Now this is surprising , also brings up a great debate. Do we want something added to our music by the amp ?? On the AX "blameless " I listen to now , I can add just a little bass to make it sound like carlo's "ES" (pix 1).
I did a little "blind test" today. I took me screwdriver and adjusted my tail current (ltp) to where i thought it sounded best. I then measured across my 2.2k tail resistor to find 8.8v 8.8/2200 = exactly 4ma !! I then did a LT FFT with stepped tail currents at 1k/10k , especially at 10k , the H3 dropped by 15db with increasing tail current. Best was 2- 2.2 ma (4- 4.4ma total). How did this happen ?? I just turned the trimmer and heard it sound better and stopped. Could I actually hear this? 😱 And I thought amps all sound the same. 🙄 Both the scope and LT show perfect linearity and ppm THD across the whole range of the "CCS adjust" (2ma - 5ma) , but 4ma sounds best.
If an amp would be dry and sterile , would that not be good. To ABSOLUTELY magnify the source ...adding NOTHING. All that aside , How the speakers interact with the OPS ? Damping factor , How low the impedance is between power supply and woofer , thickness of traces , wires gauge are other factors in the "sound".
The crazy amp with "too many devices" above will even do under 10 ppm at 100K !! It is different , why not try it ??
C2C , wife is getting mad at me again , parts are "migrating" back to "her" kitchen yet again .. too dang cold in the doghouse still... Brrrrrrr! 😀
Edit: would a halcro with a 12ax7 tube preamp sound uninvolving ?? would you just not hear the tube ?
OS
Attachments
Last edited:
What Hugh is suggesting, is very much at odds with reviews like those at the attached link.
Halcro dm38 power amplifier | Stereophile.com
Thanx , I needed to see that , (out comes the Sprint layout CAD 🙂 ) Imagine the lowly mongrel CX ... a 15K $ frontend .. ha 😀 They describe it as "fast" .. I agree ,with 27p millers , 200kHZ- .002% ... double HA 😀
screw "Goldmoney"..
Ps .. still playin' with it in LT .. the circuit has a lot of leeway (non-critical parameters ) of course , semi-precise matching for the 3 small (to-92) differentials would be of benefit. seems to "self balance" itself .. offset is not an issue with this one either (single ended primary stage.)
this even describes my ps250/ax setup ,
humm can it get better.. gonna build it !!!
OS
Last edited:
Yep. Most of my views are at odd with the accepted view. Should I apologise, or will you guys humor me? I enjoy baying at the moon!
Merry Christmas and all that.....
Hugh
Merry Christmas and all that.....
Hugh
But if I write in proper English, Andrew, ALL is lost in translation......
BTW, that should be ODDS, not just ODD.
Hugh
BTW, that should be ODDS, not just ODD.
Hugh
Pete,
I wouldn't use that 12AX7. Lousy tube. I'd use a 6SL7. Yes, it would totally change the sound, I would expect for the better. You need to very carefully select the operating point for the tube, however. 15 years ago I built an amp using a SE mosfet output stage with a tube front end. One of the most magical amps I've ever heard.
Heavily coloUred, of course (that's for Androo), but very involving and great on jazz.
Hugh
I wouldn't use that 12AX7. Lousy tube. I'd use a 6SL7. Yes, it would totally change the sound, I would expect for the better. You need to very carefully select the operating point for the tube, however. 15 years ago I built an amp using a SE mosfet output stage with a tube front end. One of the most magical amps I've ever heard.
Heavily coloUred, of course (that's for Androo), but very involving and great on jazz.
Hugh
Last edited:
Hello
Audio magazines equipments reviews are not always neutral, most of those magazines will do a bit or a lot of indirect advertisings with their equipments reviews articles. So we have to be carefull with those reviews since advertisers are a good part of the audio magazines incomes.
Reading comments from users say more about an audio gear.
Bye
Gaetan
Last edited:
My preamp is still in the imaginary mode. That can change rapidly around here If I get in one of those "building frenzies" . I'm going to do the strait "leech" (below 1)topology first before that "halcro thingy" I posted. To go with that , the PS50 (50kuf - smaller power supply) and my new stereo OP board are being designed/built now. 😎
I worked on carlos's ES/ (BX) that I have. Changed a few degeneration resistors , fine tuned the currents in the input stage and VAS. It really sounds nice now , so close to the AX. Better bass with the BX , better highs and slightly better imaging with the AX , STILL , almost too close to call as to which is "better"
.. even with "mountain ears". The BX with the 200w Op stage is SO loud , but clear... even with complex source material.Does anyone know the best way to implement an DC offset adjustment on the below "leech" topology ?? please do comment...
