Topics Topics Edit Profile Profile Help/Instructions Help Member List Member List  
Search Last 1|3|7 Days Search Search Tree View Tree View  

My amplifier does not amplify or osci...

:: EPE Chat Zone ­:: ­Radio Bygones Message Board :: » EPE Forum Archives 2010 - » Archive through 20 January, 2011 » My amplifier does not amplify or oscillate « Previous Next »

  Thread Last Poster Posts Pages Last Post
Lm358 ampdselec01/01/11  06:26 pm
  ClosedClosed: New threads not accepted on this page        

Author Message
Top of pagePrevious messageNext messageBottom of page Link to this message

istedman
Frequent Contributor
Username: istedman

Post Number: 189
Registered: 05-2005


Rating: N/A
Votes: 0 (Vote!)

Posted on Tuesday, 28 December, 2010 - 11:41 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Hello,

I'm starting to lose my sanity with an audio mixer circuit. The design shown here is for a simple, active audio mixer. The main difference is that it only uses a single +12V supply. The LM358 datasheet states that this part can be used with a single supply.

The audio input (pin 2) into the first amplifier is fine, the output (pin 1) is around 11V, only 1V less than the supply rails, ergo the amplifier has hit the rails. I can not see a reason why. Another amplifier is used to create 1/2 Vcc for the positive input, my virtual earth. A 5K1 resistor is used to compensate for any currents flowing around the negative input, though not required for the LM358.

Tried changing the LM358 for the LF353 (Bipolar to FET) but it made no difference.

In a simulation, this circuit worked perfectly!

Any suggestions or obvious mistakes you can see?

schematic
Top of pagePrevious messageNext messageBottom of page Link to this message

bruce
Frequent Contributor
Username: bruce

Post Number: 515
Registered: 04-2008

Rating: N/A
Votes: 0 (Vote!)

Posted on Wednesday, 29 December, 2010 - 12:04 am:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Ian,
If the simulation works OK, then surely looking at the circuit diagram cant help, can it?

Now, this could be b*****s, but if the i/p resistor of pin 2 ( of the first op amp ) is connected directly to the potentiometer, then effectively pin 2 is grounded and so the o/p will hit the +ve rail. Do you not require another capacitor to isolate the potentiometer (DC wise ) from the 10K i/p resistor?

Bruce
Top of pagePrevious messageNext messageBottom of page Link to this message

alec_t
Frequent Contributor
Username: alec_t

Post Number: 383
Registered: 03-2009

Rating: N/A
Votes: 0 (Vote!)

Posted on Wednesday, 29 December, 2010 - 10:01 am:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

I agree with Bruce. Pin2 is connected to 0V via R12 and the input pot R20, so the amp output goes high to try and pull pin 2 back up to match the virtual earth voltage. You could connect the pot common to the virtual earth instead of 0V.

Alec
Top of pagePrevious messageNext messageBottom of page Link to this message

bruce
Frequent Contributor
Username: bruce

Post Number: 516
Registered: 04-2008

Rating: N/A
Votes: 0 (Vote!)

Posted on Wednesday, 29 December, 2010 - 10:59 am:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Alec,
It's gratifying to hear that someone believes me, but odd that a Spice simulation let it go!

Bruce
Top of pagePrevious messageNext messageBottom of page Link to this message

alec_t
Frequent Contributor
Username: alec_t

Post Number: 385
Registered: 03-2009

Rating: N/A
Votes: 0 (Vote!)

Posted on Wednesday, 29 December, 2010 - 04:13 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Well I've just run an LTSpice sim of the circuit as shown (using an LT1077A op-amp model and assuming the pot is 10k) and alas it didn't work perfectly for me! The sim amp output had a dc operating point of nearly 11V (as you found in practice). Nevertheless it could output a few tens of mV undistorted. Try an input sig of >1V or so amplitude and the sim amp clips the waveform, as you'd expect.
I re-ran the sim with the pot common connected to a 6V virtual earth instead of 0V. That DID run perfectly, with the amp having a dc operating point of 6V!

Alec
Top of pagePrevious messageNext messageBottom of page Link to this message

violin
Frequent Contributor
Username: violin

Post Number: 137
Registered: 11-2008

Rating: N/A
Votes: 0 (Vote!)

Posted on Wednesday, 29 December, 2010 - 07:51 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

I myself am really sceptical of simulating software to say the least, as I personally consider that nothing in the virtual world testing cannot beat real-world testing and in this case I would prefer to deal with this via a signal generator and oscilloscope and not to rely on this type of software for designing. Just my personal opinion... Regards
Top of pagePrevious messageNext messageBottom of page Link to this message

istedman
Frequent Contributor
Username: istedman

Post Number: 190
Registered: 05-2005


Rating: N/A
Votes: 0 (Vote!)

