Re: Low distortion sinewave oscillator without big capacitor.

On 9/04/2025 1:56 am, Edward Rawde wrote:

"Bill Sloman" <bill.sloman@ieee.org> wrote in message news:vt2bck$1gmbm$1@dont-email.me...

On 7/04/2025 11:10 pm, Edward Rawde wrote:

"Bill Sloman" <bill.sloman@ieee.org> wrote in message news:vsvspr$33gdd$1@dont-email.me...

On 7/04/2025 3:25 am, Edward Rawde wrote:

"Bill Sloman" <bill.sloman@ieee.org> wrote in message news:vsuacb$1d4ec$1@dont-email.me...

On 7/04/2025 12:39 am, Edward Rawde wrote:

"Bill Sloman" <bill.sloman@ieee.org> wrote in message news:vsu092$14oc7$1@dont-email.me...

On 6/04/2025 2:12 am, JM wrote:

On Fri, 4 Apr 2025 23:55:11 -0400, "Edward Rawde"
<invalid@invalid.invalid> wrote:
>

"JM" <sunaecoNoChoppedPork@gmail.com> wrote in message news:uop0vjp3d13t441ujfboi5aeeg08anm1je@4ax.com...

On Fri, 4 Apr 2025 16:29:27 -0400, "Edward Rawde"
<invalid@invalid.invalid> wrote:
>

"JM" <sunaecoNoChoppedPork@gmail.com> wrote in message news:add0vjdh2gcma0n9pfunq76n04cfbkhtnj@4ax.com...

On Fri, 4 Apr 2025 14:25:29 -0400, "Edward Rawde"
<invalid@invalid.invalid> wrote:
>

"Bill Sloman" <bill.sloman@ieee.org> wrote in message news:vsnufh$2ou7j$1@dont-email.me...

On 4/04/2025 11:33 am, JM wrote:

On Thu, 3 Apr 2025 19:25:33 -0400, "Edward Rawde"
<invalid@invalid.invalid> wrote:
>

"JM" <sunaecoNoChoppedPork@gmail.com> wrote in message news:qq8tujlpciqc2jrd0ibljmjr9pd37ip6hi@4ax.com...

On Sun, 30 Mar 2025 14:54:56 -0400, "Edward Rawde"
<invalid@invalid.invalid> wrote:
>

Not long ago JM posted a 1KHz sinewave oscillator with very low distortion.
It used a 470uF non polarized capacitor which in practice would probably be made from two 1000uF capacitors.
There's nothing wrong with that but I wanted to see whether I could make a working circuit without needing such a
large
capacitor.

>

You will need to adjust the feedback to suit.  Start with a -5 or -6
gain block after the integrator and adjust it's gain until the startup
is clean (no saturation).

>
Here's my version of John May's variation.
>

...
>
Yes, that works but only 120dB down at 4KHz.
I put the damping resistor back to 47k since I don't care what happens during the first few seonds as long as it
happens.
If it's necessary to wait one minute for the purest tone, that's fine with me.
>

>
Linewraps are going to be a problem - delete all "\n" from the last few lines

>
You don't want to remove \n just remove the wraps.
>
BCM61B does not have two independent transistors.
>
So you probably want BCM847BS which has two independent matched transistors and a very low price at digikey, so may as
well
use
two
of them as shown below.
>
I took the model from
https://github.com/peteut/spice-models/blob/master/nxp/complex_discretes/complex_discretes.txt
>
It should only be necessary to unwrap the last line of the following.
Don't remove \n just remove the wraps, you may need to use a horizontal scroll bar.
>
Version 4.1

>
>
Best to just offset the integrator output so the amplitude is brought
under control sooner.
>

>
Wow. That has much lower distortion too.
>

>
If you just replace Q1,2 duals with a simple long tailed pair I think
you will get better performance.

