|
Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members. |
|
Thread Tools |
22nd Mar 2020, 11:10 am | #21 | |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: Low distortion oscillator
Quote:
Though you couldn't really call it hifi, the Texan made a lot of people happy for several years. All the "Um, could you have a look at this thing I've been making..." preambles served as aversion therapy for me. So I built a Mullard blue book 25 watter for a friend, and the later edition 50 watter for myself. My disc preamp was from Hugh Walker's mixer articles in WW. After that I went with own-designs. I wonder just how many Texans were built? I guess there are some that ran into problems still hidden out of mind in cupboards around the country.
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
|
22nd Mar 2020, 11:19 am | #22 |
Nonode
Join Date: Jun 2016
Location: Bognor Regis, West Sussex, UK.
Posts: 2,288
|
Re: Low distortion oscillator
I built a Texan, it did work and eventually sold it on. I also settled on a Mullard blue book 25 Watter which I used for many years.
For a while it was used at work, we sold some very small LVDT tranducer amplifiers that ran on 240V, we were delighted when we won an order from the Swedish nuclear industry for a signicant quantity. We were less delighted when we saw the testing specification. One requirement was to vary the mains frequency over a range (I think it was 50Hz +/- 10 or 15%) I connected a signal generator to my Mullard amp and drove a 6V transformer in reverse to give the required voltage. It certainly worked but would blow the output transistors occasionally. Fortunately my firm paid for a plentiful supply of replacements. Peter |
22nd Mar 2020, 1:19 pm | #23 |
Dekatron
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,942
|
Re: Low distortion oscillator
Mine worked fine. Built the preamp as an individual board first, and then added a second board housing the power amps. It never missed a beat.
Gave it to my dad (long time gone now), then powered our kitchen system here for a number of years, and now it resides in my loft. I'm betting if I power it up it would work perfectly. Well as perfectly as a 741 ever did! |
22nd Mar 2020, 3:06 pm | #24 |
Hexode
Join Date: Apr 2007
Location: W Yorks, UK.
Posts: 406
|
Re: Low distortion oscillator
If you only want one fixed frequency, one of the easiest approaches is to build a basic sinewave oscillator without worrying about ultra low distortion, then filter the heck out of the signal. Keep filtering and re-amplifying (with decent opamps of course) until you reach the residual distortion level you're aiming for.
|
22nd Mar 2020, 3:41 pm | #25 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: Low distortion oscillator
Oddly enough, if you are trying for very low distortion, making a filter which doesn't produce its own distortion is as big an undertaking as a low distortion oscillator.
If you build a passive L-C filter you run into problems with the linearity of magnetic cores and to a lesser extent, even with capacitors (which is where we circle back to Bateman's articles and the purpose of his work) If you build an active filter, you're back with the issue of the linearity of amplifiers whether discrete or monolithic. At least you don't have to juggle with the business of stabilising oscillator amplitude at the same time. I did go down the filter route when I needed very low distortion (for the era) sources when I was playing with power amp design. I needed comparable bandstop designs to tart up the performance of the measuring receivers. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
22nd Mar 2020, 5:02 pm | #26 |
Dekatron
Join Date: Jun 2015
Location: Biggin Hill, London, UK.
Posts: 5,191
|
Re: Low distortion oscillator
My late father assembled a Texan from the kit from Henry's Radio. He used it for the rear channel speakers of a quadraphonic setup (remember those?). I still have the Texan, I must get it going again.
But the real reason for this post is that some years ago I was given an instrument that was otherwise going in the skip 'You might get some useful bits from it'. It turned out to be a Radford DMS3A distortion meter. I didn't know how respected they were back then but I kept it and didn't raid it for part. What I don't have is a suitable low distortion oscillator to use with it. Should I try making one? |
22nd Mar 2020, 6:42 pm | #27 |
Dekatron
Join Date: Nov 2010
Location: Oxford, UK.
Posts: 4,942
|
Re: Low distortion oscillator
I tried looking for the spec of the Radford, and there is a dearth of information out there. I found mention that it is good to 0.001% (10ppm) which is pretty good, and a useful number. You should really be aiming for ten times lower distortion in a test signal (ie oscillator), so 1ppm. Doing that and providing a variable frequency is not straighforward.
Oddly I found this document while searching, which is an interesting read (I just did a quick scan) http://www.angelfire.com/ab3/mjramp/distortiontest.pdf This is the guy's website http://www.renardson-audio.com/ which seems to have a load of interesting stuff on there, Craig |
22nd Mar 2020, 7:38 pm | #28 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: Low distortion oscillator
Of course, the Radford distortion meter is the perfect instrument to use when developing an oscillator to go with it. It may floor out at 0.001% but when you've developed an oscillator good enough to show only little above that floor, then you've reached the 1ppm sort of level that Craig mentions. You don't need to find a 0.0001% distortion meter to check your oscillator.
Remember to keep a sense of proportion. The hardest thing is to know when you've done enough. A good many respected and sought after amplifiers have distortion figures orders of magnitude more than this level. One of the best bits of a distortion meter is to stick a scope or spectrum analyser on the distortion residual output and look at the frequency content coming out. As said above, low order distortion products are significant parts of music and voices. Higher order ones sound unnatural and get noticed. This might be a good application for a soundcard/PC/Software spectrum analysis app. I gave my brother a distortion meter as a christmas present a few years ago. He messes around with guitars and guitar amplifiers. I said he might find it useful to check that he was getting enough distortion David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |