Low Distortion Oscilator schematic anybody?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
tschrama said:
Hi All,

I've done a search on this on the forum and indeed found some, but some values are missing or the link is dead. Maybe someone has got a nice little schematic for <0.01% oscilator, fixed frequency is OK.
Have you considered an ordninary sound card?

My "junk" card gives 0.0087% distortion, Turtle Beach Santa Cruz.

Tone generator (excellent freeware): http://audio.rightmark.org I haven't checked if the software can produce sinus signals but I think so.
 
this idea appeared in AudioXpress -- you take a square wave generator (which can be a simple multivibrator type, xtal etc.) and pass the signal through 2 stages of filtering with the Linear Tech LT1063 low pass filters -- the distortion will be well below what you are looking for and the filter is very easy to tune

you just use a variety of counters to select the frequency

it helps if you have a "post filter" to remove the vestigial clocking noise from the circuit.

this won't get you a lot below 0.01%, but in the ballpark and without the need of ganged potentiometers, etc.

I have even experimented with this using the PicBasic Pro FREQOUT function on a 16F84 multiprocessor and gotten good results.
 
Hi everybody,

I just read the LT1037 opamp datahseet ...

The oscilator shown uses a #327 lamp ..... hmmmmmm I gues that means 5 Watt ;)

Can anybody give me some clues about what kind of lamp I need .. something like 12V / 200mA or more like 6V / 10mA?

Regards,
Thijs
 
Re: Re: Low Distortion Oscilator schematic anybody?

Have you considered an ordninary sound card?

Hi,

The advantage of this is that you get a distortion profile, not just THD + noise. Here a picture of my 200 bucks 24 bit soundcard in loop through mode, line in connected to line out. Measuring ultra low distortion is not easy. Noise and hum entering through cables and caused by ground loops are not easy to cope with.

Good affordable software for an audio test bench can also be found here:

http://www.audiotester.de/mainE.htm

Cheers ;)
 

Attachments

  • dist2496.gif
    dist2496.gif
    3.8 KB · Views: 459
The M-Audio Audiphile 24/96 can manage this with ease, and it's not that expensive.

Makes a good basis for a PC-based analyser.

Frequency response (from 40 Hz to 15 kHz), dB: +0.00, -0.00
Noise level, dB (A): -101.4
Dynamic range, dB (A): 101.0
THD, %: 0.0003
IMD, %: 0.0042
Stereo crosstalk, dB: -99.3
 

Attachments

  • thd.png
    12.6 KB · Views: 363
Hi,

First, thanks for or the links. I got around measuring my own sound card in 'loop through' mode, but was a bit dissapointed... I use a asus 8NX-X motherboard with sound-on-board. The specs are not that good.

So I read more about the wien-bridge, I think I understand the principles now...

Got myself a few lamp: I understand now that:
-you need low current draw to make it easy for the opamp to drive
-you need large mass to keep thermal-dissipation time constant large
-all lamps typically have a 1/10th impedance at zero current compared to rated current and you need about 1/10th of the rated current to make some practical use of the possitive themperature thingy...

So a 230Volt / 3Watt lamp would be around 1.2KOhm at zero current and 12KOhm at rated current: perfect, it will draw about 1mA (that's low enough), requires 3k3 feedback resistor (high enough not to stress the opamp), and gives about 3 Volt out.

I got 0.006 % :) distortion, allmost all second harmonic (????) using a NE5534 opamp. I also used ad797, opa627, opa134, opa228, LT1028, LT1115, op27, op177, u747, lm308, .. alllllllllll my opamp in stock actually .... ne5534 had lowest distortion, which supprissed me... The high speed LT1028, ad797 and LT1115 all needed a output resistor of 470 or higher to drive my coax BNC cable.. the NE5534 worked fine without it.

Then I build a two-opamp inverting mode wien-bridge that help to limmit common-mode distortion... it works! ... 0.0045% :) using a opa134.. opa132 and LT1113 has slightly higher distortion, but it was difficult to see.

I also used a 6V/50mA lamp (much to fast setteling time, high distortion) and a 230V/15W lamp, which, I think, needed to much current to give low distortion. The 230V/3W lamp seems perfect, time comstant is about 0.8 sec.

Well that it for now .... goodnight ;)
Thijs
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.