuit schematics is not working.
I've also search the pass labs forum without luck (the link that was refering to pass' website is down too).
regards
Since that's my website ...... which link where exactly?
I had to redo many of them when Nelson upgraded his website and split-off Passdiy.com. Obviously I have missed something 🙁
See the download page:
http://members.ozemail.com.au/~dkfinnis/passlabs/aleph4_download.htm
I have checked these just now and all appears in order.
regards, mark
I had to redo many of them when Nelson upgraded his website and split-off Passdiy.com. Obviously I have missed something 🙁
See the download page:
http://members.ozemail.com.au/~dkfinnis/passlabs/aleph4_download.htm
I have checked these just now and all appears in order.
regards, mark
I agree totaly - as long as it pleases you, either to make it or listen to it, then it's got to be worthwhile.
😀
😀
Just a couple points i'd like to make:
1/ Simple distortion measurements are pretty much meaningless. for example: A single THD number tells you almost nothing (although a number that is very low will scare me off), but a picture of the harmonic distortion's spectrum can tell you alot.
Lots of feedback to get a really low number, tends to just move the distortion elsewhere (ie presence of more objectionable higher harmonics -- time distortion).
2/ from listening, not building. Amps with 1 (SE) or 2 (PP) devices usually sound better than amps with more devices. Corallary to this is that the best sounding amps usually have lower power.
dave
1/ Simple distortion measurements are pretty much meaningless. for example: A single THD number tells you almost nothing (although a number that is very low will scare me off), but a picture of the harmonic distortion's spectrum can tell you alot.
Lots of feedback to get a really low number, tends to just move the distortion elsewhere (ie presence of more objectionable higher harmonics -- time distortion).
2/ from listening, not building. Amps with 1 (SE) or 2 (PP) devices usually sound better than amps with more devices. Corallary to this is that the best sounding amps usually have lower power.
dave
lohk said:
I'm pretty sure Hugh has said as much himself. The circuit is not radical, just that care has been taken with the execution.
dave
Go to www.uspto.gov and search for patent # 5,892,398
This is the halcro patent. Study the schematic carefully.
The tricks used to get the distortion down very very low
are pretty clever.
Yet a few of the reviews of this amplifier say its sound
does not corellate with the extremely low distortion.
This is the halcro patent. Study the schematic carefully.
The tricks used to get the distortion down very very low
are pretty clever.
Yet a few of the reviews of this amplifier say its sound
does not corellate with the extremely low distortion.
Wow, quite an old thread that got resurrected.
I know I'll get flammed for this, 😀 but I sort of like the Opti-MOS amp. I've read both of Randy Sloan's books and like his design philosophy.
However, I have Self's book on order, and he makes some different conclusions regaring BJT's and MOSFET's.
Net result is I'm still 'arm chairing' but will be building something soon!
Randy Sloan makes the point that THD should be measured both at 1Khz AND at 20Khz. I rarely see the latter. Without both measurements it's essentially a meaningless statistic. Also, IMHO slew rate is important - it's one distinguishing characteristic in feed back designs.
"The topology is a mirror-image, L-MOSFET type, incorporating a dual-cascode input stage (for ultra low-noise operation - SNR better than -100 dB) and a unique type of "nested feedback" system incorporated into the VA stage, providing exceptional linearity and stability. It is capable of providing continuous 200-watt RMS outputs into 8-ohm speaker loads. Distortion performance is excellent (THD less than 0.005% @ 1 KHz @ 0 dBm), with high-frequency distortion reduced to only about 0.014% (20 KHz @ 0 dBm) due to the implementation of a double two-pole compensation network. Input impedance is 10 Kohm. Input stage zener overload protection and an improved multislope output overload and short-circuit protection network provide virtual "bullet-proof" reliability. "
Seal Electronics - Opti-MOS
I think this is the one I'll build.
I haven't read Self's books yet but he has some PCB's and info on his web site:
Douglas Self's Amplifier Institute
I know I'll get flammed for this, 😀 but I sort of like the Opti-MOS amp. I've read both of Randy Sloan's books and like his design philosophy.
However, I have Self's book on order, and he makes some different conclusions regaring BJT's and MOSFET's.
Net result is I'm still 'arm chairing' but will be building something soon!
Randy Sloan makes the point that THD should be measured both at 1Khz AND at 20Khz. I rarely see the latter. Without both measurements it's essentially a meaningless statistic. Also, IMHO slew rate is important - it's one distinguishing characteristic in feed back designs.
"The topology is a mirror-image, L-MOSFET type, incorporating a dual-cascode input stage (for ultra low-noise operation - SNR better than -100 dB) and a unique type of "nested feedback" system incorporated into the VA stage, providing exceptional linearity and stability. It is capable of providing continuous 200-watt RMS outputs into 8-ohm speaker loads. Distortion performance is excellent (THD less than 0.005% @ 1 KHz @ 0 dBm), with high-frequency distortion reduced to only about 0.014% (20 KHz @ 0 dBm) due to the implementation of a double two-pole compensation network. Input impedance is 10 Kohm. Input stage zener overload protection and an improved multislope output overload and short-circuit protection network provide virtual "bullet-proof" reliability. "
Seal Electronics - Opti-MOS
I think this is the one I'll build.
I haven't read Self's books yet but he has some PCB's and info on his web site:
Douglas Self's Amplifier Institute
Sound? Music?
I'm interested to hear how the Opti Mos actually sounds, no mention of this on the Seal Site!!! Have you built one yet?
Mozfet
I'm interested to hear how the Opti Mos actually sounds, no mention of this on the Seal Site!!! Have you built one yet?
Mozfet
- Status
- Not open for further replies.