Low Distortion Signal Generator

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
(Mods...had no idea where to put this...seemed more related to testing of the SS gear I fool with, so here it is. Move if you feel appropriate)

For years I've had a very old Waveforms tube sine-wave generator, but lately it has developed some problems so I'm looking for a more modern replacement.

I'm using a distortion analyzer to check out my work, and I'd like distortion of 0.01% or less. I have a opportunity to grab a used Leader LAG-126, which is supposed to be capable of less than 0.005% thd. Attractive specs, but it will cost around $200 US. Seems high for a used piece nearly 10 years old.

What other ultra-low distortion generators are out there? What do you all use? I'm really too busy to build something, and the DIY circuits I've seen are not capable of distortion readings as low as the Leader I found.
 
I stuck it in "Everything Else" for lack of a better place to move it.

In any case, what frequency range do you need? If you don't need really high frequencies, two really cheap and easy solutions are using a soundcard and software or a CD player with a volume control. The distortions are surprisingly low- with the soundcard that came stock with my POS HP, I get 0.01% THD at 1 kHz. S/N ratios can be in the -90 to -100 dB range.
 
I also considered the soundcard option, but I'd really like to be able to go to 100KHz, which a soundcard ain't gonna do. Still, for distortion measurements, the standard audio frequencies would be OK.

I have a computer near my work area with an Aureal soundcard in it. Using software from Daisyweb Labs. It's actually a pretty decent soundcard, but with this software the best it can manage is about 0.2%. Better software maybe? What did you use?

Also, the concerns that John voices are also on my mind. Plus, a computer is pretty cumbersome commpared to a dedicated piece of gear.
 
I used DACtester, with the level set up as high as I could go. If you don't use all the range of the DAC, distortion will be higher. I've had good results with the Daisyweb signal gen, but haven't measured it in detail to see how it compares- the user interface is clusier than I'd like.

John, there may be higher order stuff there, but I couldn't see it with the HP3581A you loaned me. If I were worried about it, I'd just throw together a simple lowpass filter. But if it's good enough for HP...
 
I use a Wien-bridge oscilator that I made myself. To get the lowest THD I use a 220Volt/3Watt lamp for amplitude stabilisation and it works fine .. < 0.001% THD.. opamp was a either a opa627 or NE5532, made no difference for THD.. all in all is is an enormously little simple thing that has very low THD... I think because I used a lamp with the biggest thermal timebase I could find..
 
I use a Loftech TS1 for frequency response and general testing and an HP 204CD for distortion. There's tons of the HP's out there for cheap. Even if you don't want to build one, Linear Technology published an excellent article on sine wave oscillators. It was in one of their data books from the mid 80's.
 
Member
Joined 2004
Paid Member
Commercial Low Distortion Oscillators

I built Jim Williams oscillator and it works. However it took much longer and was less robust than I had hoped. I still have some boards left over.

Commercial Ultra low distortion Oscillators are a short list.

Oscillators only:
1) HP 239
2) Tek SG505
3) Krohn Hite 4400, 4403 (and i think the 4500)
4) Boonton 1100
5) Sound Technology (I don't remember the model numbers)

There are a few really obscure others. Better value would be a complete audio analyzer. Best deals are Sound Technology 1700 or 1701 or 1710. All of which have a system residual below .002% THD. Older ones can be tricky to fix and cal.
Or an HP 339 or an Amber. Docs on the Ambers seem hard to find.

I have several Boonton 1120's that all work real well with residuals of .0008% when calibrated.

All of the above can be had for less than $300 on ebay. The next step up is an Audio Precision for lots more $$$

FYI the best THD on an HP 200CD is rated at >1% and the HO5 takes that down to .05% And I pitched several a few years ago as being too much work to fix, needing caps, tubes and ???.

Good searching -Demian
 
I very much appreciate your input. Can't find any of these for sale at the moment on eBay...the Tek505 sells for over $300 in most cases, and still needs a mainframe (although I saw a 504 with some sort of Tek power module that bolted to it to get the job done...was a sweet setup and I wonder if that sort of module would work on the 505?)

I take it that you aren't too spiffy on the Leader LAG-126? Distortion specs look pretty good. Also, Kenwood makes a low distortion AG-253, but it don't sell cheap either...
 
Member
Joined 2004
Paid Member
Your right, There is a dearth of stuff on eBay right now. I would wait a few weeks before I start compromising, there is a steady turnover in this stuff. I would wait for a ST 1700A which will do everything you could want and they are typically around $150 to $250. I have not been impressed with Leader for the most part. If you are at the stage where you can make use of a low distortion source then the whole package makes a lot of sense. And a full analyzer is faster to use. And easier than a PC system.

The RE204 (http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=25423&item=3840625495&rd=1) may be an option but I don't know enough about it. And with 6 already I can't justify getting another analyzer.
-Demian
 
Hi,

How about an easy to build 1khz sinewave oscillator with thd
of .01% or less, with stable amplitude ? My 8038-based
generator is just not good enough for distortion testing.
Hobbyists need a simple one like this and with a diy 1khz
filter plus a diy millivoltmeter one can see if one's design
is good enough.

efren
 
rcavictim said:


The old tubed HP gear was made so well and designed so conservatively that this piece probably does still run clock-wise. CW stands for Continuous Wave.
Lol!! I see...

The RE204 still needs a GPIB card, and it will have to be a PCI, as none of my systems have a ISA slot. And I can't find any info on what it's capable of.

I have a nice Tek 4084 distortion analyzer that works wonderful, but has no generator.

efren sabio... My Waveforms tube gen was capable of 0.01 when it was working, so I'd really really like to do better than that.
I have several Boonton 1120's that all work real well with residuals of .0008% when calibrated.
Sell one?:)
 
8038 way too much THD

about 5% typical.

1 - wein bridge - agreed I've made these with very low THD. Frequency accuracy is poor, but that's not usually an issue.

2 - A uController driving a CODEC or DAC. You can hard code a sine wave into memory. Use the timer to set up different freqs. Battery powered, portable, dead on freq accuracy. THD equal to DAC's capability

3 - Take a crystal-based oscillator. Divide to 100x desired frequency, then divide the 100X. Use the 100x signal (i.e. 100kHz) to clock a switched-capacitor filter. Apply the 100x divided signal (in this case, 1kHz square wave) to the input of the SC filter. The output is a very pure sine wave with exact frequency. You will need to add an external filter to remove the SC clock feedthru.

3 - All methods. Add an LC based low pass filter to the output. A third order filter suppresses the second harmonic by 18dB, and the subsequent harmonics by even more. This will make a good sine wave source better.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.