i made a test board for this schematic and it doesnt work ..... powering it up from a variac with current limmiter but keep the 15-15 volts for the op amp always on then after you supply more than 15+15 volts to the amp the current limiter goes on ....
the circuits is exactly as shown in the scematic only i used 2sa 1943 and 2sc5200 for the transistors ....
any ideas ????
the circuits is exactly as shown in the scematic only i used 2sa 1943 and 2sc5200 for the transistors ....
any ideas ????
Start your test your circuit
with having R12-R15 = 1 Ohm
And having R22 and R23 a bit smaller .. say 22/33 Ohm., I do not know exactly.
Then you control how much current you get in output.
(Measure across one those 1 Ohm resistors, for example R12)
When is alright, you can put R12-R15 back 0.22 Ohm
... and if need you put R22 and R23 a bit bigger again .. for your normal current in output.
------
When you get too much Current, the current limiter will BREAK.
Be happy it did!
We do not want transistor to BURN in smoke
with having R12-R15 = 1 Ohm
And having R22 and R23 a bit smaller .. say 22/33 Ohm., I do not know exactly.
Then you control how much current you get in output.
(Measure across one those 1 Ohm resistors, for example R12)
When is alright, you can put R12-R15 back 0.22 Ohm
... and if need you put R22 and R23 a bit bigger again .. for your normal current in output.
------
When you get too much Current, the current limiter will BREAK.
Be happy it did!
We do not want transistor to BURN in smoke
and then another thing
i disconected from the circuit R6 and R7 that goes to the base of the driver transistors and then R9 with C5 that brings the feedback ..... menaing that only the output section is connected to the rail .....
i think that by this way there shouldnt be any current draw correct or not ?????
i disconected from the circuit R6 and R7 that goes to the base of the driver transistors and then R9 with C5 that brings the feedback ..... menaing that only the output section is connected to the rail .....
i think that by this way there shouldnt be any current draw correct or not ?????
hi,
i am not good in electronics but i had exp about TL071, i also made few times some circuit using this but dsn't works, so i got fdup at the end and used others and its works...i beleve that TL071 have some problem even you perchanse new. i dn't know what problem, its not easy to use TL071. may be its not original...
this is my own xpc..
i am not good in electronics but i had exp about TL071, i also made few times some circuit using this but dsn't works, so i got fdup at the end and used others and its works...i beleve that TL071 have some problem even you perchanse new. i dn't know what problem, its not easy to use TL071. may be its not original...
this is my own xpc..
Re: just did
Okay.
I think you have an unstable circuit.
There are oscillation going on!
Is the current draw symmerical? = same current in both V+ and V-
If putting this into Simulation with Spice
you may find a way to stabilize it.
Using some compensation capacitors.
This is way too unknown and critical design, to be tested in live.
Besides, changing from the original can easy make many amplifier unstable.
Because some capacitor values were chosen to make the original stable.
1. You have any DUMMY Load 8 Ohm connected?
You should not run it unloaded.
2. Also PLEASE connect a ZOBEL filter at output.
You know those: 100nF in series with 10 Ohm
3. Also put Base Stoppers resistors in those 4 power transistors.
10 Ohm close to & going into the Base pin on each.
There are a number of reasons this amplifier can be un- stable.
As I can see from schematic.
This type of output stage is not always easy to get stable.
C10 and C11 is a try to make it more stable.
You can also try making C5 a bit bigger: 47 / 100 pF
10 pF is not much.
sakis said:
Okay.
I think you have an unstable circuit.
There are oscillation going on!
Is the current draw symmerical? = same current in both V+ and V-
If putting this into Simulation with Spice
you may find a way to stabilize it.
Using some compensation capacitors.
This is way too unknown and critical design, to be tested in live.
Besides, changing from the original can easy make many amplifier unstable.
Because some capacitor values were chosen to make the original stable.
1. You have any DUMMY Load 8 Ohm connected?
You should not run it unloaded.
2. Also PLEASE connect a ZOBEL filter at output.
You know those: 100nF in series with 10 Ohm
3. Also put Base Stoppers resistors in those 4 power transistors.
10 Ohm close to & going into the Base pin on each.
There are a number of reasons this amplifier can be un- stable.
As I can see from schematic.
This type of output stage is not always easy to get stable.
C10 and C11 is a try to make it more stable.
You can also try making C5 a bit bigger: 47 / 100 pF
10 pF is not much.
Also
this amplifier is NOT suitable to test with variac.
As I can see.
Because of R27 - R6 - R7 - R26.
That line of 4 resistors needs to be active at correct voltages
to set the Q1 & Q2 into proper work.
I would do some math to see what those 4 resistors would be
at for example +- 30 Volt supply.
When you get this amplifier to work into lower voltages supply.
then you can do some math again to see what resistor values to use for +- 80 Volt.
But I never build such high voltage power amplifiers.
I do not think those 4 power transistor
will be able to put out power much higher than 40-50 Volt.