OS
Attachments
Pete,
I'd work on the CCSs supplying the input stage. Simply disconnect junction of R21 and R33 from ground, connect to an integrating, inverting opamp with 15V rails. That should do it. You should have a reasonable variation of reference voltage off the LEDs in the CCSs, but if you don't you can always add a 15R resistor in series with each LED.
Nice and remote from the signal path.....
Hugh
I'd work on the CCSs supplying the input stage. Simply disconnect junction of R21 and R33 from ground, connect to an integrating, inverting opamp with 15V rails. That should do it. You should have a reasonable variation of reference voltage off the LEDs in the CCSs, but if you don't you can always add a 15R resistor in series with each LED.
Nice and remote from the signal path.....
Hugh
So , just a simple adjustable ground reference for the 2 CCS's ?? I want a IC free design that will be easily integrated to the capacitance multipliers of the mongrel pb250/120. You were right about "character" .. the bootstrap is becoming my favorite of the 2 blameless varients. I also have found another "secret" in how to deploy nested feedback to a bootstrapped topology. It's all about the sound... 🙂 , isn't it. The journey continues.. even as some say there are no differences , I say B_$.
OS
Just 1 or both of the LX1.0 CCS's , if made adjustable , will make for a dandy offset adjust. Thanks Hugh , you got me to the right path. Just one extra part , simplicity is bliss.
OS
OS
The best way is what Hugh suggests, driving the bootstrap IS A BAD IDEA (many commercial manufactures use this INCORRECT way, messing the sound at low spectrum), you will inject a spurious signal in the input path. It could be even more worse if the biasing signal is inside de audio BW (it could make your amplifier a subsonic oscillator). It is a MUCH better way to bias the current at the input CS's. You could manage to do a transistorized integrator but it will not have the precision (nor thermal stability) of a fast IC ie: LM318 that will set the DC offset below 1mV, without messing audio.
Cheers
Arturo
Cheers
Arturo
OS
Why are the BJTs better for the front end differentials than the JFets? I am just wondering out loud. It seems that is holding true for all you do.😕
You are have been very very prolific in the last month, keep up the Great work. 😱
😉Lots of what you have posted has educated me so much THANKS🙂
Why are the BJTs better for the front end differentials than the JFets? I am just wondering out loud. It seems that is holding true for all you do.😕
You are have been very very prolific in the last month, keep up the Great work. 😱
😉Lots of what you have posted has educated me so much THANKS🙂
BJT's are cheaper , easier to source , and are just as good as the JFET. I tried the K192A's I have (below) and they needed 22r degeneration to match the Bjt ,sounded exactly the same and were more of a pain to match. They worked very well , but at 30-40c apiece, why bother ??
I am still educating myself .. if any of it enriches others , cool !!
BTW , I am using MPSA18's and SS9014's , both are VERY quiet .. much more than 2sc1845's or BC556's. SO , my BX is at least 6-10db quieter than an ES , the only drawback is the need for a cascode (2 more devices + 1cap / 2 resistors) ... but that also allows for very high rails as well (70-90V is BX's max). 😎
OS
I am still educating myself .. if any of it enriches others , cool !!
BTW , I am using MPSA18's and SS9014's , both are VERY quiet .. much more than 2sc1845's or BC556's. SO , my BX is at least 6-10db quieter than an ES , the only drawback is the need for a cascode (2 more devices + 1cap / 2 resistors) ... but that also allows for very high rails as well (70-90V is BX's max). 😎
OS
Attachments
Last edited:
Thanks Arturo!
Pete,
You can dispense with the IC if you wish, but the servo adjustment, using a single transistor, will act on one CCS only. Run a 15V zener supply off the negative rail to handle about 10mA. Then place a 4K7 resistor in series with a 5K trimpot from the zener anode (the -15V supply) to the emitter of a 2N5551 npn device. The collector of this device is connected via one 15K resistor to the positive supply CCS LED, and the existing cap to the positive rail. Now the base of the device is connected to output via a 47K resistor, and to a 47uF cap to ground. This is the integrator. Adjust the trimpot for correct offset and you are away....
Cheers,
Hugh
Pete,
You can dispense with the IC if you wish, but the servo adjustment, using a single transistor, will act on one CCS only. Run a 15V zener supply off the negative rail to handle about 10mA. Then place a 4K7 resistor in series with a 5K trimpot from the zener anode (the -15V supply) to the emitter of a 2N5551 npn device. The collector of this device is connected via one 15K resistor to the positive supply CCS LED, and the existing cap to the positive rail. Now the base of the device is connected to output via a 47K resistor, and to a 47uF cap to ground. This is the integrator. Adjust the trimpot for correct offset and you are away....
Cheers,
Hugh
Last edited: