Hi,
I've built the amp you can find at:
http://home.t-online.de/home/rat.moeller/280watt.zip
This amp works only if there's an input connected, if there's nothing connected the transistors will blow up or burn out; can anyone help me with this problem???
Thanks anyway,
Hugo.
I've built the amp you can find at:
http://home.t-online.de/home/rat.moeller/280watt.zip
This amp works only if there's an input connected, if there's nothing connected the transistors will blow up or burn out; can anyone help me with this problem???
Thanks anyway,
Hugo.
I don't think it's R1, otherwise the amp would
not function with an input.
This is a very high gain amp at 83X, and I suggest
that the output is bleeding back into the input
and causing oscillation. You can correct this by
a low pass filter at the input, say at 50 - 100 KHz,
or by reducing R1 to 10 Kohm, or by attenuating the
input with a resistive divider, or by not running the
amp without an input.
Alternatively, check for proximity of input and output
hot leads, and make sure any shielded cables have the
shield properly grounded.
not function with an input.
This is a very high gain amp at 83X, and I suggest
that the output is bleeding back into the input
and causing oscillation. You can correct this by
a low pass filter at the input, say at 50 - 100 KHz,
or by reducing R1 to 10 Kohm, or by attenuating the
input with a resistive divider, or by not running the
amp without an input.
Alternatively, check for proximity of input and output
hot leads, and make sure any shielded cables have the
shield properly grounded.
I agree with Mr Pass.
Furthermore, I think that the open loop gain of the amplifier is too high, this is unreasonnable. I suggest to connect in series with each emitter of the input differential pair a small resistor (33...68 ohms) in order to keep the open loop gain at usual value. This will improve the stability of the amplifier at high frequencies.
Also, R1 must be equal to R5, i.e. 8.2 kohms, for correct ac/dc balance of the differential amplifier.
Film capacitors of 0.47 uF or more must be connected across input supply terminals (one between +48 and Gnd, the other betwwen -48 and Gnd), directly on the PCB, with shortest possible wires, in order to avoid disastrous feedback from high frequency output currents into input stages.
Regards, P.Lacombe
Furthermore, I think that the open loop gain of the amplifier is too high, this is unreasonnable. I suggest to connect in series with each emitter of the input differential pair a small resistor (33...68 ohms) in order to keep the open loop gain at usual value. This will improve the stability of the amplifier at high frequencies.
Also, R1 must be equal to R5, i.e. 8.2 kohms, for correct ac/dc balance of the differential amplifier.
Film capacitors of 0.47 uF or more must be connected across input supply terminals (one between +48 and Gnd, the other betwwen -48 and Gnd), directly on the PCB, with shortest possible wires, in order to avoid disastrous feedback from high frequency output currents into input stages.
Regards, P.Lacombe
Amplifier High Frequncy Instability
I would have to agree and say that say that the amplifier is going into oscillation when there is no audio source connected to the input and thus making both output devices conduct at the same time. This would indicate very high frequency oscillation where the speed of the output devices will not allow one device to turn off before the other starts conducing heavily.
I would suppose that the original designer had no problem with this so there is likely to be some small difference between your unit and the one the designer built. It is also very likely the circuit is somewhat unstable under the best of conditions. Excessive gain along with enough phase shift at the frequency the amplifier gain is more than 1 will turn the amplifier into a oscillator under some conditions. The frequency this occurs at can exceed 30 Mhz (30,000,000 HZ) quite easily in some designs.
Very likely is problem is aggravated when the input impedance for the audio input is raised by disconnecting the signal source. You can lower the resistance on the input but this is not a good cure. The real problem will still be there.
The best way to attack this problem may well be to simulate the circuit with computer software and make the required circuit changes to make it stable. In lieu of this the fix may very well depend on how your particular unit was built. However I would say that reducing the gain by adding the resistors in the emitters of the input differential pair is a very good place to start. The devices you used may well have more gain than those used by the designer.
I would also suggest that you fuse the positive and negative supplies to the output devices with some small value fuses, perhaps 1 amp to help prevent blowing output devices while you are trouble shooting. You may also have to add some small value feedback capacitors in the circuit somewhere to help reduce the amplifier gain at the low RF frequency range.
John Fassotte
Alaskan Audio
[Edited by alaskanaudio on 11-15-2001 at 12:32 AM]
I would have to agree and say that say that the amplifier is going into oscillation when there is no audio source connected to the input and thus making both output devices conduct at the same time. This would indicate very high frequency oscillation where the speed of the output devices will not allow one device to turn off before the other starts conducing heavily.
