Hi Bojan!
"I have enough energy to explain this, step by step, if you want to."
Thanks!
The 1st thing I don't understand is the purpose of Q5-Q6. Do they amplify anything? Second is how does the output transistor can get enough basis current?
ACD!
"Even though that the resistor is connected to collectors, they are still in series with the speaker."
Yes, if you want so, but then there is a practically infinite impedance (collector) in series with them too! And infinite+1 is about infinite! 😉 The feedback lowers both!
"I have enough energy to explain this, step by step, if you want to."
Thanks!
The 1st thing I don't understand is the purpose of Q5-Q6. Do they amplify anything? Second is how does the output transistor can get enough basis current?
ACD!
"Even though that the resistor is connected to collectors, they are still in series with the speaker."
Yes, if you want so, but then there is a practically infinite impedance (collector) in series with them too! And infinite+1 is about infinite! 😉 The feedback lowers both!
protection circuit
Regard ACD protection
ACD, your protection is bassis circuit from old applications like RCAs and was often used in some amps in the past 25 years.
This circuit is problematic because of HARD SOUNDIN CLIPPING (very dangereus at high levels) because cliping is in one moment at maksimum level 100%. This circuit do this. Without circuit like this, clipping is much more soft - something lik by tubes.
I think it isn't clever idea, new producers avoid this.
Regard ACD protection
ACD, your protection is bassis circuit from old applications like RCAs and was often used in some amps in the past 25 years.
This circuit is problematic because of HARD SOUNDIN CLIPPING (very dangereus at high levels) because cliping is in one moment at maksimum level 100%. This circuit do this. Without circuit like this, clipping is much more soft - something lik by tubes.
I think it isn't clever idea, new producers avoid this.
Just for the record
Feedback or not feedback, collector or emitter.....
Lets say you use 1 Ohm resistors, your speakers are 8 Ohm and you put out 8 Volt to your speaker.
There will when be 1 Volt across your 1 Ohm resistor.... Right ???
If so, then it's is serie with the speaker (feedback or not)
😉
Feedback or not feedback, collector or emitter.....
Lets say you use 1 Ohm resistors, your speakers are 8 Ohm and you put out 8 Volt to your speaker.
There will when be 1 Volt across your 1 Ohm resistor.... Right ???
If so, then it's is serie with the speaker (feedback or not)
😉
Bojan Hajdinjak
Yes it's quite common, however it seems that you do not understand the function of such circuit....
First...
The circuit is intended to protect the output devices, not to give you good sound when playing louder than the amp is capable of
(which you seems to be doing since you are affected by the "HARD SOUNDING CLIPPING"
Second..
This "old" circuit acts quite good due to the caps, which ensures a softer in- and outrush of the current limiter
😉
Yes it's quite common, however it seems that you do not understand the function of such circuit....
First...
The circuit is intended to protect the output devices, not to give you good sound when playing louder than the amp is capable of
(which you seems to be doing since you are affected by the "HARD SOUNDING CLIPPING"
Second..
This "old" circuit acts quite good due to the caps, which ensures a softer in- and outrush of the current limiter
😉
cliping
It's not problem of good sound at High Levels. Its problem of loudspeakers life (tweeters 😉 life). Its very dangeres to burn out tweeters.
Some people here made amps with the only of few Watts.... but those circuits are danger with more powerfull amps even more...
It's not problem of good sound at High Levels. Its problem of loudspeakers life (tweeters 😉 life). Its very dangeres to burn out tweeters.
Some people here made amps with the only of few Watts.... but those circuits are danger with more powerfull amps even more...
Pafi said:
PAFI
No, Q5 and Q6 do not amplify anything. Output transistors are without of basis current (if drives stage is in class B, or is with a little of curent (class AB) if driver stage is supplys with more current. ..
I'll show you at the first, basis circuit from Motorola (at first glance seems totaly different, but isn't,..) and shortly explain on what idea is all this circuit. So, I have to "paint" original shematic for the start and show you some basis ideas.
