harmonic distortion

We all start somewhere
Post Reply
User avatar
ed
retired
Posts: 4714
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

#1 harmonic distortion

Post by ed »

naive question chaps...
I've finally got round to replacing my much loved wien bridge signal generator...I hope...
I'm not sure whether to buffer the output with an opamp(which I did last time) or use a simple 2 transistor gain stage.....
Can somebody give me a rough idea whether the 2 transistor approach will be as clean as the opamp, distortion wise.

I wanted simple and originally thought the transistor approach would get away with a simpler single ended supply, whereas the opamp would need a balanced supply...But on further thinking I could probably get away with asingle ended supply on the opamp as well...

so....transistor or opamp....I barely need any gain..it's really just for the buffer.
There's nowhere you can be that isn't where you're meant to be
User avatar
Nick
Site Admin
Posts: 13657
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#2 Re: harmonic distortion

Post by Nick »

Can somebody give me a rough idea whether the 2 transistor approach will be as clean as the opamp, distortion wise
No, no it wont. as to if it will be more distorted than the generator its buffering I can't say. Stick a LME49724 on the end and it will be as good as you need.
Little known fact, coherent thought can destructively interfere with itself leaving no thought at all, that’s why I prefer incoherent thought.
User avatar
ed
retired
Posts: 4714
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

#3 Re: harmonic distortion

Post by ed »

many thanks....I see from the 2017 chart I have that it's good to 4 decimal places....trouble is, like all the low thd opamps in my list, they are all small footprint. If I go the opamp route it will have to be pdip this time.

But I was gratified that you've confirmed the discrete transistor is not the way to go.

I am very naive in this area and up to this point I had picked opamps on thd and slew rate..the introduction of all these exotic package types has taken the fun out of it for me.

I am thinking it may be possible to get the pcb board factory to place an smd on the board because I think it's a service they offer...but then I thought I'm still learning to walk with the whole pcb thing, better not complicate it yet.

I realise I might be over thinking this whole thing..the chip I'm planning to use is ad9850 which is a bit of an unknown. It may not be great shakes at thd in the first place so I shouldn't sweat about the buffer.
There's nowhere you can be that isn't where you're meant to be
User avatar
Nick
Site Admin
Posts: 13657
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#4 Re: harmonic distortion

Post by Nick »

I guess it depends on the frequency range you want to cover. I don't know the ad9850 but it looks like a 10 bit DAC in it, so distortion is only going to be at 0.02% or so, so you could just use any simple DIL opamp at that sort of level. Also depends on what you want to drive. If its a 600R line, then maybe a THAT line driver would be better, or even a power op amp. SO packages are easy to solder with a normal iron if you have board with solder mask, though a power pad like the part I suggested would need air or a oven I guess. But there are loads of other op amps with distortion down there. If you wanted simple, one of James CFP's could do fine as a discrete driver stage.
Little known fact, coherent thought can destructively interfere with itself leaving no thought at all, that’s why I prefer incoherent thought.
User avatar
ed
retired
Posts: 4714
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

#5 Re: harmonic distortion

Post by ed »

thanks....

I was over thinking this...it's only to replace the wien bridge for amplifier testing. The wien bridge only goes to about 1.5v RMS but I could adjust the opamp on the output but even then it doesn't have a square wave..I suppose I could also fiddle with the comparitor output to get a variable square wave, but it all seems such a fudge...

so the new one is the same 0-20khz but will have more drive and have a square wave, and it pushes the pcb design experience a bit further.....I'll use one of the opamps in the bit box with a single supply.....Oh, and it has a prettier tft screen.

simple is best
There's nowhere you can be that isn't where you're meant to be
JamesD
Old Hand
Posts: 894
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#6 Re: harmonic distortion

Post by JamesD »

I don't think I agree with the Op amp contention that it is better than a CFP for a simple buffer - although the headline distortion of the LME49724 is impressive it only does that at one output level and that near maximum output - as the signal level output decreases - the distortion increases - and that is the opposite of what one wants for audio be it listening or testing - and when you look at an op amps open loop gain against frequency you can see that their are problems there too... although the GBW product for the beastie is good so any problems for this is probably buried in the noise floor and that is confirmed by the spec sheet curves for THD+N against frequency below 1KHZ and then it too starts increasing... This is the main reason I don't use op amps anymore - compared to class A discrete stages they are just not good enough at the signal levels I listen at - great for high levels and spec sheet headlines - poorer in practise - at least to my ears and limited testing...

Having said that if you use it at 10V output and into loads above say 600R its pretty damn good - but is that all you need the buffer to do?
User avatar
Nick
Site Admin
Posts: 13657
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#7 Re: harmonic distortion

Post by Nick »

I don’t have that much of a dog in this race, but assuming the 0.00003% is at 10v RMS, then 0.0003% at 1v RMS maybe, 0.003% at 100mv (all entirely made up numbers, I would bet the actual numbers were less than that). Even before the ad9850 was introduced, the chances are the source will have a order of magnitude higher level of distortion than that. Also, there is the issue of PSRR that the op amp will probably do a better job at.

The original request was for a sine generator for amp testing, so I was ignoring listening preferences.
Little known fact, coherent thought can destructively interfere with itself leaving no thought at all, that’s why I prefer incoherent thought.
JamesD
Old Hand
Posts: 894
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#8 Re: harmonic distortion

Post by JamesD »

Agree that its all a bit academic I just wanted to raise that OP amps headline specs don't necessarily reflect their performance on real signals and that most people underestimate what a low gain CFP can achieve - 2 transistor CFP headphone amps on diyaudio have been measured with less than 0.001% distortion at 3Volts output into 30R load - a somewhat harder load than most op amps can do with comparable distortion at the headline and better distortion at typical usage levels...

But for what ed needs a 072 would work great and is as cheap as chips...
User avatar
Nick
Site Admin
Posts: 13657
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#9 Re: harmonic distortion

Post by Nick »

But for what ed needs a 072 would work great and is as cheap as chips...
Yep, that was my view as well.

Linear used to do a very useful power buffer, the LME49600, but now gone like a lot of the interesting stuff they used to make. Like the LME49830 I used in the first mosfet amp I brought to Owston with the Heatsinks held together with meccano.
Little known fact, coherent thought can destructively interfere with itself leaving no thought at all, that’s why I prefer incoherent thought.
JamesD
Old Hand
Posts: 894
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#10 Re: harmonic distortion

Post by JamesD »

Now that was a great little beastie - I always admired Neurochrome's Modulus - 86 nested op-amp power amplifier using the LME49720 into an LM3886 yielding 0.00011% THD at 1W and 40W flat out into 8R...

I still have several op amp based phonos, preamps and power amps around the place so I know they can measure and sound good...

Sorry Ed- thread creeping off topic....
Post Reply