4

I'm using common voltage controlled oscillator IC - AS3340. It has some problem with pulse wave output on higher frequencies. The bottom of falling edge looses sharpness the higher I go.

This is how it looks at the limits of audible threshold:

enter image description here

I need some advice on cleaning up this signal so I would get nice and crisp pulse shape all the way.

This is my schematic. I'm measuring it at the PULSE_OUT node, without attaching the amplifier circuit.

enter image description here

somerandomusername
  • 97
  • 9
  • Show the circuit that generates this signal. Include links to the datasheet of the IC you're using. Show how you measure it (include details). What are "higher frequencies" 10 Hz, 10 kHz 10 MHz ? I'm too lazy to find the "18.3 kHz" i nthe corner. Don't ask "How to fix this" without understanding **WHY** the signal is like that. – Bimpelrekkie Oct 16 '20 at 13:17
  • If we don't see what circuit generates this signal, how are we supposed to suggest ways to fix it? – Justme Oct 16 '20 at 13:19
  • Sorry, I added the circuit. There really isn't much to it. I'm measuring just the pulse output with just the pulldown resistor attached As far as I understand professional audio modules reshape the pulse wave from this IC as it is generally pretty bad. – somerandomusername Oct 16 '20 at 13:22
  • A Schmitt trigger circuit? – winny Oct 16 '20 at 13:23
  • 1
    *professional audio modules reshape the pulse wave* OK, crucial info. there! (it should have been in the question, not in your comment). So the IC's pulse output waveform is terrible, not much can be done about that. Then we know **WHY** and have a good reason to look at how to "patch it up". I agree with @winny that a schmitt trigger circuit should be neatest solution. – Bimpelrekkie Oct 16 '20 at 13:31
  • I would use a logic gate schmitt trigger IC like the HEF40106 to patch up the signal. Use the 12 V supply for the HEF40106 and you'll get a 12 Vpp square wave at its output (use resistor divider to attenuate if needed). That should be a proper square wave up to at least 100 kHz. – Bimpelrekkie Oct 16 '20 at 13:31
  • Sorry yes I could have made it clearer. Thats why I didn't add the design at first because my question was how to reshape the wave, not to fix it. – somerandomusername Oct 16 '20 at 13:41
  • Reducing the pulldown resistor will clean it up. Whether it'll clean in up enough is up to you, but you won't find a simpler way if the chip can source a higher current : check that in the datasheet. –  Oct 16 '20 at 14:09
  • Yes I found that out thru some experimentation. Data sheet said to use 51K. Original CEM3340 IC (that this chip is based of) used to have 10K. In theory those chips should be interchangeable. It can handle higher current for sure, but the less heat it has the more stability in the oscillator. – somerandomusername Oct 16 '20 at 14:12

1 Answers1

9

As others have said, a Schmitt trigger will clean it up. Reducing the 51k resistor to 10k or so may help as well.

Also consider this waveform distortion is irrelevant for the typical intended electronic music function of the IC. A perfect 18kHz square wave is barely audible under good conditions. The harmonics of the imperfect square are way above audible frequency. The difference between a perfect square and your signal will not be perceptible in the context of electronic music. In an analog synthesizer, achieving perfect waveshape beyond useful audio frequencies isn't necessarily expected, required, practicable, or audible.

vofa
  • 1,739
  • 7
  • 12