Bojan!
I don't mind, show it, but I don't see the point of it, if it is different.
It's OK, that no current at no signal, that's obvious. But how can you drive it? To produce 5A output, there must be 0,5 mA in basis (approx.). It drops 5V at R7, so there must be 6V on R11. This means 9mA on R11. Where does it come from? From Q5? No, You said it doesn't amplify. Via R6? Maybe. But then there must be 42V in input. Where is the mistake?
ACD!
"Lets say you use 1 Ohm resistors, your speakers are 8 Ohm and you put out 8 Volt to your speaker.
There will when be 1 Volt across your 1 Ohm resistor.... Right ???
If so, then it's is serie with the speaker (feedback or not)"
I allready said, Yes, it's in series. But with what? With speaker? It's irrelevant, if the question is damping faktor. The expression: "output impedance" (determining DF) refers to Thevenin (controlled voltage source) modell's impedance. Without NFB this is huge allready, no matter with or without the resistor. The output impedance increased from 100 ohm to 101 ohm. Who cares? With NFB it is lessen to a value wich is allmost independent of the resistor. In closed loop Thevenin modell You absolutely don't see this resistor!
I hope you don't think that a 4,5V battery has 4,5k impedance, if 1mA flowing on the load!
I don't mind, show it, but I don't see the point of it, if it is different.
It's OK, that no current at no signal, that's obvious. But how can you drive it? To produce 5A output, there must be 0,5 mA in basis (approx.). It drops 5V at R7, so there must be 6V on R11. This means 9mA on R11. Where does it come from? From Q5? No, You said it doesn't amplify. Via R6? Maybe. But then there must be 42V in input. Where is the mistake?
ACD!
"Lets say you use 1 Ohm resistors, your speakers are 8 Ohm and you put out 8 Volt to your speaker.
There will when be 1 Volt across your 1 Ohm resistor.... Right ???
If so, then it's is serie with the speaker (feedback or not)"
I allready said, Yes, it's in series. But with what? With speaker? It's irrelevant, if the question is damping faktor. The expression: "output impedance" (determining DF) refers to Thevenin (controlled voltage source) modell's impedance. Without NFB this is huge allready, no matter with or without the resistor. The output impedance increased from 100 ohm to 101 ohm. Who cares? With NFB it is lessen to a value wich is allmost independent of the resistor. In closed loop Thevenin modell You absolutely don't see this resistor!
I hope you don't think that a 4,5V battery has 4,5k impedance, if 1mA flowing on the load!
basic circuit
I'm back to the computer again
Here is the basic circuit.
Its much easier to explain that way
In one hour I'll explain all about...
Regard your question - sorry - some more minutes... on the next reply
I'm back to the computer again
Here is the basic circuit.
Its much easier to explain that way
An externally hosted image should be here but it was not working when we last tested it.
In one hour I'll explain all about...
Regard your question - sorry - some more minutes... on the next reply
basic theory
Circuit is on the first glance a little unusual. It’s must to adjust appropriate supply voltage for the IC. +- 17 V is maximum.
Regard output transistors; we need supply voltage around 30V for the appropriate power. 30V is for our OP AMP to high, so we need voltage divider like R9 and R13 near Q3 for the positive supply Voltage. Base of those transistors like Q3 are fully open (always).
On emitter of Q3 is half of supply voltage (half of 30Volts).
Maximum current cross Q3 is Vcc (30V)/R9 when the output of the Op AMP is fully opened. In our case with OP AMP on our Vcc Voltage (30V), this current is around 40 mA. When OP AMP isn’t amplified (working) his supply current is almost zero. When our OP AMP works, we have (supply) current cross the R11, which with fall of Voltage on R11, drive Darlington on the output. It’s the same for the negative half period.
Idea is to replace OP AMP with different, discrete “OP AMP” design, with similar characteristic, with symmetric supply and with similar bias. Basic circuit for this, I’ll show you in the next reply.
