Author Topic: TL071 distortion - bad amplifier design?  (Read 6967 times)

0 Members and 2 Guests are viewing this topic.

Offline AngraMelo

  • Regular Contributor
  • *
  • Posts: 217
  • Country: br
TL071 distortion - bad amplifier design?
« on: December 09, 2018, 11:10:12 pm »
Hey guys,
I just got a couple of boards and populated them for an audio amplifier.
After completion and checking to see if everything was in the right place, I powered it up, injected a 1kHz sine wave from the generator, hooked up the speaker to the output and was greeted with a enormous distortion.

Schematic attached

probing the output of the TL071 (pin 6) shows that the distortion is coming from there.
Now, the other thing is that by powering the amp with +-40V (unregulated but stable) I can never get a output of more than 7.5Vpp
That is very weird. I thought the whole point of the +-40V rails was to have the power transistors swinging accross that voltage.

I have no clue what is going on. Any help?
 

Offline HB9EVI

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: ch
Re: TL071 distortion - bad amplifier design?
« Reply #1 on: December 09, 2018, 11:22:38 pm »
First of all, you have the opamp inputs wrong; the feedback network goes on the inverting input and the audio signal on the non-inverting.
That what looks like a VAS stage, is not really needed if you use an opamp - but you have problem; the opamp cannot deliver full swing of nearly +/-40V, since its supply rails are limited by the zeners, so the whole concept has big lacks.

My tip would be to construct a proper long tailed pair input stage, a VAS and then the emitter follower as output stage; but also in the output stage as-is are several things to correct, for example you have far to many miller capacitors, which limit you the bandwidth far over the needs.
« Last Edit: December 09, 2018, 11:26:52 pm by HB9EVI »
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9206
  • Country: my
  • reassessing directives...
Re: TL071 distortion - bad amplifier design?
« Reply #2 on: December 10, 2018, 03:09:36 am »
First of all, you have the opamp inputs wrong; the feedback network goes on the inverting input and the audio signal on the non-inverting.
its common to feed the signal in inverting manner, but the feedback path, it depends.

That what looks like a VAS stage, is not really needed if you use an opamp - but you have problem; the opamp cannot deliver full swing of nearly +/-40V, since its supply rails are limited by the zeners, so the whole concept has big lacks.
the opamp doesnt have to swing full ±40V, there's app note for voltage gain stage like this...

it seems the OP copied the design from somewhere else, he may want to post here the original diagram so somebody may check for him.

Now, the other thing is that by powering the amp with +-40V (unregulated but stable) I can never get a output of more than 7.5Vpp
That is very weird. I thought the whole point of the +-40V rails was to have the power transistors swinging accross that voltage.
the TL071 will never be supplied from ±40V it will smoke otherwise. the zeners will limit the incoming supply to maybe ±15V. full swing is achieved in the 1st or 2nd stage of the transistor amplifer, you may check the swing at TIP41/42 emitter pins or after the 100R resistors.
« Last Edit: December 10, 2018, 03:16:28 am by Mechatrommer »
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline oPossum

  • Frequent Contributor
  • **
  • Posts: 845
  • Country: us
  • The other white meat
Re: TL071 distortion - bad amplifier design?
« Reply #3 on: December 10, 2018, 03:15:26 am »
The stage following the op-amp is inverting, so the feedback to the + input of the op-amp is correct. That stage also has voltage gain, so the output can get near the rails.

I don't see anything obvious wrong with the circuit. It may be a build error. Pictures of the build and the distortion you are seeing would be helpful.
 