>
I'm not sure I understand how the multiplier could be implemented with just a long tailed pair.
>

>
Just modulate the tail current and select how much to steer to the
output by directly driving the bases rather than indirectly as in your
circuit.  The following link shows one example topology, and a four
quadrant differential I/O version.  Compare the linearity of each of
them.
>
https://1drv.ms/u/c/1af24d72a509cd48/EWVCUG7-jFJMu7-01VczCRcBzEC9JPHrV45x7TOunN90Gg?e=GXbvX5
>
It could be used as shown here.
>
https://1drv.ms/u/c/1af24d72a509cd48/EVmMVrvUD15GutoR5nCJ7QEBSeZsHWpHudqR0b8XtTLMLw?e=HIV74I

>
As I've already said, I like it. I've played with it a bit.
>
The ON-Semiconductor NSS40301MDR2G NPN dual comes with a 2mV guaranteed maximum difference in base-emitter voltages (at the
same
emitter current). Edward Rawde's Nexperia BCM61B dual part has matched current gain, but no guarantee on the Vbe. The
Nexperia
BCM847BS does offer 2mV base-emitter matching, and would presumably work just as well.
>
Putting in the ON-Semiconductor dual means that your gain control circuit doesn't have to waste output swing coping with
part-to-part variation.
>
I've added a cascode transistor (Q1, it should be Q4) to minimise any Early effect distortion.
>
I've snipped out the op amp driving the base of Q3. Once you'd gone AC-coupled, it wasn't doing anything useful.
>
And I've put a string of eight diodes in series with R10. They nominally compensate for the temperature dependence
introduced
by
the four rectifier diodes D2, D12, D13 and D14. In this version of the circuit the rectifiers knock about 0.6 volts off a
sine
wave that peaks at 3.8V, about a quarter of the 15V rail. I haven't run the numbers to fix the best number of diodes, but
something between six and eight looks okay.
>
The harmonics aren't great - most of them are about 90dB below the fundamental, but the seventh is only 85dB down.
>

>
Not sure I see the point if it's only 80dB down Bill.
@ 7Khz in LTSPice 24.1.5

>
LTSpice isn't all that reliable as predictor of low level distortion. Having an armoury of different circuits to try when you
finally get around to building  and testing something real may be useful.
>

I can almost get that from a simple phase shift oscillator and a 1KHz tuned circuit.

>
I'm sure that you think so. John Larkin thought that a bang-bang amplitude control was worth suggesting...
>

The last time I included a diode string like that in one of my circuits (which I seem to recall had better than 80dB
performance)
you told me it was nuts.

>
It probably was. In this case there are better ways of getting a rectified output than a simple series diode - I've posted
circuits which incorporate precision rectifiers which get rid of the forward drop through the diode, and I've built circuits
that
used synchronous rectifiers built around transmission gates where the output isn't shifted by a temperature dependent diode
drop.
It went into a GaAs single crystal puller as a retrofit.
>
The main point of the diode string was as a satirical comment on that aspect of the design you posted, which probably counts
as
being hostile, but I am hostile to ill-thought out designs, hard though it is to get the design time to sort them out.

>
I don't see anything particularly hostile there Bill. Just different points of view.
>

I've got stuck with sorting out other peoples half-baked designs often enough, but only after my bosses had had their noses
rubbed
in the unfortunate consequences.

>
When I started work as a fresh graduate (but one who also had practical experience with everything from TV antenna systems in
hospitals to AY-3-8500 based games) I was amazed at some of the analogue circuit design blunders I encountered.
In one case I built a piece of custom test equipment which needed +15V and -15V. An available transformer had two suitable
isolated
secondary windings so I just used two 7815 devices.
Only to be told by a more experienced "Designer" that connecting the output of a 7815 to ground would short it out and I had to
use
7915 for that.....

>
What he should have said was the you were messing up the ground returns by hooking up the +15V output of the second 7815 to the
0V
rail.
>
He was avoiding spending a long time talking about grounding and shielding, which is a rather specialised subject.

>
I see. Thank you for letting me know what he was thinking Bill.
Do you have some kind of time machine that you travel around in so you can have a better knowledge of what happened in a past
situation when compared with someone who was actually there at the time?
>

>
In this particular thread your eight transistor is bonkers, but it
works - not that I can see how.

>
Seems pretty obvious to me how it works Bill.

>
But John May could take out half the transistors and produce a circuit that still worked.

 One reason for that is that a current source/sink can be approximated by a high value resistor so instead of using a current mirror
a single resistor is likely to work fine.
You might need to reverse the polarity of the voltage drive but there are many ways to do that in this circuit.
 

And you haven't made any attempt to explain how it does work.

 As I pointed out some time ago, you never explain how your circuits work Bill.
It's like you just expect others to accept that you have superior knowledge.
If you really could see what other people think then I think you might be surprised at the amount of laughter going on.
 

>

If I didn't know how it was intended to work then it almost certainly would not work.