Circuit is on the first glance a little unusual. It’s must to adjust appropriate supply voltage for the IC. +- 17 V is maximum.
Regard output transistors; we need supply voltage around 30V for the appropriate power. 30V is for our OP AMP to high, so we need voltage divider like R9 and R13 near Q3 for the positive supply Voltage. Base of those transistors like Q3 are fully open (always).
On emitter of Q3 is half of supply voltage (half of 30Volts).
Maximum current cross Q3 is Vcc (30V)/R9 when the output of the Op AMP is fully opened. In our case with OP AMP on our Vcc Voltage (30V), this current is around 40 mA. When OP AMP isn’t amplified (working) his supply current is almost zero. When our OP AMP works, we have (supply) current cross the R11, which with fall of Voltage on R11, drive Darlington on the output. It’s the same for the negative half period.
Idea is to replace OP AMP with different, discrete “OP AMP” design, with similar characteristic, with symmetric supply and with similar bias. Basic circuit for this, I’ll show you in the next reply.
another circuit
Here is one step further
This is the same design without OP AMP.
It's the basis of our idea is to change OP AMP with some other discrete circuits and make influence on sound and operating class (to change operating class in not so simple).
Many audiophiles has something against class B, but in class B is only the output stage, all design is made with only a few stages (excelent design of) and all the circuit destroy the signal minimally. Original circuit surprisingly sound (against expectations) good, maybe better than many of, named High End.
B class output stage is super regard heatsinks. If we look on some amplifiers like Laverden (they do not need heatsinks, but sound good for the price).
It's easy to made input stage in class A, driver stage in Class A or AB and to have ouput stage in class B or AB.
Here is one step further
This is the same design without OP AMP.
An externally hosted image should be here but it was not working when we last tested it.
It's the basis of our idea is to change OP AMP with some other discrete circuits and make influence on sound and operating class (to change operating class in not so simple).
Many audiophiles has something against class B, but in class B is only the output stage, all design is made with only a few stages (excelent design of) and all the circuit destroy the signal minimally. Original circuit surprisingly sound (against expectations) good, maybe better than many of, named High End.
B class output stage is super regard heatsinks. If we look on some amplifiers like Laverden (they do not need heatsinks, but sound good for the price).
It's easy to made input stage in class A, driver stage in Class A or AB and to have ouput stage in class B or AB.
PART III
Take look on this please
Its complete amp. Driver stage with discrete "OP AMP" 🙄 😉 . Driver stage is prooven, class A design, for simetric supply, made for preamp,.... but I'm not sure, that bias is OK for this aplication. +- 15 V is OK for this circuit. I made this in a few minutes in the past, but only like idea for another research. We (HL) have diferent circuits (our kind of OP AMP made by HL) for similar circuit, but here we have some more ideas.
If this works or no, depend on bias curent of driver stage and apropriate feedback.
Take look on this please
An externally hosted image should be here but it was not working when we last tested it.
Its complete amp. Driver stage with discrete "OP AMP" 🙄 😉 . Driver stage is prooven, class A design, for simetric supply, made for preamp,.... but I'm not sure, that bias is OK for this aplication. +- 15 V is OK for this circuit. I made this in a few minutes in the past, but only like idea for another research. We (HL) have diferent circuits (our kind of OP AMP made by HL) for similar circuit, but here we have some more ideas.
If this works or no, depend on bias curent of driver stage and apropriate feedback.
Part IV
Here we have picture with some more parts. Its the same idea. Input stage is possible to made with valves, with
germanium
🙂 ,..... If wee need inpust stage depend on gain (feedback), ...
Driver stage here is in AB class, in the one reply before in class A, ...
With the apropriate driver circuit is possible to made output stage with enough stable bias curent to put it in class AB.
A class in not allways the best. Life isn't so simple. Everything depent on design, because class is only one princip in the jungle of milions on principes. Careful system matching bring us more than some ideas like FETs, Class A, Paper cones (and,.. eat only bananas 😉 ). Such design works well (experience with original circuit and with some similar designed HL products) with big, good loudspeakers, not with full range...