The following users thanked this post: newbrain

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9206
  • Country: my
  • reassessing directives...
Re: TL071 distortion - bad amplifier design?
« Reply #4 on: December 10, 2018, 03:22:10 am »
The stage following the op-amp is inverting, so the feedback to the + input of the op-amp is correct. That stage also has voltage gain, so the output can get near the rails.
the feedback should stabilize/equalize the inverting input path. if inp goes -ve, out will goes +ve, if further voltage gain stages goes +ve, it should go to inverting pin, otherwise it can go to non inverting pin. but you might be correct. if probing the output is flat low or high, the feedback path is probably wrong, but if excessive swing, maybe too much voltage gain there or even TL071 goes phase reversal. ymmv..

edit: err.. your speaker output should be connected to the last emitter follower (current gain) stage 52ee/1943 after 0.22R. not the earlier stage tip41/42. there dont have enough current power, you load that node you'll have distortion. ymmv.
« Last Edit: December 10, 2018, 03:29:41 am by Mechatrommer »
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline Bassman59

  • Super Contributor
  • ***
  • Posts: 1291
  • Country: us
  • Yes, I do this for a living
Re: TL071 distortion - bad amplifier design?
« Reply #5 on: December 10, 2018, 03:25:27 am »
The stage following the op-amp is inverting, so the feedback to the + input of the op-amp is correct.

no.
 

Offline AngraMelo

  • Regular Contributor
  • *
  • Posts: 217
  • Country: br
Re: TL071 distortion - bad amplifier design?
« Reply #6 on: December 10, 2018, 03:39:12 am »
I actually bought the boards. I got them on our local "ebay" kind of website.
Both boards were assembled exactly as shown on the schematic.

it did not matter how many volts I put in, the output would never swing more than 7.5Vpp. After that it would distort until it became a nearly square wave.

By probing the output of the TL071, the wave became a whole new thing completely, it was distortion in the theory sense of the word but what I mean is that it did not resemble a sine wave anymore.

How could the 1015-1815 and/or Tip41/42 be a VAS? They are both emitter followers. I dont understand.
 

Offline AngraMelo

  • Regular Contributor
  • *
  • Posts: 217
  • Country: br
Re: TL071 distortion - bad amplifier design?
« Reply #7 on: December 10, 2018, 03:54:39 am »
That is exactly what happened. I used a 4ohm speaker, the output went nuts and all anyone could hear was distortion galore.
I think that amp is based on this one here
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9206
  • Country: my
  • reassessing directives...
Re: TL071 distortion - bad amplifier design?
« Reply #8 on: December 10, 2018, 04:07:47 am »
sorry i thought 100R and 0.22R is disconnected, you forgot big dot in OP :D the original schematics using ±70V, you are saying you are using ±40V the circuit maybe tuned to gain from ±70V, ±40V maybe not enough to stabilize feedback. you need to decrease gain (need some expertise on that) or reduce input level. and you dont have 100R R14, R16 and R15 at 2 biasing diodes there that may screw things up significantly. ymmv.
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline oPossum

  • Frequent Contributor
  • **
  • Posts: 845
  • Country: us
  • The other white meat
Re: TL071 distortion - bad amplifier design?
« Reply #9 on: December 10, 2018, 04:48:40 am »
How could the 1015-1815 and/or Tip41/42 be a VAS? They are both emitter followers. I dont understand.

A1015 and C1815 are common emitter configuration, so they have voltage gain and invert. The two following stages are common collector (also known as emitter follower), so they have only current gain and are non-inverting.
 
The following users thanked this post: Mechatrommer

Offline oPossum

  • Frequent Contributor
  • **
  • Posts: 845
  • Country: us
  • The other white meat
Re: TL071 distortion - bad amplifier design?
« Reply #10 on: December 10, 2018, 04:55:14 am »
The voltage rating on the A1015 and C1815 is too low. You should have at least 100V Vceo devices for +/- 40 volt rails. The TIP41/42 must be the C versions (TIP41C, TIP42C).
 

Offline Treehouseman

  • Supporter
  • ****
  • Posts: 53
Re: TL071 distortion - bad amplifier design?
« Reply #11 on: December 10, 2018, 05:00:46 am »
The stage following the op-amp is inverting, so the feedback to the + input of the op-amp is correct.

no.
k
 

Offline oPossum

  • Frequent Contributor
  • **
  • Posts: 845
  • Country: us
  • The other white meat
Re: TL071 distortion - bad amplifier design?
« Reply #12 on: December 10, 2018, 05:04:00 am »
You should also have the Zobel network (10 ohm resistor + 100 nF capacitor) on the output to help with stability.
 
