Nordic,
It certainly frees up a lot of space.
Carlos,
Ah, the brain. It has to come into the discussion at some point. Funny thing the brain, I spend endless hours laughing at mine. Just one big chemical stew. Never trust anything that involves the brain.
Big capacitors: I think I'll give up on these big capacitors after the voltage reg. The board that Nordic is designing allows for them to be used, I'll stick with the small and simple ones.
It certainly frees up a lot of space.
Carlos,
Ah, the brain. It has to come into the discussion at some point. Funny thing the brain, I spend endless hours laughing at mine. Just one big chemical stew. Never trust anything that involves the brain.
Big capacitors: I think I'll give up on these big capacitors after the voltage reg. The board that Nordic is designing allows for them to be used, I'll stick with the small and simple ones.
Hi Carlos, when I do offboard input caps, I tend to make them stretch from the RCA to the PCB, this time, I will need to add an inch or 2 of cable to the cap... but nothing long...
I just completed removeing the 100nf caps, and the sound is now more to my likeing and my speakers even sound more like what I'm used to hearing from them... only mod left was the 10mfd on the output pairs...
Have to wait for wife to wake up, before I can turn the volume up, but it sounds, a litttle less warm now, a bit more neutral... I'm sure the bass is still there... I can feel it with some of the music on the tv, at low volume... watching saturday night live rerun now...
Just to avoid confusion for other member.. this mod was to the standard DX..
P.S. Carlos, I remeber when I tested using magnet wire to hook the RCA to the DX input, it was totaly microphonic... never mind picking up RF etc... just tapping the wire reproduced the sound on the speakers...
To me high frequency sensitivity is something we held over from our survival instincts in nature... although there might be alot of HF sounds in say a rainforrest.. the ones that our body knows to listen for ar hissing snakes, insects... etc... most things in nature capable of hissing can hurt you... which is why we respond so well to warm,mellow sounds, like that of a father's voice cooing a baby.
I just completed removeing the 100nf caps, and the sound is now more to my likeing and my speakers even sound more like what I'm used to hearing from them... only mod left was the 10mfd on the output pairs...
Have to wait for wife to wake up, before I can turn the volume up, but it sounds, a litttle less warm now, a bit more neutral... I'm sure the bass is still there... I can feel it with some of the music on the tv, at low volume... watching saturday night live rerun now...
Just to avoid confusion for other member.. this mod was to the standard DX..
P.S. Carlos, I remeber when I tested using magnet wire to hook the RCA to the DX input, it was totaly microphonic... never mind picking up RF etc... just tapping the wire reproduced the sound on the speakers...
To me high frequency sensitivity is something we held over from our survival instincts in nature... although there might be alot of HF sounds in say a rainforrest.. the ones that our body knows to listen for ar hissing snakes, insects... etc... most things in nature capable of hissing can hurt you... which is why we respond so well to warm,mellow sounds, like that of a father's voice cooing a baby.
Small and simple:
Big caps after VReg gone.
Big input caps gone.
Twiddled some bits to make room for the Vbe capacitor, there it is, right in the middle of the board. Messy.
Added pads so output ground can either be run with separate lines to star ground or can be linked, as with the DX, to the single board ground.
Big caps after VReg gone.
Big input caps gone.
Twiddled some bits to make room for the Vbe capacitor, there it is, right in the middle of the board. Messy.
Added pads so output ground can either be run with separate lines to star ground or can be linked, as with the DX, to the single board ground.
Attachments
Nordic,
For running the wires from the RCA+Cap to the board you can always use a three wire braid, one wire hot, two wires ground to the pcb which will provide you with a good shield. I typically use 30awg kynar wire for this which has always worked like a charm at eliminating any interference, I use the same setup for my interconnects which has been the best thus far without spending megabucks on interconnects. But for the the average 2-6 inches you shouldnt have issues, just keep those suckers away from the AC voltage runs.
Colin
For running the wires from the RCA+Cap to the board you can always use a three wire braid, one wire hot, two wires ground to the pcb which will provide you with a good shield. I typically use 30awg kynar wire for this which has always worked like a charm at eliminating any interference, I use the same setup for my interconnects which has been the best thus far without spending megabucks on interconnects. But for the the average 2-6 inches you shouldnt have issues, just keep those suckers away from the AC voltage runs.
Colin
Yep, from a layout point of view, I made the AC connections the lowest point on the back feeding the wire down the middle, on the floor, to the PSUs behind the FACE. Iinput connection is the the highest point.
As for the DX amp... those of you who built the standard versions, please go power up your soldering irons...
