hi,
I have included 2 bjt based design both complementary & quasicomplementary
for evaluation and testing by everybody .The main motive behind this thead is to make a amp with commonly available parts with good sonics and reliablity
Plz fee free to give comments and modifications required for the given amps ........ AS this is a "Amp for everybody" let us design this amp together.....
Both the amps have been simulated under Multisim 10.1 and they have reasonable good thd ratings *(0.005 for complementary & 0.08 for quasicomplementary) I had following question for the board.....
1.how power will these amps handle safetly ??
2.Can supply rail voltages be increased & by how much ??
3.How can a OPS be included in the design??
COMPLEMENTARY
QUASICOMPLEMENTARY
Regards
SEKHAR
I have included 2 bjt based design both complementary & quasicomplementary
for evaluation and testing by everybody .The main motive behind this thead is to make a amp with commonly available parts with good sonics and reliablity
Plz fee free to give comments and modifications required for the given amps ........ AS this is a "Amp for everybody" let us design this amp together.....
Both the amps have been simulated under Multisim 10.1 and they have reasonable good thd ratings *(0.005 for complementary & 0.08 for quasicomplementary) I had following question for the board.....
1.how power will these amps handle safetly ??
2.Can supply rail voltages be increased & by how much ??
3.How can a OPS be included in the design??
COMPLEMENTARY
An externally hosted image should be here but it was not working when we last tested it.
QUASICOMPLEMENTARY
An externally hosted image should be here but it was not working when we last tested it.
Regards
SEKHAR
bias needs re design ....choise of outputs is very low fi ....
to my opinion one amplifier at plan A that is so simply made will work without curent sources in the fron end almost the same good ....
to design a sophisticated front end to drive 2N3055 is not realy wise ....
in any case and given the facts as is these amps may stand up to 50+50 volts i think and may be a bit more ....
then again keeping the outputs safe can only be done with VI limiters and there will come the question what you realy like to do ..... a high end device with good sonics and good sound or a safe amplifier with full safety at all times
to my opinion one amplifier at plan A that is so simply made will work without curent sources in the fron end almost the same good ....
to design a sophisticated front end to drive 2N3055 is not realy wise ....
in any case and given the facts as is these amps may stand up to 50+50 volts i think and may be a bit more ....
then again keeping the outputs safe can only be done with VI limiters and there will come the question what you realy like to do ..... a high end device with good sonics and good sound or a safe amplifier with full safety at all times
Hi Sekhar
It should be stable.
Maybe the changes below can help?
A) Add a capacitor at the input to filter radio interference and also to make sure the input transistor has a low impedance at it's base at high frequencies.
B) Bias circuit:
1) Don't need the diodes.
2) Use a smaller pot (e.g. 1K) to adjust the output transistor bias current.
(should be about 50% for the complimentary circuit, about 25% for the quasi)
3) Add a capacitor to keep impedance low at high frequencies.
Fixing the bias should make distortion much better for the quasi.
C) I don't understand why this capacitor is here?
D) A trimpot like you show will allow adjustment for minimum distortion, but only for people with expensive test equipment. A fixed resistor is simple and nearly as good.
Regards - Godfrey
It should be stable.
Maybe the changes below can help?
A) Add a capacitor at the input to filter radio interference and also to make sure the input transistor has a low impedance at it's base at high frequencies.
B) Bias circuit:
1) Don't need the diodes.
2) Use a smaller pot (e.g. 1K) to adjust the output transistor bias current.
(should be about 50% for the complimentary circuit, about 25% for the quasi)
3) Add a capacitor to keep impedance low at high frequencies.
Fixing the bias should make distortion much better for the quasi.
C) I don't understand why this capacitor is here?
D) A trimpot like you show will allow adjustment for minimum distortion, but only for people with expensive test equipment. A fixed resistor is simple and nearly as good.
Regards - Godfrey
Attachments
The way I am reading these circuits is they look more like development amlifiers - hence pot for DC offset adjust on current mirror
I am seeing the two diodes as constant current sources for the collector for the BD 139 if mounted to heatsink for thermal impedance tracking
The 100pf at the bias chain linked between base collector of the driver attenuates the ability of the chain to operate beyond the reactive impedance "Xc" drops below 800Ω beyond 2Mhz - preventing RF instability - this is generally only done in the big power amplifiers used for power concert rigs (1000Watt to 5000Watt RMS) , because long multicores that feed signal can be 100 meters long from the main control desk - presenting a potential antenna to the amp stack - in this case input caps are not enough - clamping gain is far more effective. Not so necessary in domestic equipment where the feed is only carried on cable no more then 2 meters long and is screened properly.
With all due respect here - I think the devices used are a bit 1970's
I am seeing the two diodes as constant current sources for the collector for the BD 139 if mounted to heatsink for thermal impedance tracking
The 100pf at the bias chain linked between base collector of the driver attenuates the ability of the chain to operate beyond the reactive impedance "Xc" drops below 800Ω beyond 2Mhz - preventing RF instability - this is generally only done in the big power amplifiers used for power concert rigs (1000Watt to 5000Watt RMS) , because long multicores that feed signal can be 100 meters long from the main control desk - presenting a potential antenna to the amp stack - in this case input caps are not enough - clamping gain is far more effective. Not so necessary in domestic equipment where the feed is only carried on cable no more then 2 meters long and is screened properly.
