Hi
We often see soundstage (deep, width) as a criteria for sound evaluation. Is it part of one of the above criteria?
For criteria #5 I would expect any properly built amplifier (and preceding stages) to have a 10?
Thanks
Fab
Last edited:
AUDIO AMPLIFIER ASSESSMENT SYSTEM
Folks,
Fab has suggested soundstage is very important. ON reflection, he is completely right, so I have changed the assessment system as follows:
We select each individual rating out of 10, the multiplied by the #weighting, then all added to give a 100 figure for a perfect amp. There are now four aspects with weighting #2, and the last two are #1, making a total of #10. I would not expect a 100% figure for any amplifier.
1. Musicality (whether it sounds 'natural', like real music) #2
2. Engagement (whether it brings a smile and taps your foot) #2
3. Soundstage (width AND depth; both are important) #2
4. Visceral impact (whether it really gives you the rock concert force and power) #2
5. Clarity (no explanation for this!) #1
6. Noise (how much hash or hum is evident during silent musical passages, should be 10 but if not mark it down heavily) #1
I have marked down Musicality and Engagement to #2, added Soundstage with #2, and taken one from Clarity and added it to Impact. Give each of these aspects a rating out of 10, then multiply each rating for its respective weighting, then add the all five aspects together to give a figure out of 100. Ratings can be given down to halves, that is, 7.5/10 or 4.5/10.
Hugh
Folks,
Fab has suggested soundstage is very important. ON reflection, he is completely right, so I have changed the assessment system as follows:
We select each individual rating out of 10, the multiplied by the #weighting, then all added to give a 100 figure for a perfect amp. There are now four aspects with weighting #2, and the last two are #1, making a total of #10. I would not expect a 100% figure for any amplifier.
1. Musicality (whether it sounds 'natural', like real music) #2
2. Engagement (whether it brings a smile and taps your foot) #2
3. Soundstage (width AND depth; both are important) #2
4. Visceral impact (whether it really gives you the rock concert force and power) #2
5. Clarity (no explanation for this!) #1
6. Noise (how much hash or hum is evident during silent musical passages, should be 10 but if not mark it down heavily) #1
I have marked down Musicality and Engagement to #2, added Soundstage with #2, and taken one from Clarity and added it to Impact. Give each of these aspects a rating out of 10, then multiply each rating for its respective weighting, then add the all five aspects together to give a figure out of 100. Ratings can be given down to halves, that is, 7.5/10 or 4.5/10.
Hugh
Last edited:
For Alpha BB running at 3amps bias current and XRKaudio Transient Perfect FAST speakers.
1. Musicality (whether it sounds 'natural', like real music) #2 2x9=18
2. Engagement (whether it brings a smile and taps your foot) #2 2x10=20
3. Soundstage (width AND depth; both are important) #2 2x10=20
4. Visceral impact (whether it really gives you the rock concert force and power) #2 2x9=18
5. Clarity (no explanation for this!) #1 1x10=10
6. Noise (how much hash or hum is evident during silent musical passages, should be 10 but if not mark it down heavily) #1 9x10=9
Total = 95
1. Musicality (whether it sounds 'natural', like real music) #2 2x9=18
2. Engagement (whether it brings a smile and taps your foot) #2 2x10=20
3. Soundstage (width AND depth; both are important) #2 2x10=20
4. Visceral impact (whether it really gives you the rock concert force and power) #2 2x9=18
5. Clarity (no explanation for this!) #1 1x10=10
6. Noise (how much hash or hum is evident during silent musical passages, should be 10 but if not mark it down heavily) #1 9x10=9
Total = 95
I think ALL must be interested about BIG electrolytic capacitors with low ESR, high Ripple and cheap!
I commented about the expensive Mundorf MLGO here: Aksa Lender P-mos Hybrid Aleph (ALPHA) Amplifier
[PDF] http://www.mundorf.com/english%201.1/Broschuere%20Einzelseiten/MLGO.pdf
KEMET PEH200 Series, +85°C
[PDF] http://www.kemet.com/Lists/ProductCatalog/Attachments/391/KEM_A4034_PEH200.pdf
Cheap here:
http://www.elektrodump.nl/en/51-elco-low-voltage
Elco / KEMET low voltage 47000uF 63V 85ºC
-> http://www.elektrodump.nl/en/elco-low-voltage/2222-elco-low-voltage-47000uf-63v.html € 10
I commented about the expensive Mundorf MLGO here: Aksa Lender P-mos Hybrid Aleph (ALPHA) Amplifier
[PDF] http://www.mundorf.com/english%201.1/Broschuere%20Einzelseiten/MLGO.pdf
KEMET PEH200 Series, +85°C
[PDF] http://www.kemet.com/Lists/ProductCatalog/Attachments/391/KEM_A4034_PEH200.pdf
Cheap here:
http://www.elektrodump.nl/en/51-elco-low-voltage
Elco / KEMET low voltage 47000uF 63V 85ºC
-> http://www.elektrodump.nl/en/elco-low-voltage/2222-elco-low-voltage-47000uf-63v.html € 10
Last edited:
I done my calculus by the book,scope signal generator pretty basics differential (ltp) test
Want to ask way the 50ma ksa992fbu?
