Low-distortion Audio-range Oscillator

I am looking to add a low distortion audio sig gen to my bench. I have been watching eBay for good deals. Any worth bidding soon rise above my budget.

Anyway, I starting looking into a DIY Signal Generator. My research turned this up. Low-distortion Audio-range Oscillator

Anyone ever build this? Sounds good but is it? There is no build tips or PCB layout. Just a diagram and parts list.

This author has a book of all his circuit designs but comes at a steep price. However, my local library has a copy and I will pick it up sometime this week.
 
Looks like a variation of the Wein Bridge oscillator. Look at the schematic of the HP 209A. There is much info here leading to even improving the 209. Also check out the oscillator in the HP339 or 8903 distortion analyzers.
But for low distortion, the Wein Bridge design is the way to go.
Doc
 
For a reasonably well-documented DIY project, look at the oscillator section of Robert Cordell's THD Analayzer at < http://www.cordellaudio.com/papers/thd_analyzer.pdf >. Even after 30 years, the performance holds its own against the best commercial units. This design has been discussed in several threads here on the Forum, e.g. < http://www.diyaudio.com/forums/solid-state/154260-my-implementation-cordell-distortion-analyser.html >. For that matter, the general topic of low-distortion oscillators has been discussed in quite a few Forum threads, e.g., < http://www.diyaudio.com/forums/solid-state/141271-can-we-improve-low-distortion-sine-oscillator.html >. Use the "SEARCH" feature to locate others.

A currently active DIY project using the same basic circuit architecture as Cordell's oscillator can be found at < http://www.users.on.net/~glenk/uldo/uldo.htm >.

For an introduction to the topic, look at < http://www.moorepage.net/RC.html >. Dick Moore's web pages (see < http://www.moorepage.net/index.html >) include several DIY efforts to upgrade or optimize the performance of several commercial test equipment models. A similar effort can be found at < http://www.tronola.com/html/ig-18_mods.html >.

Dale
 
The lowest budget solution is to record test tones to a CD and use a CD player as your source.

But for everything you need to know about oscillators, Jim Williams did an application note way back in June 1990 about Bridge Circuits for Linear Technology AP #43. I don't have a current link. It had pages on great oscillators.
 

coluke

Member
2009-03-23 7:44 am
Anyone ever build this? Sounds good but is it? There is no build tips or PCB layout. Just a diagram and parts list.
I've built it a few years ago, with a few modifications (a bit of filtering after the full wave rectifier): it's a very simple and clever design (the THD cancellation trick is lovely), and it is indeed capable of low settling time and very low THD - 0.001% or better mid-band:

APFGEN.png

but low settling time has a prize: amplitude flatness is far from optimum (with no integration in the level controlling loop the steady-state amplitude is a bit undeterminate...), and THD+N is somewhat high. If you are looking for something simple and fast to build it is certainly a good choice, but if you are managing to do some serious audio work then you should go for an SVF design - something like Bob Cordell's generator; or, maybe, improve a bit this generator with a better level controlling loop - never tried, but I think suprisingly good results are obtainable.

Ciao,

L.

ps - this is not a Wien bridge variation, but an all-pass filter; the differential output stage (IC2B) is designed to notch out almost exactly third harmonic and reduce somewhat higher order odd harmonics produced by the full-wave detector - that's why output THD is so low.
 
Should anyone be considering a build of the HP339 oscillator clone as referenced in post #3 at Modifications: Heathkit IG-18 Audio Generator, Newark has only 42 pieces of the HA3-2625 IC in stock @ 5.92 ea. After they are gone part probably will no longer be available. INTERSIL|HA3-2625-5|IC, OP-AMP, 100MHZ, 35V/µs, DIP | Newark.com The HA2625 has some unique characteristics and it is unknown whether a newer chip can replace it. I have some of these but am holding off building it until I see what Dick Moore has in mind for his IG18 #3 mod. IG-18 #3
 
Should anyone be considering a build of the HP339 oscillator clone as referenced in post #3 at Modifications: Heathkit IG-18 Audio Generator, Newark has only 42 pieces of the HA3-2625 IC in stock @ 5.92 ea. After they are gone part probably will no longer be available. INTERSIL|HA3-2625-5|IC, OP-AMP, 100MHZ, 35V/µs, DIP | Newark.com The HA2625 has some unique characteristics and it is unknown whether a newer chip can replace it. I have some of these but am holding off building it until I see what Dick Moore has in mind for his IG18 #3 mod. IG-18 #3


Do you have any idea what the approximate cost of parts would be less the Heathkit? Rough estimate.

This looks like a project I could do. Thanks for sharing.
 
A very rough guess might be $80-90 if you need to buy everything. The switch is the most expensive part @ about $23.00 from Mouser. The large frequency caps also tend to be expensive for good film. It will be a bit more if you decide to build a pc board, a group buy might be nice. I believe this one is an improvement in distortion over the design you first listed. The Cordell design as shown on Dick Moore's page is also a top notch performer but perhaps a little more difficult to build.
 
I am looking to add a low distortion audio sig gen to my bench. I have been watching eBay for good deals. Any worth bidding soon rise above my budget.

Anyway, I starting looking into a DIY Signal Generator. My research turned this up. Low-distortion Audio-range Oscillator

Anyone ever build this? Sounds good but is it? There is no build tips or PCB layout. Just a diagram and parts list.

This author has a book of all his circuit designs but comes at a steep price. However, my local library has a copy and I will pick it up sometime this week.

I too liked the simplicity of this. Here's my pcb layout. I mounted + / - 15V regs on the pcb.

Brian.
 

Attachments

  • LedOsc.doc
    109.5 KB · Views: 1,250

coluke

Member
2009-03-23 7:44 am
I wonder if the distortion would be lower with something like the LME49740?

I think the answer is definitely... no. The distortion products you see in the spectrum I've posted are mainly (or entirely) due to the peak detector, whose time constant is zero - no filtering at all means very low settling time (which is good), buy the price you have to pay is that all the harmonics produced by the full wave rectifier are injected in the level control loop, i.e. in the output signal. In order to lower the output THD you have to set up an integrating loop - maybe you can try the one from the JLH's original design:
JLHWB.PNG
it's a Wien Bridge ring-topology (in order to get rid of the opamp common-mode THD), and the amplitude control loop is based on a canonical peak-detector/integrator chain (whose time constant you'll have the to tailor in order to suite your needs - ie freq range and output amplitude).

L.
 
Thanks. I don't understand all of this, so I'll just leave it at that. I can stuff a PCB though....

I kind of feel this way too. I can stuff a board. There are so many different ways to approach this. But I am learning. I have read everyone's response to my opening post. I opened every link and read what I could. I googled these different approaches and found more info to study.

They all have merit, but DDB's post on the modified Heathkit is the one I understand. Mainly because I can visualize the end product that will sit on my bench.

So this is route I am leaning towards.

Thanks everyone for your responses.