Confused about wiring of XR 2206

I am trying to wire the XR2206 chip so that I could adjust duty cycle with a pot:

I am using schematic from figure 14 and am following it.

I finally got this thing to produce some signal.

To my big surprise, when I turn the megohm pot wired to pins 7 and 8, frequency is adjusted and not duty cycle. What gives? Any thoughts?

My second problem is that I cannot figure out why amplitude modulation does not work. My signal has amplitude of about 1 volt instead of about 15 volt, as I would like. Changing voltage at pin 3 is not helping.

My third problem is that there s ripple on the - power rail that I do not like. I wonder why it happens.

I think that I understand what is happening, but not why.

What's happening is that pin 9 is level sensitive and, depending on the input level, switches frequency from being controlled by R1 (pin

The pulse width adjustable diagram on Fig 14 connects pin 9 to pin

1. Pin 11 is the square wave output. So, square wave itself changes the decision which resistor is used to control duration of the phase. Low part of the pulse uses one resistor, upper part uses another resistor. It makes duty cycle adjustable if R1 and R2 are obtained from a pot, as is my case.

The trouble is that the output level of pin 11, square wave output, is just too low, about +- 0.5V.

Why is that so, is unclear to me. I tried messing with the voltage on pin 3 (bias that supposedly sets output level), to no avail.

Well, the duty cycle is advertised as something like 1 to 98%.

That's right.

Just like the figure 14 says: middle leg is grounded, and one leg goes to 7 and another goes to pin 8.

In fact, I tested this setup a little. I removed the link from 9 to

1. If I apply voltage to pin 9 (whichever it is, + or -), the chip promptly switches to another resistor and instantly changes frequency.

OK. I tried messing with pin 3. I did the following, disconnected everything from it, used different resistors, connected it to Vcc and V-, etc. Nothing made any difference whatsoever.

Volgen PRK30U-1515.

The ripple is about 200 kHz. (roughly)

Yep. So, how could the output be such a low voltage?

I am quite stumped. Something simple eludes me.

One thing that I realized is that in some places electrolytic caps are demanded and I used non-EL caps. Not sure what is the significance.

Also, in a couple of places such as Vcc to ground next to pin 4, and next to pins 10 and 3, I used 2uF caps instead of amount specified. Removing these caps does not seem to make a difference. The amplitude of the square wave signal is too low to affect pin 9 and thus I have no duty cycle. I culd use some hack, but I would rather understand what is going on.

thanks a lot for reading this post.

(snip)

You should add an additional resistor between the wiper and ground, to set a minimum resistance for either input. The chip does not like to have either input grounded (unlimited current).

That certainly confirms the low swing on pin 11.

As I said, pin 3 is not involved in the output from pin 11.

(snip)

Is that a switching power supply? 200 kHz sounds like the ripple from a switcher.

(snip)

Is Vcc 15 volts?

Electrolytics are recommended only because they are cheap and small. The same value in a film or ceramic capacitor should work fine.

Not critical.

Sounds to me that something like a diode is across pin 9 to ground, or the output transistor has been blown, and you are seeing the internal base drive instead of the collector.

(snip)

(snip)

If 2 devices show this problem, how about posting a picture of your circuit, before you plug the third unit in? If there is something about your connections that is wrong, and in your blind spot, someone may save that third piece from damage.

I agree. I will do that in the actual setup.

Yep.

You are 100% right.

Looks like a switching power supply, yes. I thought that a cap placed across legs would help, but placing a 2 uF cap did not help.

Pretty close, about 14.75 or some such.

I see. Thanks.

I see. Well, I have another XR2206, I will try it, but I am a little doubtful that it would help. I have 3 chips total, and two already show exactly the same pattern. I will post the exact schematic, as well as pictures, tonight. I appreciate your input greatly. I want this stupid problem solved before moving on to bigger things. This is a great learning experience. In computer programming, I know that people learn more from trying things, failing and diagnozing problems, than from other forms of learning, not that other forms of learning should be dispensed with.

I agree. I will do just that.