The following users thanked this post: Mechatrommer

Offline AngraMelo

  • Regular Contributor
  • *
  • Posts: 217
  • Country: br
Re: TL071 distortion - bad amplifier design?
« Reply #13 on: December 10, 2018, 05:08:08 am »
Like this?
 

Offline oPossum

  • Frequent Contributor
  • **
  • Posts: 845
  • Country: us
  • The other white meat
Re: TL071 distortion - bad amplifier design?
« Reply #14 on: December 10, 2018, 05:17:52 am »
 

Offline Circlotron

  • Super Contributor
  • ***
  • Posts: 1648
  • Country: au
Re: TL071 distortion - bad amplifier design?
« Reply #15 on: December 10, 2018, 05:52:55 am »
By probing the output of the TL071, the wave became a whole new thing completely, it was distortion in the theory sense of the word but what I mean is that it did not resemble a sine wave anymore.
The TL071 output is inside the feedback loop so it may well look distorted. If the circuit is operating properly, the TL071 output will be whatever it needs to be to make the amplifier output "zero" distortion.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 10039
  • Country: us
  • DavidH
Re: TL071 distortion - bad amplifier design?
« Reply #16 on: December 10, 2018, 08:55:29 am »
First of all, you have the opamp inputs wrong; the feedback network goes on the inverting input and the audio signal on the non-inverting.

The first transistor stage inverts the output so feedback does need to go from the output to the non-inverting input.

Quote
That what looks like a VAS stage, is not really needed if you use an opamp - but you have problem; the opamp cannot deliver full swing of nearly +/-40V, since its supply rails are limited by the zeners, so the whole concept has big lacks.

The first stage has voltage gain so the operational amplifier can have a lower output voltage range.



There are better but more obscure ways to design an amplifier like this but it should still work and I have done it this way in the past.  The TL071 is faster than a 741 so it might require a separate feedback network from its output (not the amplifier's output) to inverting input to maintain stability.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 13179
  • Country: gb
  • 0999
Re: TL071 distortion - bad amplifier design?
« Reply #17 on: December 10, 2018, 09:10:03 am »
The stage following the op-amp is inverting, so the feedback to the + input of the op-amp is correct.

no.
Take another look. The second stage, A1015 and C1815, is in a common emitter configuration, which is inverting and has the voltage gain required to boost the output voltage from the op-amp. The third and final stages TIP41 and TIP42 and 2SC52EE and 2SA1943 are emitter followers, which just amplify the current.

I actually bought the boards. I got them on our local "ebay" kind of website.
Both boards were assembled exactly as shown on the schematic.

it did not matter how many volts I put in, the output would never swing more than 7.5Vpp. After that it would distort until it became a nearly square wave.

By probing the output of the TL071, the wave became a whole new thing completely, it was distortion in the theory sense of the word but what I mean is that it did not resemble a sine wave anymore.

How could the 1015-1815 and/or Tip41/42 be a VAS? They are both emitter followers. I dont understand.
When the input is connected to 0V and the output open circuit, please measure the following voltages, all relative to 0V:

The +/-40V supply
Output of the circuit.
TL071: pin 2, 3, 4, 6 & 7.
« Last Edit: December 10, 2018, 10:31:51 am by Hero999 »
 

Offline HB9EVI

  • Frequent Contributor
  • **
  • Posts: 391
  • Country: ch
Re: TL071 distortion - bad amplifier design?
« Reply #18 on: December 10, 2018, 09:37:47 am »
Thanks for the feedback; you never stop learning. I've not yet met such an amp, so it looked all wrong to me;
not even in Self on Audio I found such an amp.

good to know!
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 2178
  • Country: gb
Re: TL071 distortion - bad amplifier design?
« Reply #19 on: December 10, 2018, 09:50:01 am »
The stage following the op-amp is inverting, so the feedback to the + input of the op-amp is correct.

no.

Yes.  The drive stage after the op-amp is inverting, so the feedback connection to the non-inverting op-amp input is correct.

However "distortion" doesn't tell us much without an example of the output waveform.  e.g. could be the amp is oscillating, or incorrectly biased and clipping.
 

