XR2206

[9H1FQ]

Hi all 1 Need some help
I have just finished up an XR2206 based function generator.
The output is over 2 volts into 600 ohms. Would like to variable attenuate the signal, down to a few millivolts keeping 600 ohms load out. Without the use of any op amps, just clean direct output from the chip
Any suggestions please !

[K7JEM]

The data sheet shows how to do this, using a variable resistor (R3) at the pin 3 input.

http://www.jaycar.com.au/images_uploaded/XR2206V1.PDF

Joe

[AC0FP]

Hi all 1 Need some help
I have just finished up an XR2206 based function generator.
The output is over 2 volts into 600 ohms. Would like to variable attenuate the signal, down to a few millivolts keeping 600 ohms load out. Without the use of any op amps, just clean direct output from the chip
Any suggestions please !

As JEM has said the output amplitude is controlled by the resistor value connected to Pin 3. Have you actually measured the output impedance of the IC? You are aware that just because the IC delivers 2 volts into 600 ohms, that doesn't mean its output impedance is 600 ohms.

fp

[9H1FQ]

Thanks for your advice
What I really need, is a variable attenuator, possibly switching, or tapping a series of fixed resistors

[K7JEM]

What exactly are you trying to do? Do you need a precise 600 ohms, or will something in that range work? What levels do you actually need to cover?

A simple resistor divider may be all you need, or a simple 1K pot and a couple of fixed resistors, if your requirement is not for "precise" 600 ohms.

Joe

[VK2TIL]

The data sheet says that the sine/triangle output impedance is 600 ohms.

A 600-ohm step attenuator would serve.

An ebay search for "attenuator 600" produces some results; prices don't seem high as 600-ohm devices are not as sought-after as 50-ohm devices.

You could also consider building an attenuator; the classic ARRL design is in the older Handbooks and an updated design is here;

http://www.arrl.org/tis/info/pdf/9506033.pdf

Design for low-frequency would be more-forgiving than for HF/VHF; the extra features of the updated design would probably not be worthwhile.

A search for "homebrew step attenuator" gets several results that should give you ideas.

Resistor values can be calculated;

http://chemandy.com/calculators/matching-pi-attenuator-calculator.htm

[AC0FP]

The data sheet says that the sine/triangle output impedance is 600 ohms.

A 600-ohm step attenuator would serve.

http://chemandy.com/calculators/matching-pi-attenuator-calculator.htm

I wouldn't bet on that data sheet specification after looking at the schematic/block diagram of the chip.

Its easy to tell however, just put several different load impedance's like an open, 2k, 600, or 300 ohms on the output and see if the peak/peak voltage changes.

fp

[VK2TIL]

There's a resistor in series with the output to pin 2; value not given but I imagine it's there to set the output impedance.

Any attenuation added after that, as the OP proposes, will tend to "force" the final output impedance closer to 600 ohms anyway.

[AC0FP]

There's a resistor in series with the output to pin 2; value not given but I imagine it's there to set the output impedance.

Any attenuation added after that, as the OP proposes, will tend to "force" the final output impedance closer to 600 ohms anyway.

Your right of course, the resistor in series with the output was not present in the diagram I looked at.

fp

[9H1FQ]

Pin 3 is not used for leveling, but for trimming distortion, offset bias etc




The data sheet shows how to do this, using a variable resistor (R3) at the pin 3 input.

http://www.jaycar.com.au/images_uploaded/XR2206V1.PDF

Joe

[K7JEM]

Pin 3 is not used for leveling, but for trimming distortion, offset bias etc

From the data sheet:


Output Amplitude:
Maximum output amplitude is inversely proportional to
the external resistor, R3, connected to Pin 3 (see
Figure 3). For sine wave output, amplitude is
approximately 60mV peak per k
of R3; for triangle, the
peak amplitude is approximately 160mV peak per k
of
R3. Thus, for example, R3 = 50k
would produce
approximately 13V sinusoidal output amplitude.

If your application doesn't require a precise 600 ohm impedance, and most don't, all you might have to do is attach a 1K pot to the output of the chip, and adjust to what you need, level wise. If you need an adjustable output at a precise 600 ohms, you could add a 600 ohm series resistor from the output, and a 1200 ohm resistor to ground. At the junction of the two resistors you will have a very good 600 ohm source, padded down about 6 dB or so. Make the resistor at pin 3 a variable (as per data sheet), and you will be able to adjust the level over a pretty good range while maintaining that desired 600 ohm output.

Joe

[KB0MNM]

It sounds like you are pretty determined to see that the load does not see a lower IMPEDANCE than 600 ohms. Try working backwards: 1K to 8 ohm transformer, L-Pad 8 ohm (wiper arm) and measure the impedance with a low-power Z bridge. Failing the bridge, use a 1K pot to ensure your source has a light load ( high impedance ) path. Confirm clean waveshapes with a scope. This assumes that you can only obtain audio-grade L-Pads such as Waldom part number 30-372 available from Newark www.newark.com (http://www.newark.com) and possibly Farnell. The L-pad is pricier than you might think ( $9.00 to $25.00 US ), but is a recognized way to load-balance the output of an audio amp into many variable speakers all in parallel- allowing individual control without affecting the source loading too much. The system described may attenuate your 2V signal beyond use at 600 ohms- in which case a buffer transistor could be inserted before the L-pad. 2V is plenty for forward-biasing a silicon transistor( try .7V) - use class "A" bias. Or- The switch step-attenuator is a cheaper solution, try three ganged wafers. 73 KB0MNM- Jon:)

Hi all 1 Need some help
I have just finished up an XR2206 based function generator.
The output is over 2 volts into 600 ohms. Would like to variable attenuate the signal, down to a few millivolts keeping 600 ohms load out. Without the use of any op amps, just clean direct output from the chip
Any suggestions please !