Low-distortion Audio-range Oscillator

[dirkwright]

Originally Posted by EssB

The schematic in fig. 8.4 is a bit fuzzy but readable enough.

If you can read it then why don't you post it? Obviously, if I could read it then I wouldn't ask him for a clear copy!

Thanks for the link. very useful!

[EssB]

Because you already have it! see post 532 Low-distortion Audio-range Oscillator

[dirkwright]

Originally Posted by EssB

Because you already have it! see post 532 Low-distortion Audio-range Oscillator

I'm sorry, I don't follow you. I have that manual. That's where I got the fuzzy figure. There is no figure 8.4 BTW.

Do you mean the attached schematic instead?

[EssB]

Originally Posted by dirkwright

I'm sorry, I don't follow you. I have that manual. That's where I got the fuzzy figure. There is no figure 8.4 BTW.

Do you mean the attached schematic instead?

Yes!
It says Fig. 8-4 in the bottom right corner

[1audio]

The oscillator circuit it the same as the HP339a, schematic available here: http://bama.edebris.com/manuals/hp/339a look at pgs 103-106. Much better than the HP239 scans, just chopped into several pages. Much simpler than the KH4025 circuit but slower settling.

[coluke]

Here is a better quality 239A (simplified) schematics.

Resistor values:
Ra = 10k (X 10 range), 1k (X 100, X 1k, X 10k range)
Rb = 100k (X 10 range), 10k (X 100, X 1k range), 1k (X 10k range)
R49 = 2k
C42 = 220u

Ciao,

L.

[dirkwright]

Originally Posted by coluke

Here is a better quality 239A (simplified) schematics.

Resistor values:
Ra = 10k (X 10 range), 1k (X 100, X 1k, X 10k range)
Rb = 100k (X 10 range), 10k (X 100, X 1k range), 1k (X 10k range)
R49 = 2k
C42 = 220u

Ciao,

L.

Thank you!

So, bear with me a minute about what they are dong here. Are they changing these resistors and capacitors for each range to change some kind of filter for each range? Are they changing the settling time based on which range is selected?

[dirkwright]

Originally Posted by 1audio

The oscillator circuit it the same as the HP339a, schematic available here: http://bama.edebris.com/manuals/hp/339a look at pgs 103-106. Much better than the HP239 scans, just chopped into several pages. Much simpler than the KH4025 circuit but slower settling.

Thank you!

[dirkwright]

Originally Posted by coluke

Here is a better quality 239A (simplified) schematics.

Resistor values:
Ra = 10k (X 10 range), 1k (X 100, X 1k, X 10k range)
Rb = 100k (X 10 range), 10k (X 100, X 1k range), 1k (X 10k range)
R49 = 2k
C42 = 220u

Ciao,

L.

Now that I see this circuit more clearly, what is the input impedance for the node where CR1 connects to the output of the oscillator? Since the other end of CR1 is AC connected to ground via a capacitor Ca, I would think it would be a dead short for either some part of each cycle or for some corner frequency that I can't figure out right now. Should there be a buffer to drive this diode?

[coluke]

Originally Posted by dirkwright

Thank you!

So, bear with me a minute about what they are dong here. Are they changing these resistors and capacitors for each range to change some kind of filter for each range? Are they changing the settling time based on which range is selected?

Generator dynamics changes with oscillation frequency, so you have to change ALC loop gain and response time in order to optimize every range for both low THD and short settling time (and stability, too) - in general a single time constant loop can't work (and actually won't) from 10 Hz to 100 kHz with standard peak detectors. Heavy filtering will guarantee low THD but very long settling time; light filtering will shorten settling time at the expense of a much higher detector injected THD - there's a trade off. In the actual 239A THD @ 1 kHz is lower in the 10 X100 range than in the 1 X1k range, but the generator won't probably work @ 10 kHz in the X100 range...

As far as the schematics is concerned, my 239A clone includes a buffer just before the full wave detector, and it performs actually measurably better than the original one.

L.