Posted on Thursday, 30 December, 2010 - 04:54 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Bruce,
I tried another AC coupling capacitor between the 10K pot and the 10K input resistor, no change. :-(

Alec_t

I will try the virtual earth idea on the pot tonight.

The LM358 is an op-amp that can run from a single supply, does the data-sheet for the LT1077A support single supply operation?

@violin

I too am sceptical of simulation software, it has it's place but I do not rely on it, have been caught out a number of times where theory and practice do not meet.

@thread

After dinner I will retire to the workshop and try some of your suggestions. Thanks for the ideas.

The design uses a single supply as it is designed to run from an old computer PSU that only has +12V and +5V available.

Will post any updates later.

Ian
Top of pagePrevious messageNext messageBottom of page Link to this message

bruce
Frequent Contributor
Username: bruce

Post Number: 518
Registered: 04-2008

Rating: N/A
Votes: 0 (Vote!)

Posted on Thursday, 30 December, 2010 - 06:05 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Ian,
Alec and I agree that you cannot 'ground' the inverting input without the o/p crashing into the +ve supply. So far so good.
Now, if you stick another capacitor in front of the 10K i/p resistor, then the Op Amp is stable and wont saturate.
Now you are telling us that you've done that and that there is 'no change'. I'm sorry, that cannot be right.
What do you mean by 'no change'? If the Op Amp is still saturated, then I'd suspect you have a wiring error or somesuch. As written, that Op Amp is configured as a voltage follower and the o/p has to be Vcc/2 ( with no signal ).

Bruce
Top of pagePrevious messageNext messageBottom of page Link to this message

alec_t
Frequent Contributor
Username: alec_t

Post Number: 391
Registered: 03-2009

Rating: N/A
Votes: 0 (Vote!)

Posted on Thursday, 30 December, 2010 - 08:16 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Simply putting a cap between R12 and the pot (R20) won't cure the problem, because pin 2 is still connected to 0V via R1 and pot R19 !

Alec
Top of pagePrevious messageNext messageBottom of page Link to this message

alec_t
Frequent Contributor
Username: alec_t

Post Number: 392
Registered: 03-2009

Rating: N/A
Votes: 0 (Vote!)

Posted on Thursday, 30 December, 2010 - 09:56 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Forgot to say the LT1077A is a single-supply type like the LM358. Unfortunately LTSpice doesn't have a model for the LM358....that's why I chose the LT1077A. But the simulation should be similar.

You'll need to tie the commons of both pots R19 and R20 to the virtual ground (Vcc/2) instead of 0V.
Because pin 3 of both amps IC1A and IC2A is referenced to Vcc/2, pin 2 of both amps also needs to be referenced to Vcc/2.

HTH,
Alec
Top of pagePrevious messageNext messageBottom of page Link to this message

alec_t
Frequent Contributor
Username: alec_t

Post Number: 394
Registered: 03-2009

Rating: N/A
Votes: 0 (Vote!)

Posted on Thursday, 30 December, 2010 - 11:24 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

If you don't fancy connecting the pot commons to the Vcc/2 virtual earth (which on re-consideration might induce noise from earth-loop currents) you could use an isolating cap as Bruce suggested. In fact you'll need 4 caps; one from each of the pot wipers.

Alec
Top of pagePrevious messageNext messageBottom of page Link to this message

bruce
Frequent Contributor
Username: bruce

Post Number: 519
Registered: 04-2008

Rating: N/A
Votes: 0 (Vote!)

Posted on Friday, 31 December, 2010 - 12:44 am:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Alec,
The diagram is so faint I'm struggling to find P19, but I imagine you're right about that specific point: putting an isolting capacitor where I stated would fail; mea culpa, I didnt look at the whole cct.
However, I think this is a trifle pedantic: The key issue is that, somehow, you have to isolate the inverting input of the op amp from anything that shorts it to ground. When you do that, the output ( no signal ) will be Vcc/2. Only then will it amplify.
Well, that's my story and I'm sticking to it. I never liked audio anyway.

Bruce
Top of pagePrevious messageNext messageBottom of page Link to this message

bruce
Frequent Contributor
Username: bruce

Post Number: 521
Registered: 04-2008

Rating: N/A
Votes: 0 (Vote!)

Posted on Friday, 31 December, 2010 - 12:05 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Dear Sirs,
Furthur to my communication of the 30th inst. I would like to append the following remarks:

In the 1950s, popular beat combo 'Flanders and Swann' would, at the drop of a hat, sing a ditty entitled: 'A Song of Reproduction'. This was a typical F&S in-joke as it wasnt about sex; it was about audio. The last stanza of said ditty goes something like this:

High - fidelity,
Hi-Fi's the thing for me,
You can change the tone at a single touch,
Bel Canto sounds like double-dutch,
Still, I never did care for music much,
Just Hi-fi-del-it-y.

