I want to upgrade the output stage in my line amp. It's a Marshal Leach Wide Bandwidth Preamp I built in the 80s that was published in Audio Magazine. It uses a push pull emitter follower totem pole (2-n-p-n) output stage that operates Class A. I was looking at D. Self's book and web site and was wondering if I could change this to a CFP (complementary feedback pair). It uses MPS8099 and MPS8599 transistors. I matched these transistors. I was also wondering if there were better transistors from Toshiba or Hitachi like the BC546 and BC556 or the 2sc1324. I would like complementary transistors that can be heat sinked back to back with an aluminum plate between to ward off thermal drift. I also feed the need to leave a space for the base resistor to compensate for self oscillation. Any other leads would be most helpful. Thanks a whole heap! Ray Hughes
60Vce allows up to +-30V supply giving a maximum output of 20Vac (+28dbu).
What are you driving?
That is some overhead margin.
Will the transistors ever see much more than half supply voltage?
625mW will allow upto 5mA bias current but they are going to run hot with 30Vce & 5mA.
Is small signal what you need? or small power?
What are you driving?
That is some overhead margin.
Will the transistors ever see much more than half supply voltage?
625mW will allow upto 5mA bias current but they are going to run hot with 30Vce & 5mA.
Is small signal what you need? or small power?
2 CFPs one above the other, totem pole
Using + and -24 volts power supply. I think it was called quasi-complementary. Would like to drive long lines to the power amp >25 feet but still operating Class A with no chance of it ever going AB. D. Self on his web site gives an approximation of the topology;http://www.dself.dsl.pipex.com/ampins/discrete/cfp.htm
down at the bottom of the page for a 1K6 load. Thanks, Ray Hughes
Using + and -24 volts power supply. I think it was called quasi-complementary. Would like to drive long lines to the power amp >25 feet but still operating Class A with no chance of it ever going AB. D. Self on his web site gives an approximation of the topology;http://www.dself.dsl.pipex.com/ampins/discrete/cfp.htm
down at the bottom of the page for a 1K6 load. Thanks, Ray Hughes
Hi,
the first Q can be bc550, it's only passing 0.2mA Pd=5mW
Don't bother with cooling.
Re the other two, how about Zetex Eline with 1W dissipation and can be fitted with a flag cooler or a glued copper foil to each face.
6mA and 24V gives Pd=140mW. Ipk into the load will be about 2k resistive and a bit higher 3 or 4k with some capacitance and stay in ClassA.
Unfortunately Zetex don't show the hFE below 10mA for the ztx650/750 pair, maybe they are embarassed to show how low gain becomes for these high current devices.
the first Q can be bc550, it's only passing 0.2mA Pd=5mW
Don't bother with cooling.
Re the other two, how about Zetex Eline with 1W dissipation and can be fitted with a flag cooler or a glued copper foil to each face.
6mA and 24V gives Pd=140mW. Ipk into the load will be about 2k resistive and a bit higher 3 or 4k with some capacitance and stay in ClassA.
Unfortunately Zetex don't show the hFE below 10mA for the ztx650/750 pair, maybe they are embarassed to show how low gain becomes for these high current devices.
Andrew, One more question, please?
In D. Self's Fig 13 on his web page, I would assume that those two diodes biasing the base of the bottom n-p-n transistor are 1n4148??? Or should I use 1n6263? Any hint as to what the capacitor from the emitter of the top p-n-p to the base of the bottom n-p-n would be-lead or lag compensation??? Start with 10 pf and work up?? Thanks again, Ray Hughes
In D. Self's Fig 13 on his web page, I would assume that those two diodes biasing the base of the bottom n-p-n transistor are 1n4148??? Or should I use 1n6263? Any hint as to what the capacitor from the emitter of the top p-n-p to the base of the bottom n-p-n would be-lead or lag compensation??? Start with 10 pf and work up?? Thanks again, Ray Hughes
grhughes said:
Dear Sir,
please excuse me if I cannot be of any help but I have a question.
It uses a push pull emitter follower totem pole (2-n-p-n) output stage that operates Class A .... I was wondering if I could change this to a CFP (complementary feedback pair)
May I ask you why ?
I am extremely interested, since a long time, about CFP (complementary feedback pair)
Do you have any experience with them ?
How do they sound ?
Do you have any tested schematic that it could be used to build a nice and simple line stage with very small voltage gain ?
Thank you very much and kind regards,
beppe
I'm not an EE but from D. Self's WEB SITE....
emitter follower output stages are, from what I understand, generally the weakest stages in terms of linearity. The CFP configuration changes an other wise single n-p-n transistor to a composite transistors by adding a complementary p-n-p in a ring configuration that greatly linearizes the emitter follower and reduces distortion. This is also known as a Sziklai pair. It's Fig 13 at the bottom of this page.
http://www.dself.dsl.pipex.com/ampins/discrete/cfp.htm
I'm changing professor of GA TECH, WM Leach's wide bandwidth preamp published in Audio Magazine (CONSTRUCT A WIDE BANDWIDTH PREAMP) totem pole output stage to the D. Self output stage. I built this in early 1980. It's toward the bottom of this page.
http://users.ece.gatech.edu/~mleach/audiothings.html
emitter follower output stages are, from what I understand, generally the weakest stages in terms of linearity. The CFP configuration changes an other wise single n-p-n transistor to a composite transistors by adding a complementary p-n-p in a ring configuration that greatly linearizes the emitter follower and reduces distortion. This is also known as a Sziklai pair. It's Fig 13 at the bottom of this page.
http://www.dself.dsl.pipex.com/ampins/discrete/cfp.htm
I'm changing professor of GA TECH, WM Leach's wide bandwidth preamp published in Audio Magazine (CONSTRUCT A WIDE BANDWIDTH PREAMP) totem pole output stage to the D. Self output stage. I built this in early 1980. It's toward the bottom of this page.
http://users.ece.gatech.edu/~mleach/audiothings.html
Re: I'm not an EE but from D. Self's WEB SITE....
Thank you very much for your extremely kind and valuble reply.
CONSTRUCT A WIDE BANDWIDTH PREAMP from Prof. WM Leach
I know very well that article.
The output stage following the volume control is actually a Sziklai pair with overal gain of about 3.
Extremely simple and interesting circuit, from my point of view.
Have you already built it? what do you think of its sound ?
to the D. Self output stage
the circuit in the D. Self's site is not completed.
Do you have a complete schematic of the output stage you want to build.
By the way I do not understand completely.
The line stage in the WIDE BANDWIDTH PREAMP of Prof. Leach is already a Sziklai pair.
Thank you very much again.
Kind regards,
beppe
grhughes said:
Thank you very much for your extremely kind and valuble reply.
CONSTRUCT A WIDE BANDWIDTH PREAMP from Prof. WM Leach
I know very well that article.
The output stage following the volume control is actually a Sziklai pair with overal gain of about 3.
Extremely simple and interesting circuit, from my point of view.
Have you already built it? what do you think of its sound ?
to the D. Self output stage
the circuit in the D. Self's site is not completed.
Do you have a complete schematic of the output stage you want to build.
By the way I do not understand completely.
The line stage in the WIDE BANDWIDTH PREAMP of Prof. Leach is already a Sziklai pair.
Thank you very much again.
Kind regards,
beppe
The output stage in the Leach WB Preamp is an emitter follower
, a totem pole emitter follower. It is not a Sziklai pair. I'm going to change it to a Sziklai pair. I've just started on this project and will bread board it and may have to try many combinations to make it stable. CFP/Sziklai pairs are sometimes unstable so it is not clear as to whether it will be successful. Just looking for guidance here.
The preamp sounds, as WML designed it, very nice, has large head room. But you have to match transistors to make it very nice. Especially the two input FETs and the two p-n-p second stage diff amp. Most of the gain is in the second stage. I bought 250 of each transistor and matched them. Cordially, Ray Hughes
, a totem pole emitter follower. It is not a Sziklai pair. I'm going to change it to a Sziklai pair. I've just started on this project and will bread board it and may have to try many combinations to make it stable. CFP/Sziklai pairs are sometimes unstable so it is not clear as to whether it will be successful. Just looking for guidance here.
The preamp sounds, as WML designed it, very nice, has large head room. But you have to match transistors to make it very nice. Especially the two input FETs and the two p-n-p second stage diff amp. Most of the gain is in the second stage. I bought 250 of each transistor and matched them. Cordially, Ray Hughes
Re: The output stage in the Leach WB Preamp is an emitter follower
I think I understand Mr Hughes.
You were referring to the phono stage in the Leach preamp while I was talking about the output stage that is already a Sziklai Pair.
Thank you very much again.
Kind regards,
beppe
grhughes said:
I think I understand Mr Hughes.
You were referring to the phono stage in the Leach preamp while I was talking about the output stage that is already a Sziklai Pair.
Thank you very much again.
Kind regards,
beppe
Thanks! I have other examples from J.Bongiorno..
and from Motchenbacher and Fitzen LOW NOISE CIRCUIT DESIGN. Thanks again for the attachment. You remarked that the D. Self circuit was incomplete. Yes, but he was only discussing an output stage. I'm just trying to improve the output stage in the WML WB discreet op amp preamp. The transistors used in your attachment are very old. It would seem that transistor manufacturing techniques would have improved in 35 years and they would not vary in beta and hfe as much as the old transistors from unit to unit.. Old transistors are all over the place when you try to match them, varying as much as 60% from their published specs. Cordially and thanks! Ray Hughes
and from Motchenbacher and Fitzen LOW NOISE CIRCUIT DESIGN. Thanks again for the attachment. You remarked that the D. Self circuit was incomplete. Yes, but he was only discussing an output stage. I'm just trying to improve the output stage in the WML WB discreet op amp preamp. The transistors used in your attachment are very old. It would seem that transistor manufacturing techniques would have improved in 35 years and they would not vary in beta and hfe as much as the old transistors from unit to unit.. Old transistors are all over the place when you try to match them, varying as much as 60% from their published specs. Cordially and thanks! Ray Hughes
Re: Thanks! I have other examples from J.Bongiorno..