Remove the outpu pair's electrolytics and wire in those 10000uf caps, I have no doubt, you will find it a drastic improvement, like I do. Should take you all of 10 minutes... please just do it... I would describe the sound as detailed and smooth at the same time, warm and neutral at the same time....it should be called the paradox amp.
I do not know if it is hte result of the extra peak power delivery of the caps, the lower impendance or some other combination...I used plain 85C 10000ufd 63V Jamicons...
As for the DX amp... those of you who built the standard versions, please go power up your soldering irons...
Remove the outpu pair's electrolytics and wire in those 10000uf caps, I have no doubt, you will find it a drastic improvement, like I do. Should take you all of 10 minutes... please just do it... I would describe the sound as detailed and smooth at the same time, warm and neutral at the same time....it should be called the paradox amp.
I do not know if it is hte result of the extra peak power delivery of the caps, the lower impendance or some other combination...I used plain 85C 10000ufd 63V Jamicons...
Hi Nordic,
Your findings are exactly as I expected.
It is essential to have 10mF PSU electrolytics connected to the main pcb rails so that these cannot become modulated by bass current draw with respect to the board ground.
Once you have them fitted, with say parallel connected 220uF + 22uF + 470nF then the series 82 ohm resistors which drop (lose)voltage to the VAS may also be linked out !
On the DX just leave the 220uF+100nF fitted and try linking out the 82 ohm resistors.
With these large (low ESR) capacitors fitted and paralleled with the smaller electrolytics/poly there ought not even be need for series regulators, so no loss of output power either !
Cheers ........... Graham.
Your findings are exactly as I expected.
It is essential to have 10mF PSU electrolytics connected to the main pcb rails so that these cannot become modulated by bass current draw with respect to the board ground.
Once you have them fitted, with say parallel connected 220uF + 22uF + 470nF then the series 82 ohm resistors which drop (lose)voltage to the VAS may also be linked out !
On the DX just leave the 220uF+100nF fitted and try linking out the 82 ohm resistors.
With these large (low ESR) capacitors fitted and paralleled with the smaller electrolytics/poly there ought not even be need for series regulators, so no loss of output power either !
Cheers ........... Graham.
Supply rail capacitors
Graham
Your suggestions make a lot of sense for the standard version of DX, and will undoubtedly improve performance.
However,It has been the experience of several DiyAudio members in Sydney, that even with 30,000uF per supply rail. that there are further audible improvements to be obtained from regulating the front end of most Power amplifiers.
Your suggestion also goes against Carlos's findings, and his preferred schematic. Carlos has made his preference quite clear in several posts. I suggest that constructors of Carlos's latest amplifier follow Carlos's design, until they have a fully working amplifier. The more adventurous should then be free to trial
Graham's recommendations, and decide for themselves.
Regards
SandyK
Graham
Your suggestions make a lot of sense for the standard version of DX, and will undoubtedly improve performance.
However,It has been the experience of several DiyAudio members in Sydney, that even with 30,000uF per supply rail. that there are further audible improvements to be obtained from regulating the front end of most Power amplifiers.
Your suggestion also goes against Carlos's findings, and his preferred schematic. Carlos has made his preference quite clear in several posts. I suggest that constructors of Carlos's latest amplifier follow Carlos's design, until they have a fully working amplifier. The more adventurous should then be free to trial
Graham's recommendations, and decide for themselves.
Regards
SandyK
I'm also a little scared of doing the mods to the front end, but I have been toying with the idea of testing a seperate front end power supply...
However as far as the standard DX go... I would say the large caps on the output transistors should be default... I forgot, the other non standard item I have are 10uf non-botique non-polar electrolytic types... I have nice polyester caps, but nothing in that range...
I think Carlos himself eluded to some of the DX's "problems" by stateing that he used 0.1uf as input caps and also tweaked the feedback chain with another cap to increase trebble... which I found waaaaay to trebble when I tested it... maybe when I'm old and half deaf to HF, would I be able to live with that...
I plan to remove the input cap on my amp, as my preamp allready has 4u7on the output.
The large output stage decoupling increase, is like adding light to a slightly dark picture. And I believe it is simple enough to go with the orinial simple goals of the amp.
I too have read where Carlos said he tried large caps there and didn't find improvement... but I believe it may be due to some unseen coincidance, and invite him to reinvestigate (with his ears) standard DX with large caps on output pair.
However as far as the standard DX go... I would say the large caps on the output transistors should be default... I forgot, the other non standard item I have are 10uf non-botique non-polar electrolytic types... I have nice polyester caps, but nothing in that range...
I think Carlos himself eluded to some of the DX's "problems" by stateing that he used 0.1uf as input caps and also tweaked the feedback chain with another cap to increase trebble... which I found waaaaay to trebble when I tested it... maybe when I'm old and half deaf to HF, would I be able to live with that...
I plan to remove the input cap on my amp, as my preamp allready has 4u7on the output.
The large output stage decoupling increase, is like adding light to a slightly dark picture. And I believe it is simple enough to go with the orinial simple goals of the amp.
I too have read where Carlos said he tried large caps there and didn't find improvement... but I believe it may be due to some unseen coincidance, and invite him to reinvestigate (with his ears) standard DX with large caps on output pair.
Carlos's amplifier
Nordic
The Silicon Chip 100W/Ch. ULD amplifier actually uses a separate regulated front end supply obtained from extra windings on the transformer. I have heard a "blueprinted" version of this amplifier. Firstly with 30,000uF on each supply rail, and it was indeed very good. At a later date, I heard the same amplifier, still with 30,000uF per rail, but now using the original SC design's regulated front end supply. It was quite noticeably more detailed, with a further improvement to the soundstage.
"However as far as the standard DX go... I would say the large caps on the output transistors should be default..."
AGREED !
SandyK
Nordic
The Silicon Chip 100W/Ch. ULD amplifier actually uses a separate regulated front end supply obtained from extra windings on the transformer. I have heard a "blueprinted" version of this amplifier. Firstly with 30,000uF on each supply rail, and it was indeed very good. At a later date, I heard the same amplifier, still with 30,000uF per rail, but now using the original SC design's regulated front end supply. It was quite noticeably more detailed, with a further improvement to the soundstage.
"However as far as the standard DX go... I would say the large caps on the output transistors should be default..."
AGREED !
SandyK
Hi Sandy,
I was trying to help Nordic further improve his already constructed DX design. I added explanation for findings and suggested what to watch out for; you concur !
Whether front end regulation is worthwhile must relate to the input stage+VAS arrangement/devices chosen/wiring layout and their combined capability for good CMRR.
I have commented on amplifier circuitry in general, and I have most certainly NOT 'gone against' Carlos' findings from his preferred schematic, nor have I suggested that anyone be more adventurous with his circuit.
Indeed I would not be so foolish, for this recipe has yet to be savoured at the audio banquet before others can learn about our faithful connoisseur's delicacy !!!
Cheers ......... Graham.
I was trying to help Nordic further improve his already constructed DX design. I added explanation for findings and suggested what to watch out for; you concur !
Whether front end regulation is worthwhile must relate to the input stage+VAS arrangement/devices chosen/wiring layout and their combined capability for good CMRR.
I have commented on amplifier circuitry in general, and I have most certainly NOT 'gone against' Carlos' findings from his preferred schematic, nor have I suggested that anyone be more adventurous with his circuit.
Indeed I would not be so foolish, for this recipe has yet to be savoured at the audio banquet before others can learn about our faithful connoisseur's delicacy !!!
Cheers ......... Graham.
By all means I believe in starting projects with first versions... even if it is just to tag along for the learning, during evolution...
If I built an amp sounding as nice as my DX sounds now... it would be long before I tried designing something better....
The standard DX is like a meal with too much pudding... hard to find anyone complaining, cause it is so nice... but probably not as good as you should eat everyday... would be nice for those alcohol induced parties though... sounds just like a club's woofers...
Good luck with the move and all the other adjustments, maybe it is a good time now to start eating healthy, and getting off a floor early, and takeing the stairs the rest of the way.... so we can have you around for a long time still...
If I built an amp sounding as nice as my DX sounds now... it would be long before I tried designing something better....
The standard DX is like a meal with too much pudding... hard to find anyone complaining, cause it is so nice... but probably not as good as you should eat everyday... would be nice for those alcohol induced parties though... sounds just like a club's woofers...
Good luck with the move and all the other adjustments, maybe it is a good time now to start eating healthy, and getting off a floor early, and takeing the stairs the rest of the way.... so we can have you around for a long time still...
Re: Re: Nordic...this text is my Christmas gift...in advance to you
The mk vii Crimson amps were all quasi complementary.
The 70's, 80's & 90's versions of the higher powered 1704vii all had 2n3773 output devices and had 1k8 or 1k3 & 1nF as RF filter (~90kHz to 120kHz).
The later mk viii went to high speed complementary output devices and a 130kHz RF filter.
Sorry to be off topic.gaetan8888 said:
The mk vii Crimson amps were all quasi complementary.
The 70's, 80's & 90's versions of the higher powered 1704vii all had 2n3773 output devices and had 1k8 or 1k3 & 1nF as RF filter (~90kHz to 120kHz).
The later mk viii went to high speed complementary output devices and a 130kHz RF filter.
hehe...for sure there are some sligth deviations from the main idea
But, of course, people is free to engage their brain without copy the circuit exactly as published..... forum free...people free to move.... and they are not wrong.
The reality is that my supply has more than 22000 uf each rail..and the wires to the board are 10 inches long...and also i have a good condenser into the board power input, 3300uf to decouple the wire resistance that is miliohms or microohms..i do not know...the reality, is that i have already 25000uf each rail, as supply is almost assembled "over" the board....because of that i felt no need of big condensers folks...because there are already big condenser in that place...just accept those short (and very thick) wires are not disturbing as i believe.
Of course...having already enormous capacitance..and short wiring... the insertion of 10000uf or any good capacitance will not result different..because capacitance will change from big to big...from good to good....or for excelent to excelent...it is the same as 6 and half dozen....so.... in my case..condenser not needed there and made no difference.
Graham is thinking about his amplifier.... his supply...his class A sucking current...his cables....his own reality of construction....mine is different, Nordic is different.
But, as usual, Nordic may be sensitive to things the old folk here do not perceive...as removing 100n condenser may be better...ehehe.... i think Nordic may be illuded...or home noise floor reduced...was from ear change place.... well..... he may be rigth too...but those things were tested.
Also i think that Nordic, as i have suggested, could try the GEM, as he is perceiving things that prove that Graham is that good designer we all know he is...so...why to insist into modifications into Dx standard..there are other options, High Resolution and High Resolution II...and also the GEM is an exceptional amplifier.
The idea to replace, increase, decrease condensers is good... a i have made here...and those things teach us a lot...but confusions can exist and people be fooled....maybe i am being folled, as i perceived that condenser there are not needed...when i have, in the reality..the equivalent of condenser there.
regards,
Carlos
But, of course, people is free to engage their brain without copy the circuit exactly as published..... forum free...people free to move.... and they are not wrong.
The reality is that my supply has more than 22000 uf each rail..and the wires to the board are 10 inches long...and also i have a good condenser into the board power input, 3300uf to decouple the wire resistance that is miliohms or microohms..i do not know...the reality, is that i have already 25000uf each rail, as supply is almost assembled "over" the board....because of that i felt no need of big condensers folks...because there are already big condenser in that place...just accept those short (and very thick) wires are not disturbing as i believe.
Of course...having already enormous capacitance..and short wiring... the insertion of 10000uf or any good capacitance will not result different..because capacitance will change from big to big...from good to good....or for excelent to excelent...it is the same as 6 and half dozen....so.... in my case..condenser not needed there and made no difference.
Graham is thinking about his amplifier.... his supply...his class A sucking current...his cables....his own reality of construction....mine is different, Nordic is different.
But, as usual, Nordic may be sensitive to things the old folk here do not perceive...as removing 100n condenser may be better...ehehe.... i think Nordic may be illuded...or home noise floor reduced...was from ear change place.... well..... he may be rigth too...but those things were tested.
Also i think that Nordic, as i have suggested, could try the GEM, as he is perceiving things that prove that Graham is that good designer we all know he is...so...why to insist into modifications into Dx standard..there are other options, High Resolution and High Resolution II...and also the GEM is an exceptional amplifier.
The idea to replace, increase, decrease condensers is good... a i have made here...and those things teach us a lot...but confusions can exist and people be fooled....maybe i am being folled, as i perceived that condenser there are not needed...when i have, in the reality..the equivalent of condenser there.
regards,
Carlos
Attachments
My supply is the standard DX supply with about 10mfd per rail (2x 4700u), and decoupling was first 47u then 220uf (could hear no change). I suspect that even a jump to 3300 or 4700 might start to show results...
My transformers are pretty far from the PSU's, but I don't think that is a problem, and the wires from PSUs are 3 to 4 inches...
Lol, I have about the same capacitance on the output pair, as in the whole PSU... crazy world, I tell you.
I'm pretty sure about the improvement.. of course I have no improvement meter. But it is inconceiveable to me that you can't hear the diffirence...
Except, with the previous setup.. something kept feeling too overpowering...yet somehow withholding information, now its much more to my likeing... I don't think I will do any other changes between now and whenI do the HDII.. of course I should give this amp a good listen for a few days...
When you get to the point where you don't feel your amp needs immediate improvements, it is a good place to be...
Listening to the chipamp yesterday, realy opened my eyes to what I was missing in the sound again... man the DX in any form eats a chipamp... but the critters are very detailed sounding... but not as smooth... a bit more of a cringe factor at high frequencies...
Thanks Carlos
You put the ooz is smooz
My transformers are pretty far from the PSU's, but I don't think that is a problem, and the wires from PSUs are 3 to 4 inches...
Lol, I have about the same capacitance on the output pair, as in the whole PSU... crazy world, I tell you.
I'm pretty sure about the improvement.. of course I have no improvement meter. But it is inconceiveable to me that you can't hear the diffirence...
Except, with the previous setup.. something kept feeling too overpowering...yet somehow withholding information, now its much more to my likeing... I don't think I will do any other changes between now and whenI do the HDII.. of course I should give this amp a good listen for a few days...
When you get to the point where you don't feel your amp needs immediate improvements, it is a good place to be...
Listening to the chipamp yesterday, realy opened my eyes to what I was missing in the sound again... man the DX in any form eats a chipamp... but the critters are very detailed sounding... but not as smooth... a bit more of a cringe factor at high frequencies...
Thanks Carlos
You put the ooz is smooz
I am talking about HD-II...other schematics supplied, the options, red frame, blue
frame and others, have different ideas...one of them, included big condensers into the schematic as standard...so...there are options to all tastes and all kind of ears.
Into the High Resolution II, was removed the big caps into the input...the amplifier left side, and into the thread's opener, your old uncle, the official schematic.
The blue and green frame schematic, that one were Ecat made a nice green and blue design, is also using small condenser into the input side of the amplifier.. and this was done because the electronic regulation included.... into this one, big condensers into the input is a overkill, not needed, wasting money into big condensers that will produce the same work as the 220uf multiplied by the series regulator gain (capacitance multiplier... Rodd Elliot)
Nothing bad related condenser.. i like them, you like them, so,.enjoy them.... also is nice when we switch the power off...into low volumes, my prototype delayed 42 seconds to stop playing with nice, undistorted sound.
An this is also nice...to see how big is our reserve when playing low volumes.
I think i am talking about High Resolution, and you are talking about the standard Dx amplifier...i have tried those big condensers into the standard Dx and no result...but, as i have explained, my supply is very good.
You had different results..so...solder those parts there and do not move them... as you could discover an interesting thing.
Now Dx amplifier, By Nordic, his upgrade, has ozz and smoozz
The official amplifier is still published into Greg pages.... i suggest you to try ozz and smoozz and be happy.
regards,
Carlos
frame and others, have different ideas...one of them, included big condensers into the schematic as standard...so...there are options to all tastes and all kind of ears.
Into the High Resolution II, was removed the big caps into the input...the amplifier left side, and into the thread's opener, your old uncle, the official schematic.
The blue and green frame schematic, that one were Ecat made a nice green and blue design, is also using small condenser into the input side of the amplifier.. and this was done because the electronic regulation included.... into this one, big condensers into the input is a overkill, not needed, wasting money into big condensers that will produce the same work as the 220uf multiplied by the series regulator gain (capacitance multiplier... Rodd Elliot)
Nothing bad related condenser.. i like them, you like them, so,.enjoy them.... also is nice when we switch the power off...into low volumes, my prototype delayed 42 seconds to stop playing with nice, undistorted sound.
An this is also nice...to see how big is our reserve when playing low volumes.
I think i am talking about High Resolution, and you are talking about the standard Dx amplifier...i have tried those big condensers into the standard Dx and no result...but, as i have explained, my supply is very good.
You had different results..so...solder those parts there and do not move them... as you could discover an interesting thing.
Now Dx amplifier, By Nordic, his upgrade, has ozz and smoozz
The official amplifier is still published into Greg pages.... i suggest you to try ozz and smoozz and be happy.
regards,
Carlos
Hi G, no the my_ref C does not fall in the standard chipamp category... the chip is used as a current pump, as well as haveing some DC protection etc... except of course its all chips... bar a few transistors in the protection parts...
It is amazeing how your ears can deceive you, after some time it sounds pretty good... but A/B versus the DX it is a little harsh sounding. As far as my testing is concerned, other gainclones need not even apply...
It is amazeing how your ears can deceive you, after some time it sounds pretty good... but A/B versus the DX it is a little harsh sounding. As far as my testing is concerned, other gainclones need not even apply...
- Status
- Not open for further replies.