I would say 2N5551 a better transistor compared with BC546 for the input stage which on a good day is only suitable for 65 V The input stage rarely sees any voltage swing so can be made using lower voltage transistors than those further up the amplifier . The 80 V rating of BC546 is a maximum , the Vceo is 65 V . Realistically a 50 V device . I only mention it because 2N5551 is cheap enough ( 180 , 160 V ) .
I would fit a small resistor to each output transistor base as close to the transistor as possible . If carefully judged it can increase stability and if very carefully chosen prevent damage to the driver stage . 100 R is often used . That probably is too low to protect the driver . 1K is worth trying . In fact go up as high as you can . A simulation would be useful .
I would fit a small resistor to each output transistor base as close to the transistor as possible . If carefully judged it can increase stability and if very carefully chosen prevent damage to the driver stage . 100 R is often used . That probably is too low to protect the driver . 1K is worth trying . In fact go up as high as you can . A simulation would be useful .
Dear Nigel, do you mean a 100r to 1 K resistor in series with each output transistor base?
Thanks.
That's correct ( JMF ) . It is good practice . My brother told me years ago that transistors can have mystery failures which this resistor helps to avoid . It also forms a simple RC filter to the input which helps stability . The small inductance of the typical modern resistor is said by some to help . With FET's there is often a small oscillation inside the device ( about 5MHz if I remember ) . The resistor helps stop it entering the driver stage . Valves especailly benefit from this resistor . There are situations where such a resistor is a problem . I have seen an amplifier that was for driving a 1 R load . It had a 2N3055 transistor driving multiple 2N3055's . In this application 5 A was required to drive the outputs as the gain of a 3055 is very low when worked so hard ( 5 ) . In this case it was OK as 2N3055 will withstand such currents . 2N3773 and 2N3055 are much criticized . They are tough and practical . They can be made to work and sound excellent . The more modern devices often fail to work so easily .
Ok.
In that case you can not use such resistor values in Bipolar Power Transistors as in Vedmitraa's suggested amplifier.
We are talking base currents in the 1A range here.
Or higher.
At most, say, 1 ohm would be reasonable but any further will cause heavy power losses at the speaker output.
Those values your brother mentioned are fine for MosFets, which have no "Base Current" to deal with.
PS: and I *love* 2N3773 and 2N3055 , have been using for over 40 years and only stopped because of heavy counterfeiting
In that case you can not use such resistor values in Bipolar Power Transistors as in Vedmitraa's suggested amplifier.
We are talking base currents in the 1A range here.
Or higher.
At most, say, 1 ohm would be reasonable but any further will cause heavy power losses at the speaker output.
Those values your brother mentioned are fine for MosFets, which have no "Base Current" to deal with.
PS: and I *love* 2N3773 and 2N3055 , have been using for over 40 years and only stopped because of heavy counterfeiting
I agree . I should have looked more carefully . It is exactly as that other amp I described . Even 1 A seems a bit optimistic . MJE15030 ( if my eyes read it properly ) is a bit undersized I feel and is a bit challenged on voltage ( Vceo 120 V ? ) . The 2N3773 is on it's limits . One of the Motorola MJ range of power transistors might make an excellent driver . I think I would continue with 2N3773 and lower the voltage to +/- 57 V ( ideal for 63 V caps ) . Have to say it is refreshing to see a muscle amp .
Way back at post#16, this question went unanswered because it's becoming very difficult to remember where Greg's
website, the thread, the link or where any details for this handy resource are.
AndrewT, If you are referring recent members to the resource, please provide the link or the post+thread where it is located, as they will have zero chance
of finding it without assistance. It should be easy enough and may be worth bookmarking in your browser.
VAS Transistors
Yes.
I didn't want to go deeper before addressing the resistor value first, but a poor TO220 driving all those bases will have a hard time .... and that's an understatement
An extra 2N3773 Darlington needs to be added between the driver and 8 or 10 final power transistors.
Hi JMF , Ian . A bit ashamed that got past me . Proving I can't multitask I reckon .
2STC5200/2STA1943 look to be the same only better as a possible set of drivers . Same FT , 30 A ( about x 4 ) and 230V . They are reasonably cheap . As outputs MJ15024G at Farnell UK is reasonably priced and a great spec .