I would suppose that the original designer had no problem with this so there is likely to be some small difference between your unit and the one the designer built. It is also very likely the circuit is somewhat unstable under the best of conditions. Excessive gain along with enough phase shift at the frequency the amplifier gain is more than 1 will turn the amplifier into a oscillator under some conditions. The frequency this occurs at can exceed 30 Mhz (30,000,000 HZ) quite easily in some designs.
Very likely is problem is aggravated when the input impedance for the audio input is raised by disconnecting the signal source. You can lower the resistance on the input but this is not a good cure. The real problem will still be there.
The best way to attack this problem may well be to simulate the circuit with computer software and make the required circuit changes to make it stable. In lieu of this the fix may very well depend on how your particular unit was built. However I would say that reducing the gain by adding the resistors in the emitters of the input differential pair is a very good place to start. The devices you used may well have more gain than those used by the designer.
I would also suggest that you fuse the positive and negative supplies to the output devices with some small value fuses, perhaps 1 amp to help prevent blowing output devices while you are trouble shooting. You may also have to add some small value feedback capacitors in the circuit somewhere to help reduce the amplifier gain at the low RF frequency range.
John Fassotte
Alaskan Audio
[Edited by alaskanaudio on 11-15-2001 at 12:32 AM]
This might be the solution?
Maybe I found the right solution for the problem of oscillation in case of no input. Can anybody this will work properly?
you can find my solution at:
http://hugobross.tripod.com/improvement/improvement.jpg
best regards,
hugobross.
Maybe I found the right solution for the problem of oscillation in case of no input. Can anybody this will work properly?
you can find my solution at:
http://hugobross.tripod.com/improvement/improvement.jpg
best regards,
hugobross.
hugobross,
You cann't obtain good results with this op amp, which is of poor quality for audio.
Please, try the modifications gived in previous posts...you have to correct some design errors, it's the simplest way to obtain correct operation of the amplifier. Adding more circuitry will cause more problems, the simplest is the best.
Regards, P.Lacombe.
You cann't obtain good results with this op amp, which is of poor quality for audio.
Please, try the modifications gived in previous posts...you have to correct some design errors, it's the simplest way to obtain correct operation of the amplifier. Adding more circuitry will cause more problems, the simplest is the best.
Regards, P.Lacombe.
please explain this
dear P.Lacombe,
Today, at highschool, I've tested this op-amp during my laboratory sessions and I've seen that this op-amp can amplify all audio-frequencies without any distortion visible on the oscilloscope. Can you please explain me, why using this op-amp can reduce the quality? Imagine that this op-amp is perfect, do you think that the problem of unwished oscillation will be solved by reducing R1?
regards,
HB.
dear P.Lacombe,
Today, at highschool, I've tested this op-amp during my laboratory sessions and I've seen that this op-amp can amplify all audio-frequencies without any distortion visible on the oscilloscope. Can you please explain me, why using this op-amp can reduce the quality? Imagine that this op-amp is perfect, do you think that the problem of unwished oscillation will be solved by reducing R1?
regards,
HB.
Hugo,
There's more...far more to listening to music than specs, particularly a simple frequency sweep, will reveal. Opamps are convenient and (usually) cheap. There are some that are designed specifically for audio use. If you must use an opamp, choose one of these. Be prepared to pay for a quality opamp. Personally, I avoid them like the plague, although I admit that they've gotten better over the years.
There have been a few threads here about which opamp is the 'best' sounding. Use the search function to locate them.
Grey
There's more...far more to listening to music than specs, particularly a simple frequency sweep, will reveal. Opamps are convenient and (usually) cheap. There are some that are designed specifically for audio use. If you must use an opamp, choose one of these. Be prepared to pay for a quality opamp. Personally, I avoid them like the plague, although I admit that they've gotten better over the years.
There have been a few threads here about which opamp is the 'best' sounding. Use the search function to locate them.
Grey
Have you modified your circuit ?
i don't understand germany, but im interested in building the amp. but There is some think i don't understand how come it can play 280w from only 1 par of transistors?
or have i misunderstand something (Germany
I like the simple pcb and compact sice.
how does it sound . im going to use it for active monitors on stage.
jc dk.
i don't understand germany, but im interested in building the amp. but There is some think i don't understand how come it can play 280w from only 1 par of transistors?
or have i misunderstand something (Germany
I like the simple pcb and compact sice.
how does it sound . im going to use it for active monitors on stage.
jc dk.
Hugo,
It's no simple way to predict what will occurs if you use an operational amplifier, preferabily of audio type, at the input of this amplifier, depending of layout, supply decoupling, etc. Anyhow, this will add some distorsions and phase shift, and the stability of the amplifier will be poor, because this will not cure the **reals** problems, which are excessive open loop gain, and insufficient decoupling, causing severe HF oscillation.
Again, it's more simpler to correct some design errors in the design : R1 must be equal to R5, one resistor of 47 ohms or so must be insered in the emitters of each BC556B, decoupling condensers of 1 uF or more must be connected between + 48 and output GND, and - 48 and output GND.
As usual, output wires must be as far as possible from input wires. Input wires must be shielded. No connexion must exist between input and output GND except through the PCB, the chassis must be connected to GND at one point, no more, etc...
This done, you must verify that it's no residual HF oscillation at the output, with an oscilloscope, whithout, an with signal. Idle current of output transistors (without signal) must be 50 mA or so, (=50 mV between emitters of this transistors).
Regards, P.Lacombe.
It's no simple way to predict what will occurs if you use an operational amplifier, preferabily of audio type, at the input of this amplifier, depending of layout, supply decoupling, etc. Anyhow, this will add some distorsions and phase shift, and the stability of the amplifier will be poor, because this will not cure the **reals** problems, which are excessive open loop gain, and insufficient decoupling, causing severe HF oscillation.
Again, it's more simpler to correct some design errors in the design : R1 must be equal to R5, one resistor of 47 ohms or so must be insered in the emitters of each BC556B, decoupling condensers of 1 uF or more must be connected between + 48 and output GND, and - 48 and output GND.
As usual, output wires must be as far as possible from input wires. Input wires must be shielded. No connexion must exist between input and output GND except through the PCB, the chassis must be connected to GND at one point, no more, etc...
This done, you must verify that it's no residual HF oscillation at the output, with an oscilloscope, whithout, an with signal. Idle current of output transistors (without signal) must be 50 mA or so, (=50 mV between emitters of this transistors).
Regards, P.Lacombe.
Partyjups,
Calculations shows that this amplifier can deliver 250 W RMS into a 4 ohms load, and 130 W RMS into a 8 ohms load, with the specified +/- 48 V supply (stabilized).
But there is very low safety margin, as we can have seen in the fact that HF oscillation can destroy output transistors...
In my opinion, it seems more prudent to reduce supply voltage, by instance +/- 40 V, for 90 W output into 8 ohms. For 4 ohms loads, and 180 W output, it is necessary to double the output stages (and heatsinks).
Regards, P.Lacombe.
Calculations shows that this amplifier can deliver 250 W RMS into a 4 ohms load, and 130 W RMS into a 8 ohms load, with the specified +/- 48 V supply (stabilized).
But there is very low safety margin, as we can have seen in the fact that HF oscillation can destroy output transistors...
In my opinion, it seems more prudent to reduce supply voltage, by instance +/- 40 V, for 90 W output into 8 ohms. For 4 ohms loads, and 180 W output, it is necessary to double the output stages (and heatsinks).
Regards, P.Lacombe.
I've contacted the designer
This is what the designer replied to my mail I've send to him:
First im want to say sorry, for my
bad english, and for the problems,
this amplifier makes.
I know this Problems.
It seems it can happend that the
amplifier swings with a high frequenz,
and unfortunately I could not make a
change in the Plan to prevent this
absolutely.
I only can say, that with a lower
voltage of +/-30 V this Problems
can not happen.
Please tell me, where you found the
280Watt.zip, because I want to write
mails, to everywhere, where this plan
is in the internet, so that they take this
plan out of the internet.
Perhaps you can trie to change C5
to 330pF. But I don't know, if this
solve the Problems.
Bye, Robert Sontheimer
Best regards,
HB.
This is what the designer replied to my mail I've send to him:
First im want to say sorry, for my
bad english, and for the problems,
this amplifier makes.
I know this Problems.
It seems it can happend that the
amplifier swings with a high frequenz,
and unfortunately I could not make a
change in the Plan to prevent this
absolutely.
I only can say, that with a lower
voltage of +/-30 V this Problems
can not happen.
Please tell me, where you found the
280Watt.zip, because I want to write
mails, to everywhere, where this plan
is in the internet, so that they take this
plan out of the internet.
Perhaps you can trie to change C5
to 330pF. But I don't know, if this
solve the Problems.
Bye, Robert Sontheimer
Best regards,
HB.
Hello,
I'm the designer of this amplifier. My english is not so
good, but I want trie to expain some things, and awnser
some questions of this thread.
I buid this amp some years ago, and never had any problems
with it. But some other people in germany who build it, had
problems with sudden oscillation of the amp.
This problem seems to have nothing to do with the input.
Ist's not only in case of no input.
Some months ago I tried to solve the problems with some
changes in the plan 280Watt.zip, and I have seen, that the
zip-file you use, is the oldone. The new zip-file with the
changes you can download there:
http://people.freenet.de/r.sont/280Watt.zip
But the changes in this new 280Watt.zip unfortunately
are no safe solution, to prevent this oscillation
absolutely.
Now I want to say somethin to this thread:
> I suggest to connect in series with each emitter of the
> input differential pair a small resistor (33...68 ohms)
> in order to keep the open loop gain at usual value.
> This will improve the stability of the amplifier at high
> frequencies.
Yes, I think perhaps this can be a solution.
> Also, R1 must be equal to R5, i.e. 8.2 kohms, for correct
> ac/dc balance of the differential amplifier.
I think this ist wrong, because if the current in T1 is
bigger, or the current in T2 is bigger, depends on the
amplification of T3, and this can be bigger or smaller.
> i don't understand germany, but im interested in building
> the amp. but There is some think i don't understand how
> come it can play 280w from only 1 par of transistors?
> or have i misunderstand something (Germany
The used transistors TIP142 and TIP147 are very big,
but it's importent to put a fan directly on the kuehlkoerper
(sorry, I don't kwon the right word in english, but look at
the picture montage.pcx)
A second pair of transistors is not necessary, and can
not solve the oscillation-problems.
If someone now makes tests with changes at this amp
(for example emitter-resistors at T1 and T2) and
find a solution to prevent this oscillation-problems
reliable, I would like to know it. You can send me
an e-mail to r.sont@freenet.de
Best regards from germany,
Robert Sontheimer
I'm the designer of this amplifier. My english is not so
good, but I want trie to expain some things, and awnser
some questions of this thread.
I buid this amp some years ago, and never had any problems
with it. But some other people in germany who build it, had
problems with sudden oscillation of the amp.
This problem seems to have nothing to do with the input.
Ist's not only in case of no input.
Some months ago I tried to solve the problems with some
changes in the plan 280Watt.zip, and I have seen, that the
zip-file you use, is the oldone. The new zip-file with the
changes you can download there:
http://people.freenet.de/r.sont/280Watt.zip
But the changes in this new 280Watt.zip unfortunately
are no safe solution, to prevent this oscillation
absolutely.
Now I want to say somethin to this thread:
> I suggest to connect in series with each emitter of the
> input differential pair a small resistor (33...68 ohms)
> in order to keep the open loop gain at usual value.
> This will improve the stability of the amplifier at high
> frequencies.
Yes, I think perhaps this can be a solution.
> Also, R1 must be equal to R5, i.e. 8.2 kohms, for correct
> ac/dc balance of the differential amplifier.
I think this ist wrong, because if the current in T1 is
bigger, or the current in T2 is bigger, depends on the
amplification of T3, and this can be bigger or smaller.
> i don't understand germany, but im interested in building
> the amp. but There is some think i don't understand how
> come it can play 280w from only 1 par of transistors?
> or have i misunderstand something (Germany
The used transistors TIP142 and TIP147 are very big,
but it's importent to put a fan directly on the kuehlkoerper
(sorry, I don't kwon the right word in english, but look at
the picture montage.pcx)
A second pair of transistors is not necessary, and can
not solve the oscillation-problems.
If someone now makes tests with changes at this amp
(for example emitter-resistors at T1 and T2) and
find a solution to prevent this oscillation-problems
reliable, I would like to know it. You can send me
an e-mail to r.sont@freenet.de
Best regards from germany,
Robert Sontheimer
Robert,
You can see here :
http://users.ece.gatech.edu/~mleach/lowtim/instage.html
explanations about utility of emitter resistors in the input differential pairs. I use such resistors in all my desings since 10 years, and I have no more problems with HF oscillations. Furthermore, this will dramatically decrease transitory intermodulation distorsion.
Concerning doubling of output transistors, if you use 4 ohms driver as load, effective loading can be as low as 3 ohms at 400 Hz or so, causing intensity reaching 15 A peak. TIPs are specified 10 A continuous, 20 A peak, so safety margin is too low in my opinion. Because of this, I have suggested to double the output transistors in the case of 4 ohms load. This will most certainly have no effect on HF stability, I agree with this.
Regards, P.Lacombe.
You can see here :
http://users.ece.gatech.edu/~mleach/lowtim/instage.html
explanations about utility of emitter resistors in the input differential pairs. I use such resistors in all my desings since 10 years, and I have no more problems with HF oscillations. Furthermore, this will dramatically decrease transitory intermodulation distorsion.
Concerning doubling of output transistors, if you use 4 ohms driver as load, effective loading can be as low as 3 ohms at 400 Hz or so, causing intensity reaching 15 A peak. TIPs are specified 10 A continuous, 20 A peak, so safety margin is too low in my opinion. Because of this, I have suggested to double the output transistors in the case of 4 ohms load. This will most certainly have no effect on HF stability, I agree with this.
Regards, P.Lacombe.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.