Do you feel it's it would help to have some regulation for the "front"
What's wrong with usual 100mA 2sa970gr/bl. I think almost everybody have some in their boxes. Real ones.
I have never come across the ksa992 in any commercial amp that uses as ltp
Could you help me relate to the sound of bb versus.aleph2 I'm familiar with.
Depends organic aluminium caps been used for decades where needed as tantalum.
Tonza75,
I have used other PNP here (like 2N5401 or SMT version MMBT5401 on the Aksa Lender preamp) and it works about just as well. Feel free to substitute with 2N5401 (which I know everyone has - it will probably work fine, Hugh should comment though if this is ok).
However, Hugh specified it and since the KSA992 is the complement to the KSC1845, it would make sense that it is used as the KSC1845 is the one used for the VAS tied to the two legs of the LTP. They only cost $0.16ea (10 or more) and readily available (39,500 in stock at Mouser) from usual vendors. Whereas 2sa970 is not even on the Mouser catalog.
http://www.mouser.com/ds/2/149/KSA992-88789.pdf
I have used other PNP here (like 2N5401 or SMT version MMBT5401 on the Aksa Lender preamp) and it works about just as well. Feel free to substitute with 2N5401 (which I know everyone has - it will probably work fine, Hugh should comment though if this is ok).
However, Hugh specified it and since the KSA992 is the complement to the KSC1845, it would make sense that it is used as the KSC1845 is the one used for the VAS tied to the two legs of the LTP. They only cost $0.16ea (10 or more) and readily available (39,500 in stock at Mouser) from usual vendors. Whereas 2sa970 is not even on the Mouser catalog.
http://www.mouser.com/ds/2/149/KSA992-88789.pdf
Last edited:
Okay I have over 200 pcs of 5401 left in my box. Tons of smd too I will certainly substitute the KSC1845 I see you used to126 maybe a 2sc3502e or 2sc2911 in bb a so maybe KSC1845 I like it's a decent bjt in current mirror.
How about bc 547c what did you choose here in final ver.
Thanks.
Watch the leg pin orientations. On BB, the KSC3503 TO126 was used. The BC547 on the Aleph CCS is of no consequence. Almost any NPN in the same family can work assuming max ratings are not exceeded. BC546/7/8/9/550 can probably all work.
http://www.mouser.com/ds/2/149/BC547-888524.pdf
http://www.mouser.com/ds/2/149/BC547-888524.pdf
I will solder like ic pins in and try I have so much bjt left of I could use smds for ltp on a those mini adapter pcbs I have at store. Thanks again.
Many amps built when I was younger and had the passion, not so active as repair jobs takes time amps too but more computer, mobile, consumer device maintenance.
Not a bad idea to standardize scoring for audio systems. However I have to point something out.
I've listened to multiple systems separately and at the time I rated them fairly well. But then over time it becomes clear one of them is really not that good. I think this is due to reference point shift. When we say something is "good" we are always determining that based on prior experience, with a bias toward recent events. If the reference points shifts, then a scoring system such as this will tend to give flaky results. Thus it would require either a very experienced reviewer to score accurately or it would require scores to be made based on comparisons rather than plucking a number form the top of your head that "sounds about right".
So when I try to rate equipment from now on, I don't try to say something is good or bad. The most meaningful comparison I can make is to rate it as better or worse than something specific I've already listened to. This also forces you to organize your experience in terms of sound quality, which helps as rudimentary error prevention mechanism.
There are certain quality milemarkers as well. For instance can you hear the Ds in a choral latin piece? Maybe you can even tell what they're saying? Can you tell when the pianist plays two notes but releases one sooner? When testing changes on a system I'll often notice that certain consonants become more legible and I can then hear the lyrics better in some tracks. These are things I take note of which seem a bit more solid than rating a system by choosing numbers.
Numerical ratings are a very slippery kind of rating and in my experience they often fail, and rely too much on the operator being consistent in eyeballing those numbers.
