Nico Ras said:
I say you are right, Nico.
There are only very subtle differences.
Which make me say it is, as with other details in audio amplifiers:
Both CCS & Bootstrap can be used with very good results.
It is very possible there are circuits that are best suited for one or the other,
but in most cases .....
a good implemented Bootstrap is better than a less optimal use of current source.
And a good CCS is better than a Bootstrap that is not very well done.
From John:
You can achieve this, it ain't so difficult.
String a high value resistor from emitter of the CCS device to output. That should do it.... arrange the resistor for a small 5% of current through the VAS.
For an 8ma CCS and 50V rails, choose R = 120K.
This will raise and lower the VAS current with signal level, injecting some H2. It will cost some resolution, of course, but it will have other effects you may like.
Hugh
You can achieve this, it ain't so difficult.
String a high value resistor from emitter of the CCS device to output. That should do it.... arrange the resistor for a small 5% of current through the VAS.
For an 8ma CCS and 50V rails, choose R = 120K.
This will raise and lower the VAS current with signal level, injecting some H2. It will cost some resolution, of course, but it will have other effects you may like.
Hugh
AKSA said:
Hi Hugh,
Yes, similar to my bootstrap resistor on the LTP of my amp.
Have you done this before?
BTW, I haven't listens with and without it as you have suggested, but I plan on doing it.
John,
No, not this variant, but I've bench tested it, and it works fine, and I use a variation in the AKSA to power the LTP.
It increases distortion, I won't say more than that.
Hugh
No, not this variant, but I've bench tested it, and it works fine, and I use a variation in the AKSA to power the LTP.
It increases distortion, I won't say more than that.
Hugh
AKSA said:
Hi Hugh,
not a bad idea at all I have both amps built up and will play around a little this week-end, you may just solved the tastes of both worlds.
Cheers
Nico
Another subtle truth..manipulating H2,3,5,7 is what gives
an amp it's character. Same with the BS/CCS loading, its not
the actual topology , but it's implementation that determines
the final product.
Strange thing is that the bootstrap would display dynamic
characteristics with regards to freq./load/ and the physics
of the capacitor itself (black gates anyone).
CCS's are for the most part "static" , ideally showing the
same distortion characteristics throughout the audio band.
lately, I have left both "camps"
, prefering a topology thatadds no "character" giving me the option of adding it
afterwards.
OS
Aagh !!!
Hugh, I thought this was a secret only I knew...
This is the technique I have applied to my guitar amplifier, and it makes an amazing difference in tone.
Hugh, I thought this was a secret only I knew...
This is the technique I have applied to my guitar amplifier, and it makes an amazing difference in tone.
ostripper said:
Another subtle truth..manipulating H2,3,5,7 is what gives
an amp it's character. Same with the BS/CCS loading, its not
the actual topology , but it's implementation that determines
the final product.
Strange thing is that the bootstrap would display dynamic
characteristics with regards to freq./load/ and the physics
of the capacitor itself (black gates anyone).
CCS's are for the most part "static" , ideally showing the
same distortion characteristics throughout the audio band.
lately, I have left both "camps"
adds no "character" giving me the option of adding it
afterwards.
OS
I take it you will share your secret.
AKSA said:
Hi Hugh,
I simulation what you suggested and contrary to popular belief it had the opposite effect, it actually improved THD by a fair margin especially at higher frequencies with almost an order of magnitude at 10kHz.
Feeding back directly on the emitter of the CCS proved a little less effective than feeding back into a split emitter resistor as shown below. At low frequencies <100 Hz there were no improvement.
This modification is sort of localized negative feedback into the CCS which IMPROVED the THD not worsen it.
I think there is merit in Hugh's "bootstrapped" CCS and should be investigated further.
Nice suggestion Hugh
I would like to make the mods on the actual amp now, but unfortunately things are a bit hectic before the Southern Hemisphere shuts down for their summer holidays.
Nico
Attachments
Serendipity is a wonderful thing...
Nico, this is exactly what I have done with my amplifer !
It is all bipolar at the moment, with TIP142/TIP147 outputs (blech),
but the next step is to use a JFET/bipolar diff. input (JFET for the input, bipolar for the feedback) followed by a MOSFET VAS drving the Darlingtons.
As you say, the biggest improvement is in the high-frequency IM distortion.
Nico, this is exactly what I have done with my amplifer !
It is all bipolar at the moment, with TIP142/TIP147 outputs (blech),
but the next step is to use a JFET/bipolar diff. input (JFET for the input, bipolar for the feedback) followed by a MOSFET VAS drving the Darlingtons.
As you say, the biggest improvement is in the high-frequency IM distortion.
Hi Hugh,
managed a few moments this afternoon to make the mod to the CCS and only had a few minutes to listen to it before having to do the grocery run with my wife before the holiday makers flock in their thousands to the coast were we live.
No comparison yet, just this amp and it sounds quite fabulous, the slight "tizzyness" is gone completely.
managed a few moments this afternoon to make the mod to the CCS and only had a few minutes to listen to it before having to do the grocery run with my wife before the holiday makers flock in their thousands to the coast were we live.
No comparison yet, just this amp and it sounds quite fabulous, the slight "tizzyness" is gone completely.
Hi Guys,
I cannot make up my mind, both topologies have merit neither outperforming the other by any margin whatsoever. The character of an amplifier I would put down to two aspects:
1) Manipulation of harmonic power distribution
2) Time relationship between the parts of a complex stimulus.
By the second statement I mean how the individual components of an instrument, voice, etc. is shifted in time changing the character of the sound.
My conclusions and you may differ from this view is that the bootstrap and the CCS manipulates the overtones differently and this would affect the nature of the sound e.g. warmer, cold, clinical, fast, slow, etc.
The second issue is the character of the sound, i.e. although this is very subtle, it is when the same instrument sounds different through different amps.
Anyway guys from my fooling around I can say that I preferred the bootstrap in this simple amplifier and that this was an interesting exercise (for me anyway) and I hope that it was useful.
Just a matter of final interest, the overshoot in the square wave in the amp using CCS was cured by a small cap across the NFB resistor.
But I must warn that this need be done using an instrument (I used a spectrum analyzer) because this cap combined with the miller cap could easily cause a very high frequency oscillation (in the MHz range) and you will never know it.
Anyway guys thanks for your contributions and comments it was appreciated. And Hugh thanks for dropping by I am sorry that a previous thread had a dig at you.
Besides I drive a 1985 BMW, not because I cannot afford a brand new one, I actually like this one.
Kindest regards
Nico
I cannot make up my mind, both topologies have merit neither outperforming the other by any margin whatsoever. The character of an amplifier I would put down to two aspects:
1) Manipulation of harmonic power distribution
2) Time relationship between the parts of a complex stimulus.
By the second statement I mean how the individual components of an instrument, voice, etc. is shifted in time changing the character of the sound.
My conclusions and you may differ from this view is that the bootstrap and the CCS manipulates the overtones differently and this would affect the nature of the sound e.g. warmer, cold, clinical, fast, slow, etc.
The second issue is the character of the sound, i.e. although this is very subtle, it is when the same instrument sounds different through different amps.
Anyway guys from my fooling around I can say that I preferred the bootstrap in this simple amplifier and that this was an interesting exercise (for me anyway) and I hope that it was useful.
Just a matter of final interest, the overshoot in the square wave in the amp using CCS was cured by a small cap across the NFB resistor.
But I must warn that this need be done using an instrument (I used a spectrum analyzer) because this cap combined with the miller cap could easily cause a very high frequency oscillation (in the MHz range) and you will never know it.
Anyway guys thanks for your contributions and comments it was appreciated. And Hugh thanks for dropping by I am sorry that a previous thread had a dig at you.
Besides I drive a 1985 BMW, not because I cannot afford a brand new one, I actually like this one.
Kindest regards
Nico
Nico,
Thanks for an open mind, trying out a strange idea, and Bala, serendipity is a wonderful thang.....
More than twenty years ago I came up with a weird way of altering the cam dwell on an operating gasoline engine. It was an expensive idea, required a lot of complex machining.
To my astonishment, the VVT on a Ferrari started to use the same concept about five years ago.
This is proof that when an idea comes of age, several people 'invent' it at much the same time.
This CCS concept surprises me. I wonder why it reduces distortion?
Very odd.......
I wasn't aware you'd had a dig at me, Nico? In any event, I've forgotten, so it can't have been very nasty!
A 1985 Beemer, that would be a 528, perhaps, one of their best?
Cheers,
Hugh
Thanks for an open mind, trying out a strange idea, and Bala, serendipity is a wonderful thang.....
More than twenty years ago I came up with a weird way of altering the cam dwell on an operating gasoline engine. It was an expensive idea, required a lot of complex machining.
To my astonishment, the VVT on a Ferrari started to use the same concept about five years ago.
This is proof that when an idea comes of age, several people 'invent' it at much the same time.
This CCS concept surprises me. I wonder why it reduces distortion?
Very odd.......
I wasn't aware you'd had a dig at me, Nico? In any event, I've forgotten, so it can't have been very nasty!
A 1985 Beemer, that would be a 528, perhaps, one of their best?
Cheers,
Hugh
AKSA said:
Hi Hugh,
I did not have a dig at you but was following the thread that started interestingly and then got a bit out of hand. I did not feel it my place to contribute because I did not believe that to be the purpose of this forum.
I think each audio designer whether professional or amateur makes what he/she perceives to sounds good to him/her. The popularity of their designs is an indication of how many other individuals share the same passion.
BTW it is a 735i was my company car then and still is. I kept it in an immaculate condition and it still serves me as an obedient dog.
Kindest regards
Nico
Hugh sorry I did not conclude your 120K suggestion. Adding the resistor reduced the 2H significantly. Well, by about 8 dB, which reduced the overall THD.
However, without the resistor the harmonics raised steadily with increasing frequency, but with the resistor the higher order products increased at a lower rate with higher frequency.
Okay that probably does not make much sense, but if I could explain this in RF terms, it would appear to improve the third order intercept point or maybe reducing the OLG. It is not feasible to measure the OLG of this circuit since the input transistor emitter is derived from the output.
I think one should also bare in mind that harmonics are not treated as individual signals by a non-linear transfer function, but could be a result of sum and difference frequencies (due to mixing).
Adding of this resistor may not have the same results in more complex designs.
However, without the resistor the harmonics raised steadily with increasing frequency, but with the resistor the higher order products increased at a lower rate with higher frequency.
Okay that probably does not make much sense, but if I could explain this in RF terms, it would appear to improve the third order intercept point or maybe reducing the OLG. It is not feasible to measure the OLG of this circuit since the input transistor emitter is derived from the output.
I think one should also bare in mind that harmonics are not treated as individual signals by a non-linear transfer function, but could be a result of sum and difference frequencies (due to mixing).
Adding of this resistor may not have the same results in more complex designs.
Hi Nico and Hugh, this resistor is interesting, for adding this resiter to more complex designs it has the exact same effect. I have about 15 designs which use LTP and just a quick simm showed improvement in all, most notable high frequencies. I first saw this resistor used in a motorolla amp design guide I think it was. I ll have to have closer look at OLG like Nico sugests to see how exactly this effects the amp, in the meantime Im still 
Hi Andy,
You've been helpful with LTSpice models here, and I and others are grateful. I hope you continue to contribute to DIYaudio in this way.
Don't throw it all away by demeaning yourself with mindless denigration of people you have never met, about whom you know very little.
Why don't you contribute to the future of others by going back to teaching?
Hugh
You've been helpful with LTSpice models here, and I and others are grateful. I hope you continue to contribute to DIYaudio in this way.
Don't throw it all away by demeaning yourself with mindless denigration of people you have never met, about whom you know very little.
Why don't you contribute to the future of others by going back to teaching?
Hugh
- Status
- Not open for further replies.