Enjoy

Bruce
Top of pagePrevious messageNext messageBottom of page Link to this message

alec_t
Frequent Contributor
Username: alec_t

Post Number: 395
Registered: 03-2009

Rating: N/A
Votes: 0 (Vote!)

Posted on Friday, 31 December, 2010 - 12:23 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

I think you ought to sing that for us, Bruce, and post it as a wav file

Alec
Top of pagePrevious messageNext messageBottom of page Link to this message

istedman
Frequent Contributor
Username: istedman

Post Number: 191
Registered: 05-2005


Rating: N/A
Votes: 0 (Vote!)

Posted on Friday, 31 December, 2010 - 07:51 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Hello

alect_t and bruce, you were correct. I added a capacitor between the potentiometer and R12 and disconnected R1 (feed from the other pot) and the amplifier started to work! The output level was centred around 6V, +/- 0.5V which was correct as I was feeding in a 1V signal.

The first time I tried adding the capacitor into the circuit, I used the wrong end of R12, the capacitor was connected to the - input of the op-am!

The final circuit will only have AC coupling capacitors between the centre tap of the potentiometers and the 10K input resistors. Also removed the 5K1 compensation resistors on the + input as I do not need to compensate for the input currents.

There was a non-linearity of the output, this was due to a lack of a load resistor on the output. A bias resistor, 10K, is required between the output and ground to keep the amplifier stable. Application note AN-116 from National Semiconductor details this. Tested the design upto 50 KHz before the gains started to roll off, did not get to the -3dB point but as this is an Audio mixer, the bandwidth is good enough. Will finish testing tomorrow, also looking at the low frequency bandwidth.

Thanks for your help guys. This has been a learning experience.

Ian
Top of pagePrevious messageNext messageBottom of page Link to this message

alec_t
Frequent Contributor
Username: alec_t

Post Number: 396
Registered: 03-2009

Rating: N/A
Votes: 0 (Vote!)

Posted on Friday, 31 December, 2010 - 08:32 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Quote: "The final circuit will only have AC coupling capacitors between the centre tap of the potentiometers and the 10K input resistors"

I take that to mean you will remove caps C1 C8 C9 C10. That should be ok provided your CD/phono sources have output coupling caps built in already. If they don't, you will have dc currents through the pots, which are a potential source of noise (e.g. crackling sounds if the pots are adjusted while the amp is powered up).
Oh, and I'd be inclined to add some uF in parallel with C2 to beef up the filtering of the virtual earth voltage, as any supply ripple will be fed to the non-inverting inputs of the op-amps.

Alec

(Message edited by alec_t on 31 December, 2010)
Top of pagePrevious messageNext messageBottom of page Link to this message

bruce
Frequent Contributor
Username: bruce

Post Number: 522
Registered: 04-2008

Rating: N/A
Votes: 0 (Vote!)

Posted on Saturday, 01 January, 2011 - 11:28 am:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Alec,
I was going to try and improve your NY celebrations with a rendition of 'song of reproduction', but unfortunately, as I started to sing it, the computer held out a white flag.

Bruce
Top of pagePrevious messageNext messageBottom of page Link to this message

alec_t
Frequent Contributor
Username: alec_t

Post Number: 401
Registered: 03-2009

Rating: N/A
Votes: 0 (Vote!)

Posted on Sunday, 02 January, 2011 - 02:55 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

@Bruce
Shame.

@Istedman
A further point. If the coupling caps are only 100nF then, with the resistor values as shown, the frequency response of the mixer rolls off below about 250Hz (according to LTSpice), so bass frequencies will be severely attenuated.

Alec
Top of pagePrevious messageNext messageBottom of page Link to this message

ant
Frequent Contributor
Username: ant

Post Number: 744
Registered: 05-2006

Rating: N/A
Votes: 0 (Vote!)

Posted on Sunday, 02 January, 2011 - 03:06 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Hello all,

Thankyou again Bruce, but rather than hijack this thread, I'll post "Song of Reproduction" on "Everything Else"...

Happy New Year - Ant
Top of pagePrevious messageNext messageBottom of page Link to this message

dselec
Member
Username: dselec

Post Number: 6
Registered: 11-2010

Rating: N/A
Votes: 0 (Vote!)

Posted on Sunday, 02 January, 2011 - 03:29 pm:   Edit Post Delete Post Print Post    Move Post (Moderator/Admin Only)

Hello

1)all measurements u do will be ref to virtual ground thus look at it like a reg amp with +/- supply
2)for a clean balanced sine wave pin 1/7 should show zero volt in ref 2 vcc/2 with all pots at min i constructed your lm358 amp on board and all works well
3) the amplification of the amps is r2 divided by (r12and r1 in parallel)all pots zero multiplied by the input

good luck

(Message edited by dselec on 01 January, 2011)

Administration Administration Log Out Log Out   Previous Page Previous Page Next Page Next Page