I have the feeling that sziklai pair was quite common in the mid '70 beacuse a lot of preamp adopted this topology, maybe with some variations.
Please excuse me Mr Hughes
I swear is the last kind question
What would be your "top" solution for a line preamp if you have to build one ?
Would you use a WML WB discreet op amp also to build a line preamp ?
As you very well understand a nice, simple but good sounding line preamp is the missing link in my present audio chain.
Thank you very much indeed and if you like please post here your creations.
My kindest regards,
beppe
grhughes said:
I have the feeling that sziklai pair was quite common in the mid '70 beacuse a lot of preamp adopted this topology, maybe with some variations.
Please excuse me Mr Hughes
I swear is the last kind question
What would be your "top" solution for a line preamp if you have to build one ?
Would you use a WML WB discreet op amp also to build a line preamp ?
As you very well understand a nice, simple but good sounding line preamp is the missing link in my present audio chain.
Thank you very much indeed and if you like please post here your creations.
My kindest regards,
beppe
Re: Thanks! I have other examples from J.Bongiorno..
Originally posted by grhughes
...........
The transistors used in your attachment are very old.
It would seem that transistor manufacturing techniques would have improved in 35 years and they would not vary in beta and hfe as much as the old transistors from unit to unit..
Old transistors are all over the place when you try to match them, varying as much as 60% from their published specs. Cordially and thanks! Ray Hughes
I am pretty sure it should work also with newer and better pair.
That is indeed a great idea !
What transistors do you intend to use ?
I am very interested in this extremely simple, elgant but also very promising topology, as you can see from some of my old posts.
I had the opportunity to listen to the original preamp, the Radford ZD22 and was sincerely impressed by the sound.
By the way the original was way to complex in its execution.
I am strongly convinced that a nice 20K log pot followed by this gain stage would be a very good line preamp, able to drive adequately any power amp at hand.
I am interested only in line preamp because I listen only to CDs anymore.
I wait for your suggestion about the best bjts !
Thank you very much again !
kind regards,
beppe
Originally posted by grhughes
...........
The transistors used in your attachment are very old.
It would seem that transistor manufacturing techniques would have improved in 35 years and they would not vary in beta and hfe as much as the old transistors from unit to unit..
Old transistors are all over the place when you try to match them, varying as much as 60% from their published specs. Cordially and thanks! Ray Hughes
I am pretty sure it should work also with newer and better pair.
That is indeed a great idea !
What transistors do you intend to use ?
I am very interested in this extremely simple, elgant but also very promising topology, as you can see from some of my old posts.
I had the opportunity to listen to the original preamp, the Radford ZD22 and was sincerely impressed by the sound.
By the way the original was way to complex in its execution.
I am strongly convinced that a nice 20K log pot followed by this gain stage would be a very good line preamp, able to drive adequately any power amp at hand.
I am interested only in line preamp because I listen only to CDs anymore.
I wait for your suggestion about the best bjts !
Thank you very much again !
kind regards,
beppe
It's difficult to be a MINIMALIST with transistors...
but I could be mistaken.... I think we are headed in two different directions. You, a minimalist and me trying to use what I have. To improve a transistor circuit generally requires complexity; matching, constant current sources, etc.. Transistors are rather crude devices, prone to instability. Linearity, I understand, comes at the price of complexity. I'm just trying to find the latest and best transistors that can be readily ordered and partially trying to use and build upon what I now have. I am now 58 and as I age and my hearing deteriorates I am requiring more tone control. I am pursuing a discreet op amp to put tone controls in the feedback network and a rather complicated op-amp topology will allow this much better than a minimalist approach using only two transistors. Professor Leach abandoned his 2 transistor output stage in preference for the FET discreet op amp tone control circuit. But to my knowledge no improvement has been made to this circuit. I just saw the D. Self's web site and wondered if by adding a p-n-p to the existing two n-p-n output transistors would improve what I have now. I will probably just add the tone controls to the existing WML, FET discreet op amp.
Cordially, Ray Hughes
but I could be mistaken.... I think we are headed in two different directions. You, a minimalist and me trying to use what I have. To improve a transistor circuit generally requires complexity; matching, constant current sources, etc.. Transistors are rather crude devices, prone to instability. Linearity, I understand, comes at the price of complexity. I'm just trying to find the latest and best transistors that can be readily ordered and partially trying to use and build upon what I now have. I am now 58 and as I age and my hearing deteriorates I am requiring more tone control. I am pursuing a discreet op amp to put tone controls in the feedback network and a rather complicated op-amp topology will allow this much better than a minimalist approach using only two transistors. Professor Leach abandoned his 2 transistor output stage in preference for the FET discreet op amp tone control circuit. But to my knowledge no improvement has been made to this circuit. I just saw the D. Self's web site and wondered if by adding a p-n-p to the existing two n-p-n output transistors would improve what I have now. I will probably just add the tone controls to the existing WML, FET discreet op amp.
Cordially, Ray Hughes
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.