Rating with a hierarchy rather than an arbitrary number system has the benefit that you are not presuming the limits of quality by rating something high; you can always put something higher at the top without having to go above 100% or split the top 99% into 99.1%, 99.2%...
What do you do when your 96% from today seems more like 60% after listening to a different system? I don't like the percentages, I would rather just say X is better than Y, and Y is better than Z. And with Y you can hear the quiet choral Ds, but with X you can hear those and ALSO understand clearly what they are saying - in Latin.
Another weakness of rating systems is when a system sounds so good, you don't know how to make a valid complaint against it.
I've listened to multiple systems separately and at the time I rated them fairly well. But then over time it becomes clear one of them is really not that good. I think this is due to reference point shift. When we say something is "good" we are always determining that based on prior experience, with a bias toward recent events. If the reference points shifts, then a scoring system such as this will tend to give flaky results. Thus it would require either a very experienced reviewer to score accurately or it would require scores to be made based on comparisons rather than plucking a number form the top of your head that "sounds about right".
So when I try to rate equipment from now on, I don't try to say something is good or bad. The most meaningful comparison I can make is to rate it as better or worse than something specific I've already listened to. This also forces you to organize your experience in terms of sound quality, which helps as rudimentary error prevention mechanism.
There are certain quality milemarkers as well. For instance can you hear the Ds in a choral latin piece? Maybe you can even tell what they're saying? Can you tell when the pianist plays two notes but releases one sooner? When testing changes on a system I'll often notice that certain consonants become more legible and I can then hear the lyrics better in some tracks. These are things I take note of which seem a bit more solid than rating a system by choosing numbers.
Numerical ratings are a very slippery kind of rating and in my experience they often fail, and rely too much on the operator being consistent in eyeballing those numbers.
Rating with a hierarchy rather than an arbitrary number system has the benefit that you are not presuming the limits of quality by rating something high; you can always put something higher at the top without having to go above 100% or split the top 99% into 99.1%, 99.2%...
What do you do when your 96% from today seems more like 60% after listening to a different system? I don't like the percentages, I would rather just say X is better than Y, and Y is better than Z. And with Y you can hear the quiet choral Ds, but with X you can hear those and ALSO understand clearly what they are saying - in Latin.
Another weakness of rating systems is when a system sounds so good, you don't know how to make a valid complaint against it.
I really like the idea of a numerical rating system as an attempt to quantify the subjective qualities of an amp. What a useful metric for somebody like me who is trying to pick an amp to build from the dizzying array here on DIYA.
Yet I can see Keantoken's point, too. We all have different reference points based on what we have heard before, and especially what we have on hand to make direct comparisons with. Perhaps each reviewer's total score should be weighted based on the total number of amps he has owned, before his score gets rolled into a group average. Lol. No, more likely we will just use our own discretion for deciding whose reviews we trust the most.
Yet I can see Keantoken's point, too. We all have different reference points based on what we have heard before, and especially what we have on hand to make direct comparisons with. Perhaps each reviewer's total score should be weighted based on the total number of amps he has owned, before his score gets rolled into a group average. Lol. No, more likely we will just use our own discretion for deciding whose reviews we trust the most.
Yes, Stellar, I agree. I can see Kean's reasonable comments, yet, I am trying hard to objectivise the subjectives using a 'global' system which can be argued either way but should give a reasonably predictable figure related to the sound quality. This is an absurd game, I recognise, but we have to do this eventually since THD is a poor indicator of sound excellence.
Gedlee proposed a more objective system with his index, but it's not come off yet, this is a simple attempt to pick off the issues that make a audio playback system 'musical', whatever the hell that is....... groan.......
Hugh
Gedlee proposed a more objective system with his index, but it's not come off yet, this is a simple attempt to pick off the issues that make a audio playback system 'musical', whatever the hell that is....... groan.......
Hugh
Last edited:
Hallo Gabor and All.
May i can help a little.
I have running one of Hughs favorites FQA40N25 in the MoFo, driven by single tiode from 5687/7119 Tube, actually.
It seems a robust kandidat, not to hard biase yet and sounds very clear here.
FQA bought for Big Quasi, finnished Boards have to wait....like other Projects.
Alpha has to wait any longer.....
Regards.
May i can help a little.
I have running one of Hughs favorites FQA40N25 in the MoFo, driven by single tiode from 5687/7119 Tube, actually.
It seems a robust kandidat, not to hard biase yet and sounds very clear here.
FQA bought for Big Quasi, finnished Boards have to wait....like other Projects.
Alpha has to wait any longer.....
Regards.