With all due respect here - I think the devices used are a bit 1970's
neways the pcb files for the complementary amp
PCB
SIlKSCREEN
have fun
regards
sekhar
PCB
An externally hosted image should be here but it was not working when we last tested it.
SIlKSCREEN
An externally hosted image should be here but it was not working when we last tested it.
have fun
regards
sekhar
A personal preference of mine is in epitaxy planar transistors such as 2SA1216 and 2SC2922 complementaries -#
Sorry about being a bit unkind to older technology, but there is something a bit Tweed Jacket and Pipe about the ubiquitous 3055 these days --- from the era of Slippers, Roberts radios, Gramophones and Celestion Dittons
hehee
Sorry about being a bit unkind to older technology, but there is something a bit Tweed Jacket and Pipe about the ubiquitous 3055 these days --- from the era of Slippers, Roberts radios, Gramophones and Celestion Dittons
hehee
Last edited:
Member
Joined 2009
Paid Member
I enjoy the challenge of "inferior" classic components with time proven
durability and low price. Nothing wrong with 2N3055 that engineering
can't work around. A superior design should be extremely forgiving of
component spec anyway.
The 2N3055 (in the big can) is so cheap right now, its practically free.
You might spend for a chunk of aluminum with all the right holes...
I got some toxic beryllium spacers under mine, woohooo!!!...
durability and low price. Nothing wrong with 2N3055 that engineering
can't work around. A superior design should be extremely forgiving of
component spec anyway.
The 2N3055 (in the big can) is so cheap right now, its practically free.
You might spend for a chunk of aluminum with all the right holes...
I got some toxic beryllium spacers under mine, woohooo!!!...
Last edited:
Member
Joined 2009
Paid Member
I wouldn't be concerned about how old the origin of the design might be, afterall, 99% of the schematics in this Forum are based on decades old designs. The old designs are proven and many of them sound excellent. Sometimes it's hard to recognize them, people have changed tube circuits into transistors, or transistors into MOSFETs, but if you study carefully you will see a lot of familiar things.
So with so many similar circuits floating around it's getting a bit crowded for a design that might appeal to 'everyone'.
Personally, I'm quite looking forward to building an amp around my old 3055 devices sometime in the future.
I would encourage you to build something for the fun of it !
p.s. perhaps there are people local to you who have similar constraints to you in terms of access to parts or whatever and you might come up with something of interest for them. Carlos has done this kind of thing for the enthusiasts where he lives in Brazil and you might like to look at his amplifiers (they are mostly called DX something or something DX)
Last edited:
Absolutely. I agree to a point. Cheap parts are certainly viable. You just have to design for the parts. If you used enough 2N3055 and (or) 2N2955 type cheap parts with balanced current sharing and some sort of error correction to deal with the beta droop particularly at HF, you could build an amp that will rival the best. The best amp I've made, yet
, uses cheap (less than 1$/pc) hex type switching mosfets as the OPT’s. They have superb properties for driving a speaker if you can minimize their nonlinearities.
of course there is nothing wrong with 2N3055 there is deisgns switable to operate with this as well ....
but diyaudio ( always to my understanding of course ) has been spending years on these amplifiers working out every small detail every small pcb error ,every topology ,and every alternative
i see no reason to go back 10 years to work on something that is all ready been produced in million other ways with the same million of mistakes that need to be debugged
please understand i have no intention to reduce the qualities of your design but may be something more inovative had a lot more interest
but diyaudio ( always to my understanding of course ) has been spending years on these amplifiers working out every small detail every small pcb error ,every topology ,and every alternative
i see no reason to go back 10 years to work on something that is all ready been produced in million other ways with the same million of mistakes that need to be debugged
please understand i have no intention to reduce the qualities of your design but may be something more inovative had a lot more interest
encouragment
Yeah I think you should be encouraged to continue - the designs in themselves are really good considerations -- it just maybe your Amp could just benefit from a few fresh devices to try out. otherwise it could be like taking the engine out of a Morris 1000 and sticking it in an Audi R8
Yeah I think you should be encouraged to continue - the designs in themselves are really good considerations -- it just maybe your Amp could just benefit from a few fresh devices to try out. otherwise it could be like taking the engine out of a Morris 1000 and sticking it in an Audi R8
Sekhar, I think you should pursue this project. Like a few people above have said, using classic parts is a challenge, but you can pick up these devices for very good prices nowdays and create a good amp for very little outlay - so, good luck with your project!
Good practical feedback from Godfrey BTW.
For the VI limiter, I'll try to post something up over the week end. You should not build this amp without some basic protection for the o/p devices in my view.
BTW, can you post up your 2955 and 3055 models?
Good practical feedback from Godfrey BTW.
For the VI limiter, I'll try to post something up over the week end. You should not build this amp without some basic protection for the o/p devices in my view.
BTW, can you post up your 2955 and 3055 models?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.