Here we have picture with some more parts. Its the same idea. Input stage is possible to made with valves, with
germanium 🙂 ,..... If wee need inpust stage depend on gain (feedback), ...An externally hosted image should be here but it was not working when we last tested it.
Driver stage here is in AB class, in the one reply before in class A, ...
With the apropriate driver circuit is possible to made output stage with enough stable bias curent to put it in class AB.
A class in not allways the best. Life isn't so simple. Everything depent on design, because class is only one princip in the jungle of milions on principes. Careful system matching bring us more than some ideas like FETs, Class A, Paper cones (and,.. eat only bananas 😉 ). Such design works well (experience with original circuit and with some similar designed HL products) with big, good loudspeakers, not with full range...
Re: basic theory
Some mistake - sorry. Maximum current cross Q3 is 30V/R11
Bojan Hajdinjak said:
Some mistake - sorry. Maximum current cross Q3 is 30V/R11
Bojan!
What does "fully open" means?
"When our OP AMP works, we have (supply) current cross the R11, which with fall of Voltage on R11,"
Where does this current flow? (I know, into the IC. But then? Vanishing?)
What does "fully open" means?
"When our OP AMP works, we have (supply) current cross the R11, which with fall of Voltage on R11,"
Where does this current flow? (I know, into the IC. But then? Vanishing?)
Another mistake
Here is another "mistake" on modified circuit. Feedback i positive. But this feedback is shown symbolic...
Here is another "mistake" on modified circuit. Feedback i positive. But this feedback is shown symbolic...
modules
This picture is the same like the first, modifyed version. I show this (with low resolution) only because to show modules to play with.
Regard bananas 😉 I was expected to comments on, how class B isn't enough audiophile option..... but ...
till said:
This picture is the same like the first, modifyed version. I show this (with low resolution) only because to show modules to play with.
Regard bananas 😉 I was expected to comments on, how class B isn't enough audiophile option..... but ...
Pafi said:
If you question is regard OP AMP, fully open menas when input signal is on the maximum level and OP AMP amplifies with the maximum...
If you question regard Q3, than it means, that the voltage on volktage divider R9 and R13 is such (15V + 15V), that this full open Q3.
This curent flows in original circuit cross VCC and R11, cross R7 into output Darlington, because we have no bias (almost).
If driver is in class A (depend on design), i think, we need some capacitor between the driver stage and output darlington (depend on driver circuit and lovel od DC curent)
Hi Pafi,
I will comment only on the "original" circuit only, since it seems that Bojan does not know how the circuit works. Output of the opamp should be grounded, either directly, or through low value resistor (i.e. 10-100R). Now opamp is biased by two common base gain stages Q3 and Q4. Any signal on inputs of the opamp will try to send current through opamp´s output stage therefore modulating opamp supply current. This supply current is reflected through common base gain stages Q3 and Q4 on resistors R11 and R12, modulating the output darlington pair. I will leave to Bojan to explain his part of the circuit, especially what is the maximum power dissipation on those 1R (2R?) current limiting resistors under worst case conditions, and what is the recommended load resistance to get 100W output power from 30V supply and with these current limiting resistors in series with the load.
Best regards,
Jaka Racman
I will comment only on the "original" circuit only, since it seems that Bojan does not know how the circuit works. Output of the opamp should be grounded, either directly, or through low value resistor (i.e. 10-100R). Now opamp is biased by two common base gain stages Q3 and Q4. Any signal on inputs of the opamp will try to send current through opamp´s output stage therefore modulating opamp supply current. This supply current is reflected through common base gain stages Q3 and Q4 on resistors R11 and R12, modulating the output darlington pair. I will leave to Bojan to explain his part of the circuit, especially what is the maximum power dissipation on those 1R (2R?) current limiting resistors under worst case conditions, and what is the recommended load resistance to get 100W output power from 30V supply and with these current limiting resistors in series with the load.
Best regards,
Jaka Racman