+-80 Volt is way too high for such a simple amplifier.
this amplifier is NOT suitable to test with variac.
As I can see.
Because of R27 - R6 - R7 - R26.
That line of 4 resistors needs to be active at correct voltages
to set the Q1 & Q2 into proper work.
I would do some math to see what those 4 resistors would be
at for example +- 30 Volt supply.
When you get this amplifier to work into lower voltages supply.
then you can do some math again to see what resistor values to use for +- 80 Volt.
But I never build such high voltage power amplifiers.
I do not think those 4 power transistor
will be able to put out power much higher than 40-50 Volt.
+-80 Volt is way too high for such a simple amplifier.
Bias network?
Saki
I have in my documents a similar circuit. In the drive and in the output stage it is almost similar with your circuit, except the values of resistors etc. The main difference is an active bias network between the bases of drivers, which is absent from your circuit. In the past as i remember i had builded some amplifiers of this kind with success but only with +40/-40 Volts supply.
In another two docs i have the plans of two big power amplifiers with an IC in first stage but there are also seperate VAS stages after the IC and before the drivers. I'll try to send you the schematics because i must take photos from the plans because these are from three well known comercial P.A. amplifiers and i can't find them in the web.
Fotis
Saki
I have in my documents a similar circuit. In the drive and in the output stage it is almost similar with your circuit, except the values of resistors etc. The main difference is an active bias network between the bases of drivers, which is absent from your circuit. In the past as i remember i had builded some amplifiers of this kind with success but only with +40/-40 Volts supply.
In another two docs i have the plans of two big power amplifiers with an IC in first stage but there are also seperate VAS stages after the IC and before the drivers. I'll try to send you the schematics because i must take photos from the plans because these are from three well known comercial P.A. amplifiers and i can't find them in the web.
Fotis
Hi Sakis
To me the bias doesn't seem right. Removing the op-amp from the circuit, for DC bias 79.5/(15K + 130R)=5.25mA. 5.25mA X 130R=0.68V. 79.5/(15K + 220R)=5.22mA. 5.22mA X 220R=1.15V. 1.15V + 0.68V=1.83V. This is the base spread voltage for Q1 & 2, no? Subtracting Vbe for each one 1.83V - 1.2V=0.63V. 1K||22R= 21.5R. 0.63V/21.5R=29mA. Assuming a beta of 100 for Q1 & 2, Ib=290uA. 28.7 X 51R=1.46V across R22/23. Subtract Vbe of the output transistor and that leaves 0.864 across each 0.22 Ohm emitter resistor which is 3.93A for each output. A little high I think for class AB bias. Also 28.7mA X ~78V= 2.24W on each driver, Q1 & 2. For the CFP circuit, these drivers must be thermally compensated, not the output transistors or thermal run away will take place. This is a little bit different than I have seen. I remember something similar except there was a feedback loop setting the gain of the voltage amplifier (output stage). A voltage divider from the output to the emitters of Q1 & 2 (node of R16 & 17) and the op-amp feedback was still from the output node. You cannot set a gain of 1 with the CFP and expect a stable result, in my experience anyway.
To me the bias doesn't seem right. Removing the op-amp from the circuit, for DC bias 79.5/(15K + 130R)=5.25mA. 5.25mA X 130R=0.68V. 79.5/(15K + 220R)=5.22mA. 5.22mA X 220R=1.15V. 1.15V + 0.68V=1.83V. This is the base spread voltage for Q1 & 2, no? Subtracting Vbe for each one 1.83V - 1.2V=0.63V. 1K||22R= 21.5R. 0.63V/21.5R=29mA. Assuming a beta of 100 for Q1 & 2, Ib=290uA. 28.7 X 51R=1.46V across R22/23. Subtract Vbe of the output transistor and that leaves 0.864 across each 0.22 Ohm emitter resistor which is 3.93A for each output. A little high I think for class AB bias. Also 28.7mA X ~78V= 2.24W on each driver, Q1 & 2. For the CFP circuit, these drivers must be thermally compensated, not the output transistors or thermal run away will take place. This is a little bit different than I have seen. I remember something similar except there was a feedback loop setting the gain of the voltage amplifier (output stage). A voltage divider from the output to the emitters of Q1 & 2 (node of R16 & 17) and the op-amp feedback was still from the output node. You cannot set a gain of 1 with the CFP and expect a stable result, in my experience anyway.
Good post CBS240.
Think you have done our math here, and found out the bad bias level.
And like you, I very much doubt
that such an critical output arrangement can be used without stability troubles.
It is way trickier to use Power transistors with output from Collectors.
= Have voltage gain in final output stage.
Power followers are quite much more predictable.
They most often work very well.
Think you have done our math here, and found out the bad bias level.
And like you, I very much doubt
that such an critical output arrangement can be used without stability troubles.
It is way trickier to use Power transistors with output from Collectors.
= Have voltage gain in final output stage.
Power followers are quite much more predictable.
They most often work very well.
lineup said:
I agree absolutelly with you my good friend Lineup.
I have a career of 25 years builded mainly in P.A. audio. During this they have passed many comercial amplifiers of enormous power from my hands. I have many service manuals from these devices. My expertise builded in these mainly and not in books or articles published on the web.
Sakis, as i believe, is trying to build a simple as far as possible and powerfull P.A. amplifier. In such type devices, the designer primarilly looks at the stability and rugedness. Therefore the CFP topology it is absolutelly rejected. Also the symetrical input stage - which offers as i know very well a better slew rate from the conventional simple LTP input stage - it is rejected in the 90% of P.A. designs as i have seen in a lot of plans.
I know that you are a lover of the symetrical LTP input stage
And not unjustly. Regards
Fotios
Yes.
for PA ( Public Announcement ) amp it is good to use a conventional LTP pair input. With power power followers output.
I never build PA amplifiers. And not very often Power amplifiers.
I very often design, MultiSim & publish Headphone & Line PreAmplifiers.
Because this I can do. Well
And when using lower Voltage = 2 x15, 2x24 and lower powers we can make very fast, rapido amplifiers, with complementary design and even output from DRAIN or COLLECTORS.
And use some small capacitors to make stable.
You have got my amplifier desire, philosophy right, fotios.
I like you very much. But I do not know you much.
I know you, precise like me
.. you are a man making many thoughts other things than audio.
You are Too, one Philosopher.
Like your country man Sokrates was .. long time before.
Regards Lineup
----------
Audio is not life. Is only a LITTLE part of life.
Good Friends is life. They are BIG part of life.
for PA ( Public Announcement ) amp it is good to use a conventional LTP pair input. With power power followers output.
I never build PA amplifiers. And not very often Power amplifiers.
I very often design, MultiSim & publish Headphone & Line PreAmplifiers.
Because this I can do. Well
And when using lower Voltage = 2 x15, 2x24 and lower powers we can make very fast, rapido amplifiers, with complementary design and even output from DRAIN or COLLECTORS.
And use some small capacitors to make stable.
You have got my amplifier desire, philosophy right, fotios.
I like you very much. But I do not know you much.
I know you, precise like me
.. you are a man making many thoughts other things than audio.
You are Too, one Philosopher.
Like your country man Sokrates was .. long time before.
Regards Lineup
----------
Audio is not life. Is only a LITTLE part of life.
Good Friends is life. They are BIG part of life.
Hi
oopps, I did make a mistake, that 0.63V would be across 2 X 21.5R or 43R. This leaves 14.5mA in the drivers. 14.3mA X 51R=0.73V, so each output would have ~660mA. Still a little high and 14.3mA X 78V=~1.1W for the drivers so they will need compensated. I suspect instability, both oscillations and thermal imbalance.
oopps, I did make a mistake, that 0.63V would be across 2 X 21.5R or 43R. This leaves 14.5mA in the drivers. 14.3mA X 51R=0.73V, so each output would have ~660mA. Still a little high and 14.3mA X 78V=~1.1W for the drivers so they will need compensated. I suspect instability, both oscillations and thermal imbalance.
thanks guys .....
i read and understand all the posts and will take them under serious considaration but as usual let us not forget that this is a comercial working amplifier from hill audio that has been on production some time in the past ....
for sure i cannot tell if it was good or stable
the only changes i made was the input ic and the output transistors .... will go for 5534 this afternoon and see what changes
i read and understand all the posts and will take them under serious considaration but as usual let us not forget that this is a comercial working amplifier from hill audio that has been on production some time in the past ....
for sure i cannot tell if it was good or stable
the only changes i made was the input ic and the output transistors .... will go for 5534 this afternoon and see what changes
Yes, there is a lot worng with the schematic. First off, R6 and R7 need to be some sort of thermistor or diode stack with that can be adjusted for proper bias. If you overbias it'll burn out or run away on you.
As-is, the output will be limited to +/-13V. The outputs are configured as followers the way you have them drawn. To make this simple circuit work, you need to turn the main power supply inside out. GROUND the collectors of the output transistors and take the speaker out from the center tap of the +/-79V supply. Then reverse the input phase (swap the + and - of the op amp). Do not use a TL071, it cannot provide enough drive current - the 5534 can. And the +/- supply for the op amp should remain isolated from the main supply (don't just use regulators, use a separate supply). You can boot strap it off the main supply, but that gets a little tricky if you've never done it before.
As-is, the output will be limited to +/-13V. The outputs are configured as followers the way you have them drawn. To make this simple circuit work, you need to turn the main power supply inside out. GROUND the collectors of the output transistors and take the speaker out from the center tap of the +/-79V supply. Then reverse the input phase (swap the + and - of the op amp). Do not use a TL071, it cannot provide enough drive current - the 5534 can. And the +/- supply for the op amp should remain isolated from the main supply (don't just use regulators, use a separate supply). You can boot strap it off the main supply, but that gets a little tricky if you've never done it before.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.