>
And  suspect that that there were a lot of variations that didn't work

 You can suspect what you like but that's just another example of you twisting the world into something you would like it to be.
 Anyone with any understanding of this circuit at all knows that a very basic explanation of how it works is that you need to
multiply the gain control level by a sample of the oscillator signal and feed the result back into the oscillator.
 So if your signals are in the form of two currents which never change polarity then a one quadrant current multiplier should work
fine.
 You might then do some research to see if you can find any examples of circuits which meet that requirement.
 https://www.google.com/search?q=one+quadrant+analog+current+multiplier&udm=2
 Much like you might refer to Horowitz/Hill to see if there are any relevant example circuits.
 

>

I'm surprised that you can't see how it works.

>
I've got a rough idea, but not good enough to let me do anything interesting, or to write any kind of useful explanation of what
is going on. Writing that kind of explanation is usually a necessary part of getting a circuit into production.

 I've yet to see you write an explanation of how a circuit works Bill.
Instead I've only seen circuits using three or four times as many components as the last one JM posted in this thread and never any
better than 70dB.
Then you get upset when the rest of the world tries to tell you that although your circuit may have a few good ideas in it, it's
worthless from a meeting the requirements point of view.
 

Management needs it before they will spend the extra money on document the design for production, and the final test technicians
and service engineers need it make sure that the circuit works and to be able to fix it when it doesn't.
>

John May could see how it worked, and how it could be simplified to four transistors - and eventually down
to two matched pairs.

>
You seem to be very good at knowing what other people can/could see Bill.

>
It's a necessary engineering skill. See above.

 I prefer not to make any assumptions about what other people can or can't see.
Particularly when I know nothing at all about them other than what has been posted here.
 

>

He doesn't like it, and prefers the single
long-tailed pair approach, despite the fact that LTSpice says it
offers poorer performance.

>
He provided three examples. All of which obviously work.
The best one from my point of view is the one with lowest distortion.

>
Which you haven't measured on a real circuit yet. Since you don't seem to be all that sensitive to the risks from power supply
feedthrough, doing that might be painfully educational

 Of course Bill. I have absolutely no doubt that you would take great delight in my pain.
And I have no doubt that you would not offer any meaningful assistance with the power supply design because you would rather watch
me suffer.
Offering assistance could reduce my suffering which wouldn't do at all.
 I'd probably use 6V batteries arranged for +/- 18V for a first test of a circuit like this.
With some series components for filtering and to stay below LT1679 36V max.
 

>

After all the goal, when I started looking into sinewave oscillators, was to see whether it's really necessary to use
thermistors,
lamps, opto devices or FETs.

>
Have you come to a conclusion yet? As I seem to have mentioned before, it strikes me that the answer is that it's obvious
necessary.

 What's obviously necessary Bill?
The last circuit JM posted has no measurable distortion worth mentioning that I can see.
 

>

I like the long-tailed pair approach myself - I can see exactly how it works - but it's probably worth my time to work out
exactly
how the four transistor circuit actually works, and why it seems to offer lower distortion in LTSpice simulations. It's not the
kind of project that anybody would fund, and the chance that I'd learn anything interesting is remote.

>
Ah well I'm sorry if you aren't likely to learn anything interesting Bill.

>
That's text-chopping.
>

I've learned a lot and will continue to do so.

>
We can hope.

 I'd translate that into "I hope you never learn anything so I can continue to tell you how stupid you are".

That is a remarkably hostile reaction. Most engineers can eventually learn stuff, but some are slower at it than others. Encouraging the slower ones takes quite a lot of effort because you have to go into quite a lot of detail before you can get the message across, and it can get tedious, and isn't all that rewarding. I'd love to be able to congratulate you for your insights, but they've been a bit thin on the ground.

, but I'm not swamped with work at the moment.

>
There are other things I should be doing but analogue circuit design has a certain attraction, particularly when assisted by
someone
as capable as JM.

>
I doubt if thinks that he is assisting you. Correcting is probably closer to the mark.

 Of course Bill.
https://www.youtube.com/watch?v=Z9t_KDGqOmE
 I have no doubt that your own circuits never need to be corrected Bill.

You must be joking. I've had some very clever colleagues who could do it fast. There's at least one case where I didn't spot my error until years later - the original solution worked fine but I eventually realised that there was a simpler way of getting there.

Maybe that's why they are so unnecessarily complex.
If no-one can see how it works, no-one can correct it.

It's not a matter of seeing how it works, so much as finding the right point of view on the way it works.
I could put together transformers that worked for a long time before I got to grips with the transformer equation, which gave me a clearer insight into what was going on.
--
Bill Sloman, Sydney