ABO said:Hi Epupa,
I'm afraid I don't have any friends with even the slightest interest in electronics or audio. I can sometimes use some equipment of our local univeristy. But only on occasion.
I took a look at another amps:
Elektor crescendo ME: vas-bias 35 mA, inputcap of mosfets 1700 pF
Zenquito: VAS bias 10 mA, input cap. 1200 pF
Indeed VAS bias is somewhat low. OTOH, input capacitance is a more linear. At 20 kHz the reactance of 1200 pF is some 6600 ohms. This cannot be neglected at all. I think that improvement can be made here.
Now I really wished I had more facilities. I can't see myself experimenting with other transistors and emitter resistances with only a DMM.
I still can't understand why all those French Zenquito builders are so quiet. Come on guys, you should know. Help us out here!
Salut,
Mini Crescendo :
http://www.mif.pg.gda.pl/homepages/tom/files/70MOS.gif
http://www.jls-info.com/julien/audio/photosJF/crescendo.jpg
http://perso.wanadoo.fr/pol.bct/JFG/Dsc02036.jpg
Elektor Crescendo VAS Bias BF469/BF470 12~15mA si je me souviens bien ...
2SK135/2SJ50 Distorsion thermique nulle 60mA
2SK1058/2SJ162 Distorsion thermique nulle 110mA
Hexfet 60
http://www.mif.pg.gda.pl/homepages/tom/files/120MOS.gif
Nonante 90
http://www.mif.pg.gda.pl/homepages/tom/files/140IGB.gif
http://www.promelec.ru/pdf/bf469_471.pdf
http://perso.wanadoo.fr/jm.plantefeve/2N3440.pdf
Pour Zenquito il n'y a aucuns soucis au niveau des trans bipolaire 2N3440/2N5416 si ce n'est qu'ils chauffent beaucoup et donc il faut leurs mettre un radiateur conséquent H= ~ 40mm. Il faut également ajuster le courant en modifiant R 68 ~ 150 ohms en fonction du Vgs des Mos (K1058/J162) et du courant de repos souhaité dans l'étage de puissance (110mA/Mos classe AB ou bien plus pour de la classe A )
A mon goût le problème ne vient pas de la, mais plutôt d'une mauvaise adaptation d'impédance avec la source ...
@ + JF
Posted by JFG: (Translation Babelfish):
For Zenquito it n'y has aucuns concern on the level of the trans bipolar 2N3440/2N5416 if this n'est qu'ils heats much and thus it is necessary their to put a consequent radiator H = ~ 40mm. It is also necessary to adjust the current by modifying R 68 ~ 150 ohms according to Vgs of Mos (K1058/J162) and the quiescent current desired in l'étage of power (110mA/Mos class AB or more for class A) To my taste the problem does not come from, but rather d'une bad adaptation d'impédance with the source...
HI JFG,
I don't know exactly what this means, what is the source here? If you mean that the amp may have trouble driving the speakers, well, that's not it. A huge improvement comes from improving power supply for the VAS stages. I'm not done with it yet, but a simple low-pass filter with diode makes a HUGE (wish I had larger font size here) difference. Everything, really.
I will also do some experiments with the audible effects of larger bias currents, for both VAS stage and output stage. Note that I currently use 9 mA (VAS) and 350 mA bias current in the output.
BTW: I saw pictures where you French guys use giant heatsinks for the VAS transistors. I can't seem to find them in Holland.
For Zenquito it n'y has aucuns concern on the level of the trans bipolar 2N3440/2N5416 if this n'est qu'ils heats much and thus it is necessary their to put a consequent radiator H = ~ 40mm. It is also necessary to adjust the current by modifying R 68 ~ 150 ohms according to Vgs of Mos (K1058/J162) and the quiescent current desired in l'étage of power (110mA/Mos class AB or more for class A) To my taste the problem does not come from, but rather d'une bad adaptation d'impédance with the source...
HI JFG,
I don't know exactly what this means, what is the source here? If you mean that the amp may have trouble driving the speakers, well, that's not it. A huge improvement comes from improving power supply for the VAS stages. I'm not done with it yet, but a simple low-pass filter with diode makes a HUGE (wish I had larger font size here) difference. Everything, really.
I will also do some experiments with the audible effects of larger bias currents, for both VAS stage and output stage. Note that I currently use 9 mA (VAS) and 350 mA bias current in the output.
BTW: I saw pictures where you French guys use giant heatsinks for the VAS transistors. I can't seem to find them in Holland.
Hi ABO,
Has you say it JFG has already good stuff about building ampli. He helps many guys (me too) in tehe wiring / grounding conception.
Here's my ZENQUITO EVO 5x50W. I drive Fostex full range speaker with. I buikd a BOSOZ from N.Pass to go with. And sound's quite good : clear, détailed, strong bass, impressive sound stage in depth from front to rear. I have haven't noticed any bad thing.
"A mon goût le problème ne vient pas de la, mais plutôt d'une mauvaise adaptation d'impédance avec la source ...
" JFG means here, i think, that the outpout inpedance from your cd drive (if its a cd drive) is not correct matched with your zenquito impedance entry.
Marc
Has you say it JFG has already good stuff about building ampli. He helps many guys (me too) in tehe wiring / grounding conception.
Here's my ZENQUITO EVO 5x50W. I drive Fostex full range speaker with. I buikd a BOSOZ from N.Pass to go with. And sound's quite good : clear, détailed, strong bass, impressive sound stage in depth from front to rear. I have haven't noticed any bad thing.
"A mon goût le problème ne vient pas de la, mais plutôt d'une mauvaise adaptation d'impédance avec la source ...
" JFG means here, i think, that the outpout inpedance from your cd drive (if its a cd drive) is not correct matched with your zenquito impedance entry.
Marc
To ABO : Before building of any Mosfet amp, is good to read something about theory. Very good explanation you get from Erno Borbelly's articles published in The Audio Amateur and Wireless World or now on his own pages. After reading you will have quite different look at this amp - simplicity isn't all .
.I see what you mean.
Even before I built this amp I knew that the design was 'flawed' in a technical sense. So are some other very popular designs such as Alephs. Still, people seem to really love them.
Being an engineer myself I am very reluctant to believe that an amplifier with bad specification could sound good. On the other hand, you can never be sure before you try it yourself. After some inquiring I decided to try Zenquito: It has the same cult status (in France at least) as Pass amps and does not dissipate 500W, ideal for me.
What I've learned so far is that Zenquito, with modifications, actually sounds better than my previous amp. But there seem to downsides too. I won't tell before I've learned more about Zenquito. I'm sure you'll find that it lacks transparency with complex classical music
I have the opportunity to compare it with a 2500 EUR commercial amp (tonight) and in a few weeks I can compare it with a AKSA55 Nirvana from someone on the forum. These should be very interesting experiences. Remember, I am just starting to learn about audio.
All in all, I think your objective approach is the right one. This is the only way to make any real progress. Tinkerers like Peter Daniel and AKSA (with all respect, no kidding) are able to get the best out of a given topology, but it stops right there.
Things for me to find out are:
- is it really true that amps with bad specs can sound better?
- What then?
BTW: You claim that you can tell how an amp sounds right from the schematic (I actually do believe you). Could you describe the sound of my amp?
Thanks for the interest
Even before I built this amp I knew that the design was 'flawed' in a technical sense. So are some other very popular designs such as Alephs. Still, people seem to really love them.
Being an engineer myself I am very reluctant to believe that an amplifier with bad specification could sound good. On the other hand, you can never be sure before you try it yourself. After some inquiring I decided to try Zenquito: It has the same cult status (in France at least) as Pass amps and does not dissipate 500W, ideal for me.
What I've learned so far is that Zenquito, with modifications, actually sounds better than my previous amp. But there seem to downsides too. I won't tell before I've learned more about Zenquito. I'm sure you'll find that it lacks transparency with complex classical music
I have the opportunity to compare it with a 2500 EUR commercial amp (tonight) and in a few weeks I can compare it with a AKSA55 Nirvana from someone on the forum. These should be very interesting experiences. Remember, I am just starting to learn about audio.
All in all, I think your objective approach is the right one. This is the only way to make any real progress. Tinkerers like Peter Daniel and AKSA (with all respect, no kidding) are able to get the best out of a given topology, but it stops right there.
Things for me to find out are:
- is it really true that amps with bad specs can sound better?
- What then?
BTW: You claim that you can tell how an amp sounds right from the schematic (I actually do believe you). Could you describe the sound of my amp?
Thanks for the interest
Bad sound better ?
It is neverending discussion, which was here manytimes and future will be again ( now goes in thread " Hawksford " ). When people take listening like impression, where distortion can to bring " nice coloration ", every technical arguments falling . Impession is impression and is different for each one, it is the same like women . You are asking me about sound of this amp. I never heard it, so judge, how I hit it . As I see on schematic, connection is allsymetrical, which is good. On input are jfets, which bring to reproduction little more naturality in opposite to BJT's. At low level of listening and by jazz or folk music, will be soud likeable, but probably not so transparent ( resistors in gates of mosfets have too high value - I'm using ten times lower ). Problem started, when you listen classical music, which must be listened on high level ( symphony ). When volume rise, sound will be less and less transparent ( sorry, but my english is not so good ) and started lossing of stereo and space " picture ", which is caused by intermodulation distortion and by long settling time, 'cos reserve of current in VAS / driver stage isn't too high. Have I thruth ?
It is neverending discussion, which was here manytimes and future will be again ( now goes in thread " Hawksford " ). When people take listening like impression, where distortion can to bring " nice coloration ", every technical arguments falling
. Impession is impression and is different for each one, it is the same like women . You are asking me about sound of this amp. I never heard it, so judge, how I hit it . As I see on schematic, connection is allsymetrical, which is good. On input are jfets, which bring to reproduction little more naturality in opposite to BJT's. At low level of listening and by jazz or folk music, will be soud likeable, but probably not so transparent ( resistors in gates of mosfets have too high value - I'm using ten times lower ). Problem started, when you listen classical music, which must be listened on high level ( symphony ). When volume rise, sound will be less and less transparent ( sorry, but my english is not so good ) and started lossing of stereo and space " picture ", which is caused by intermodulation distortion and by long settling time, 'cos reserve of current in VAS / driver stage isn't too high. Have I thruth ?ABO said:I will also do some experiments with the audible effects of larger bias currents, for both VAS stage and output stage. Note that I currently use 9 mA (VAS) and 350 mA bias current in the output.
BTW: I saw pictures where you French guys use giant heatsinks for the VAS transistors. I can't seem to find them in Holland.
Upupa Epops said:
Oh YES, ajustez R de 68 ohms à 150 ohms collector VAS pour que I soit entre 10 et 15mA et
http://www.diyaudio.com/forums/attachment.php?s=&postid=415756
OK, avec 3 heatsinks collés l'un sur l'autre c'est parfait ... 50° à 55° est tout à fait normal
@ + JF
Hi Epupa and others,
It is difficult for me to comment about sound quality at this time. Last night I picked up my new speakers from the original designer. The speakers are completely different (far better) than my old speakers. Today I will assemble the crossover and mount the drivers etc. I will do no modifications this weeks. I need to become familiar with my new sound first.
We did do a comparison with my Zenquito (modded) and his. To my surprise the difference was very small. The 'reference', a Vecteur integrated amp, had better bass (deeper and more punchy, slam), but at normal level listening the Zenquito held his own very well. Its transperancy at high freqencies was very good. We couldn't play at high volumes because of sleeping kids. All in all it was a comforting experience.
Makes me wonder. In France most DIYers use those big filter caps with M5 bolts. These have far better current capability and lower ESR than the caps I use. My experiments show that the amp is extremely sensitive to PSU quality. This may be the cause of the difference in sound quality between me and the French.
I still feel that in any case a separate supply voltage for the VAS stages is desirable since it makes so much difference.
After I have burned-in to my new speakers, I will try some improvements in PSU, higher bias and lower gate resistors.
PS: I have only one recording of a complete symphonical orchestra (Denon CD for testing CDP's). But I will compare differences between low and high volumes.
It is difficult for me to comment about sound quality at this time. Last night I picked up my new speakers from the original designer. The speakers are completely different (far better) than my old speakers. Today I will assemble the crossover and mount the drivers etc. I will do no modifications this weeks. I need to become familiar with my new sound first.
We did do a comparison with my Zenquito (modded) and his. To my surprise the difference was very small. The 'reference', a Vecteur integrated amp, had better bass (deeper and more punchy, slam), but at normal level listening the Zenquito held his own very well. Its transperancy at high freqencies was very good. We couldn't play at high volumes because of sleeping kids. All in all it was a comforting experience.
Makes me wonder. In France most DIYers use those big filter caps with M5 bolts. These have far better current capability and lower ESR than the caps I use. My experiments show that the amp is extremely sensitive to PSU quality. This may be the cause of the difference in sound quality between me and the French.
I still feel that in any case a separate supply voltage for the VAS stages is desirable since it makes so much difference.
After I have burned-in to my new speakers, I will try some improvements in PSU, higher bias and lower gate resistors.
PS: I have only one recording of a complete symphonical orchestra (Denon CD for testing CDP's). But I will compare differences between low and high volumes.
ABO said:
Effectively PSU needs to be on top, as far wiring and grounding. It's more an evidence when spearkers are hard to drive and need high amout of current. The most suitable case arrangement is when transformer and rectifier are séparate from the amp modules with a sheet of steel. The pictures after illustrate this.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Quality of filter caps have clearly their importance in providing dynamique, bass depth and so on also as PI filtration (CLC or CRC filtering).
Zenquito has great sound. I compare it with several commercial productions and every time the zenquito overkill them. I compare also the zenquito with à Zen4 (a good one constructed) and one more time I prefere the zenquito, same clear sound, but more sound stage in font-end depth, more dynamique, stronger bass....But remind, if amp need good parts they have to be well dimensionned, and arranged in the case. Wirring arrangement need particular attention too. The amp need to be in adaquation with the other hifi system parts (player- preamp - speakers).
Marc
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.