Offline t1d

  • Frequent Contributor
  • **
  • Posts: 626
  • Country: us
Re: TL071 distortion - bad amplifier design?
« Reply #20 on: December 10, 2018, 09:59:08 am »
I actually bought the boards. I got them on our local "ebay" kind of website.
Well, that may be a good portion of the problem. Even with the extremely knowledgeable and gracious experts here on the forum sorting out the design, the components are probably not up to specs and there may be no chasing down the problem, due to that fact. Don't get me wrong, I buy Ebay stuff, when appropriate. So, for learning the circuit, maybe what you have is a good tool. But, for your discerning ear, maybe not. Again, I am not faulting you, by any means.
 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: TL071 distortion - bad amplifier design?
« Reply #21 on: December 10, 2018, 10:39:46 am »
Hi AngraMelo

That is a pretty standard audio power amplifier architecture and has the potential for high quality reproduction, but there are two areas that you need to know about.

[1] The drive from the TL071 to the voltage amplification stage is troublesome.

[2] The amplifier will generate severe cross-over distortion.

To fix [1] above change the two 4.7nF capacitors for 47uF. The amplifier should then work.

I am quite busy at the moment but, maybe in about ten hours time. I will post a revised circuit to fix both issues properly.

« Last Edit: December 10, 2018, 10:41:55 am by spec »
 
The following users thanked this post: AngraMelo

Offline perieanuo

  • Frequent Contributor
  • **
  • Posts: 293
  • Country: fr
Re: TL071 distortion - bad amplifier design?
« Reply #22 on: December 10, 2018, 10:55:06 am »
And for god's sake put limiting resistors on base on final resistors, once they saturate you're in trouble.
On input I will use in your place bigger capacitors like 10 uF


Envoyé de mon iPad en utilisant Tapatalk
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 13179
  • Country: gb
  • 0999
Re: TL071 distortion - bad amplifier design?
« Reply #23 on: December 10, 2018, 11:51:24 am »
And for god's sake put limiting resistors on base on final resistors, once they saturate you're in trouble.
On input I will use in your place bigger capacitors like 10 uF


Envoyé de mon iPad en utilisant Tapatalk
You mean the final transistors, not resistors, i.e. the 2SC52EE and 2SA1943?

If so, how can they saturate? They're emitter followers.
 

Offline perieanuo

  • Frequent Contributor
  • **
  • Posts: 293
  • Country: fr
Re: TL071 distortion - bad amplifier design?
« Reply #24 on: December 10, 2018, 11:56:01 am »
And for god's sake put limiting resistors on base on final resistors, once they saturate you're in trouble.
On input I will use in your place bigger capacitors like 10 uF


Envoyé de mon iPad en utilisant Tapatalk
You mean the final transistors, not resistors, i.e. the 2SC52EE and 2SA1943?

If so, how can they saturate? They're emitter followers.
They can.for example, from thermal considerations or if driver goes short.that's why between emiter driver and final transistor base it's good idea to put resistor.see the second schematic that does have.

Look here and try to understand in practice it can arrive this situation"The bias resistor must connect to a voltage that is at least 0.7V higher than the supply voltage for the upper emitter-follower transistor to saturate." (eg if you have voltage loss at high power on your wires inside the amplificator, yes, the wires between the final transistor and the +Vdd, I seen this in a clone philips schematic, in practice) :
https://www.electro-tech-online.com/threads/current-limiter-using-transistor-how-to-saturate-an-emitter-follower-bjt.118323/

What you are saying should never arrive, but in practice can arrive, that's the trick.
I agree final transistor will die in short-circuit load scenario and 0.22 resistor should prevent thermal derive.
For me this schematic is not the most stable choice possible, but maybe you're right ....
Anyway there are better ways to prevent signal to be overamplified, thermal and oscillation tendances.
It's too complicated to talk about all the aspects with so simple amplifier.
Pierre


Envoyé de mon iPad en utilisant Tapatalk
« Last Edit: December 10, 2018, 01:05:15 pm by perieanuo »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf