The only triangle to sine convertor I have seen is in the Waveteks. I see alot of generators using Wein bridge oscillators.
My Wavetek does use a triangle to sine convertor and is rated 0.5 % distortion, which is good enough for troubleshooting but not measurement IMO. I notice with age however is does have some visible distortion at very low frequencies.
My HP 339A I believe uses a Wein bridge and is down to 0.0016 % which IS good enough for measurement but I am not going to break it out for troubleshooting. the Wavetek is much better for that and has alot better selection of output. Also, mine is a VCG which means I can frequency modulate it. Came in handy checking the operation of a CD-4 demodulator. It also has a synch function, though I'll probably never use it. But who knows ? Better to have it and not need I say.
My Wavetek is a 111 and I only changed filter caps in it and put on a power cord. My buddy has a 103 which is broke down at the moment, partly because he took it apart and the wires are not marked. But initially it had a problem I was having trouble running down. the Waveteks of that design to startt with a triangle wave, and I gotta admit I am impress by the sine convertor. That they got it down to 0.5 % is impressive. It is a three stage dual compressor. I uses diodes and transistors to accomplish this - nothing reactive. (which means the distortion in mine at low frequencies is probably a coupling cap to the output somewhere) Talk about precision.
His had a small amount of DC offset in the main triangle generator that would just not go away. The SOB drove me nuts. I even found resistors that were off spec but changing them did not fix it. So I took and put a capacitor inline to the sine convertor and got a nice sine wave out of it. But he is a bit anal about test equipment and took it apart, and the way those things are, well the wires are all over the place and they use these funky connectors that are not insulated at all. But really, the design (electronically) looks like something Tektronix would do.
The 111 is another one of my things that might be a POS but is not for sale. It puts out almost 40 volts at 600 ohms. With that kind of output I can feed a chopper transistor in an SMPS if needed, unless the thing has no gain at all. I can also connect it directly to a speaker and it will drive it in free air well enough to align it during a refoam. I know there is alot of loss but it does not seem to make the generator distort, which is what is needed.
So now my buddy has this HP which has less distortion but nowhere near the output. Now we need to throw together some kind of amp for it I guess. I think he has the famed 204A of something like that. All fine and good, but the output just has no balls.
Y'know for alot of things you can just use the calibrator of a scope. Not for everything, but for quite a few things. Being a square wave it is no good for finding the source of distortion, but you can see where the signal gets lost. I use them to test capacitors.
Anyway, if you can find a good working HP 339A you will have your audio generator. Now that I got mine go ahead and buy one and raise the market value all you want
. I paid $200 in not good cosmetic condition, and a few broken knobs. But it works. There is supposedly a Cds cell in there that goes bad and finding one is a PITA and then replacing it requires alignment usually that I have no idea how to do. I guess I could figure it out but better it just works.
Another thing I have (recently found in my junk) is an audio test CD from MCM. I have lost the paper TOC and the case but still have the CD and it plays. I wonder how that compares to a generator. I bet it is pretty good but you can't set anything. You get the frequencies they give you and that is that. I also have a DVD of test patterns that you used to need a $600 generator to get.
But now, if you download those tones and whatever and use a high end DAC, wouldn't that be the highest quality ?