MJ15024G - ON SEMICONDUCTOR - TRANSISTOR, NPN, TO-3 | Farnell United Kingdom
Sometimes if cheap I would love to try something like BUX 80 . A switching transistor with good speed . I have a hunch with so many devices it might just work . It certainly would need a big driver .
I like this amp . It deserves to be tweaked . Any ideas for half power distortion at 10 kHz ? Also 1 W ?
2STC5200/2STA1943 look to be the same only better as a possible set of drivers . Same FT , 30 A ( about x 4 ) and 230V . They are reasonably cheap . As outputs MJ15024G at Farnell UK is reasonably priced and a great spec .
MJ15024G - ON SEMICONDUCTOR - TRANSISTOR, NPN, TO-3 | Farnell United Kingdom
Sometimes if cheap I would love to try something like BUX 80 . A switching transistor with good speed . I have a hunch with so many devices it might just work . It certainly would need a big driver .
I like this amp . It deserves to be tweaked . Any ideas for half power distortion at 10 kHz ? Also 1 W ?
Can't blame you; I trip over my own feet all day long
I tried "TV transistors" before, looked promising with all that voltage capability and reasonable current and dissipation .... blew in a wink with speaker loads.
Turns out speed and voltage are achieved sacrificing SOA.
Look at the typical SOA curve and you'll see that although they may reach 150 to 400 VCE, the permissible dissipation curve bends *down* above only 40 to 50V, turning tigers into weak kitties.
My area is much abused live sound, so I prefer slow stuff such as 2N3773 or 2N3055H (hometaxial) if available, over much faster epitaxial ones.
Not much production of such, if any, because of relatively low production yield so they end up being expensive
I tried "TV transistors" before, looked promising with all that voltage capability and reasonable current and dissipation .... blew in a wink with speaker loads.
Turns out speed and voltage are achieved sacrificing SOA.
Look at the typical SOA curve and you'll see that although they may reach 150 to 400 VCE, the permissible dissipation curve bends *down* above only 40 to 50V, turning tigers into weak kitties.
My area is much abused live sound, so I prefer slow stuff such as 2N3773 or 2N3055H (hometaxial) if available, over much faster epitaxial ones.
Not much production of such, if any, because of relatively low production yield so they end up being expensive
The BUX 40 says inductive load . As you say a TV output transformer is their idea of inductive . Thanks for your forgiveness .
I was looking for TO3 devices . They are becoming rare .
I saw in the past Totem Pole transistors to get high power from low voltage types . I understand Crown used a floating bridge circuit that worked very well . A friend has a bucket full of 3055H which he is begging me to use . I suspect like the amp above on +/- 35 V in bridge mode .
Below is the 1 ohm amp . Emitter resistors are fuses ( Says in the text . I have it in French if anyone ever wanted to read it ) . It is Gogny 1967 and is to drive their ribbon speakers . The inputs are either high Z to suit tube preamp or low Z for transistor according to the text . The input is not quite the usual . I like the shunt feedback . Forgive the image quality . Note the Darlington made of 3055's . Distortion was 0.05% and 50W 1 ohms if my memory is right ( peak at 0.8R I think ? ) . I imagine BD135/136 and BC550 would be suitable if wanting to try this ? A CCS to the VAS also ( bootstrap ) ? . Bias looks OK to get it working . It is a good example of the op-amp power amp as early as I ever found . RCA the next ( Naim and others ) . Sinclair also Z30/50 ( 1971) . Note also the very low 1st stage current ( TID was not known then ) . Note also an attempt to balance input stage current ( or optimize ) , not bad for what was a clone of nothing before it .
I was looking for TO3 devices . They are becoming rare .
I saw in the past Totem Pole transistors to get high power from low voltage types . I understand Crown used a floating bridge circuit that worked very well . A friend has a bucket full of 3055H which he is begging me to use . I suspect like the amp above on +/- 35 V in bridge mode .
Below is the 1 ohm amp . Emitter resistors are fuses ( Says in the text . I have it in French if anyone ever wanted to read it ) . It is Gogny 1967 and is to drive their ribbon speakers . The inputs are either high Z to suit tube preamp or low Z for transistor according to the text . The input is not quite the usual . I like the shunt feedback . Forgive the image quality . Note the Darlington made of 3055's . Distortion was 0.05% and 50W 1 ohms if my memory is right ( peak at 0.8R I think ? ) . I imagine BD135/136 and BC550 would be suitable if wanting to try this ? A CCS to the VAS also ( bootstrap ) ? . Bias looks OK to get it working . It is a good example of the op-amp power amp as early as I ever found . RCA the next ( Naim and others ) . Sinclair also Z30/50 ( 1971) . Note also the very low 1st stage current ( TID was not known then ) . Note also an attempt to balance input stage current ( or optimize ) , not bad for what was a clone of nothing before it .
Last edited:
Well, 2 pairs of 2N3055H at +/-35V is perfect to drive 4 ohms forever, any quirky load you can imagine.
Make as many as you want.
No need to kludge 1KW out of a complex amp with lots of them, make multiple amps if needed.
In the late 70's / early 80's I powered the Moscow Circus with a bunch of plug-in 100W power amps.
They did specifically *not* want "a big 'un".
Understandable: they run continuous shows, Monday to Monday, from 1PM to 1 AM, non stop.
No time for maintenance, except the 15 minute intervals between shows.
Make as many as you want.
No need to kludge 1KW out of a complex amp with lots of them, make multiple amps if needed.
In the late 70's / early 80's I powered the Moscow Circus with a bunch of plug-in 100W power amps.
They did specifically *not* want "a big 'un".
Understandable: they run continuous shows, Monday to Monday, from 1PM to 1 AM, non stop.
No time for maintenance, except the 15 minute intervals between shows.
remember that amplifiers in that power do not come from a schematic only thermals are important , control of the amp is important , limiting is important , protection is important
Presenting simulations , and theoretical approach of 1817 WPC is very bad practice Only this is not enough to make a working professional amplifier ...
As for the Op i will go with Andrew's sayings again ...Make a working proven design from the forum to gain experience,,then study ....and study ..and study and you may come up with something new or more innovative
Kind regards
sakis
Presenting simulations , and theoretical approach of 1817 WPC is very bad practice Only this is not enough to make a working professional amplifier ...
As for the Op i will go with Andrew's sayings again ...Make a working proven design from the forum to gain experience,,then study ....and study ..and study and you may come up with something new or more innovative
Kind regards
sakis
I would think 40 V rms in bridge mode is not out of the question for 3055's ? If 4 R that's a nice 400 W .
My friend John who builds magnetometers especially likes fairground amplifiers . As you say robust is the only parameter .
Here is the old Sinclair Z50 if anyone wants to look . It is said to have started a revolution . Simple and with substituted 3055's very robust . It was a cheap kit amp that would outperform perhaps 90 % of what was supposed to be hi fi at the time . Sinclair before that had made a class D amp . Clive Sinclair was a technical writer in the 1960's for electronics magazines . Apparently a very badly paid job . He had to write lots to make money . He changed the world , especailly if you let Acorn into the story ( ARM Chip ) .
Google Image Result for http://sweb.cz/servisdok/sinclairC18_01.jpg
My friend John who builds magnetometers especially likes fairground amplifiers . As you say robust is the only parameter .
Here is the old Sinclair Z50 if anyone wants to look . It is said to have started a revolution . Simple and with substituted 3055's very robust . It was a cheap kit amp that would outperform perhaps 90 % of what was supposed to be hi fi at the time . Sinclair before that had made a class D amp . Clive Sinclair was a technical writer in the 1960's for electronics magazines . Apparently a very badly paid job . He had to write lots to make money . He changed the world , especailly if you let Acorn into the story ( ARM Chip ) .
Google Image Result for http://sweb.cz/servisdok/sinclairC18_01.jpg
Not exactly
Most had 60V/uSec Slew Rate at full power (a more meaningful rating than simple "Frequency Response" which is usually measured ar 1W); the biggest ones (such as the one shown) "only" 45V/uSec.
The choice of the best Recording Studios in the World, including the BBC.
http://www.hhamplification.co.uk/dbfiles/techpdfs/mosfetvx.pdf
Most had 60V/uSec Slew Rate at full power (a more meaningful rating than simple "Frequency Response" which is usually measured ar 1W); the biggest ones (such as the one shown) "only" 45V/uSec.
The choice of the best Recording Studios in the World, including the BBC.
http://www.hhamplification.co.uk/dbfiles/techpdfs/mosfetvx.pdf
Barely meets the Jung criteria of 0.5V/µS per peak volt of output.
I would prefer to see a little more current available to push all that gate capacitance.
"The choice of the best Recording Studios in the World, including the BBC."
Hyperbole, no comment.
(my standards are a bit higher than theirs)
I would prefer to see a little more current available to push all that gate capacitance.
"The choice of the best Recording Studios in the World, including the BBC."
Hyperbole, no comment.
(my standards are a bit higher than theirs)
MOS FET's ( Exicon ) only need about 1 mA to work acceptably . Sure the gate capacitance at up to 1nF is a problem . My experiments say 5.6 mA is about optimum . 600 Watts at 20 kHz for music ? More like 10 W and hope the tweeters can cope , ears of the punters won't . 8mA is not my cup of tea ( too good ) . This seems to be a default value in many amps . Distortion caused by not bandwidth limiting such an amp is folly . There is no value in entertaining mice . Myself I would have the amp as hi fi with one set of outputs . Then bandwidth limited with other FET's switched in . I would have a heat-sink for normal use backed up with fans for fairground use . The punch offered by many FET's is a positive , plenty of HF maybe not ?
I was in Cuba recently . I am sure their amplifier was fine . I just wish I had brought some Motorola horns with me . It would have been nice to hear 3 kHz and above . A few other bits and I might have integrated them nicely ( stop them quacking also ) .
Now to the crunch . I bet that an over engineered power supply and what I have just described will sound better to the punters more than most of the sterile hi fi amps . The bonuses are . Robust ,simple , no emitter resistors . Can be driven by a ridiculously small VAS and can sound silky smooth . What must be remembered is many who bought these amps were changing over from valve equipment . They wouldn't tolerate the unpleasant squeaks of typical hi fi . Most people who enjoy this forum would last 5 minutes in that world . And now the crunch . At the drop of a switch this amp with no complexity becomes a nice domestic amp . The usual criticism of the Hitachi design was it is too simple to deal with HF distortion . The Hitachi graphs say totally different to that . An op amp with serious current . To me this is as serious a defect in understanding . Like saying a diesel car could never win a race , or if it did only something minor like Le Mans !
https://www.google.co.uk/search?q=g...%2Fgoldmund-mimesis-3-schmatic.html;1400;1050
I was in Cuba recently . I am sure their amplifier was fine . I just wish I had brought some Motorola horns with me . It would have been nice to hear 3 kHz and above . A few other bits and I might have integrated them nicely ( stop them quacking also ) .
Now to the crunch . I bet that an over engineered power supply and what I have just described will sound better to the punters more than most of the sterile hi fi amps . The bonuses are . Robust ,simple , no emitter resistors . Can be driven by a ridiculously small VAS and can sound silky smooth . What must be remembered is many who bought these amps were changing over from valve equipment . They wouldn't tolerate the unpleasant squeaks of typical hi fi . Most people who enjoy this forum would last 5 minutes in that world . And now the crunch . At the drop of a switch this amp with no complexity becomes a nice domestic amp . The usual criticism of the Hitachi design was it is too simple to deal with HF distortion . The Hitachi graphs say totally different to that . An op amp with serious current . To me this is as serious a defect in understanding . Like saying a diesel car could never win a race , or if it did only something minor like Le Mans !
https://www.google.co.uk/search?q=g...%2Fgoldmund-mimesis-3-schmatic.html;1400;1050
Hyperbole is something pulled out of thin air .... this is *fact* ... and I see it hurts.
Just google "HH " and "BBC" and you'll get a million hits.
All mentioning something like "commonly used by many broadcasting studios, including the BBC. " , "many of which are still in use in studios all over the world" and so on.
Well, *this* is hyperbole.
Well, that's a quite precise evaluation of the situation.
Just adding that not only they were changing from "valve equipment" but from "the best valve equipment"
Oh well.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.