| Electronics > Beginners |
| Sine to square wave circuit instability question |
| (1/2) > >> |
| Chris Wilson:
I built the circuit below to tun a 0dBm sine wave at around 275kHz into a square wave to drive the input of a 74F74 IC. Initially I built on a breadboard and with no input signal i saw a lot of noise, maybe harmonics of our 50Hz mains, on the output. The square wave output with the circuit on a clean 5V supply looked good. i did a proper build on double sided board with a full top ground plane and an SMA input socket and a BNC output socket in an alloy screened box. I put in a 5V regulator with a 4.7uF Tantalum on the input pin and a 2.2uF electrolytic on the output so it could be run off a 12V supply. With NOTHING connected to the BNC input socket the output sits high and dead stable. but as soon as I connect even a BNC packaged 50 Ohm dummy load to the BNC input socket the output goes noisy with seemingly a load of harmonics that are multiples of 50hz. The 275 Hz source (a Kenwood TS-590 transceiver) does need even need to be powered up for the same noise to occur. Just having a decent terminated co-ax cable connected to the input starts the instability off. I tried a 50 Ohm resistor to ground on the middle pin of the input BNC with no change. When fed a sine wave from the 590 all is fine with a clean and nice looking square wave output Can I stop this happening and what do i need to try please? Thanks! |
| awallin:
self-oscillation is usually casued by positive feedback... you could also just simply bias the input to somewhere mid-way between Vcc and GND like here: https://www.george-smart.co.uk/radio/472_khz/472_khz_transmitter/ note no feedback resistor in that circuit. |
| Benta:
I'd simply use a CD40106 instead and bias the input to 1/2 VDD. It doesn't need to be precise (for 50% DC), the 74F74 is edge-triggered anyway. |
| ArthurDent:
What Benta posted above. If you put a 50 ohm termination on the input you are effectively grounding the input end of the capacitor making the circuit shown on the left side of the attached drawing I copied on line. I don't think the hysteresis matters. |
| basinstreetdesign:
--- Quote from: Chris Wilson on September 14, 2018, 12:47:56 pm ---I built the circuit below to tun a 0dBm sine wave at around 275kHz into a square wave to drive the input of a 74F74 IC. Initially I built on a breadboard and with no input signal i saw a lot of noise, maybe harmonics of our 50Hz mains, on the output. The square wave output with the circuit on a clean 5V supply looked good. i did a proper build on double sided board with a full top ground plane and an SMA input socket and a BNC output socket in an alloy screened box. I put in a 5V regulator with a 4.7uF Tantalum on the input pin and a 2.2uF electrolytic on the output so it could be run off a 12V supply. With NOTHING connected to the BNC input socket the output sits high and dead stable. but as soon as I connect even a BNC packaged 50 Ohm dummy load to the BNC input socket the output goes noisy with seemingly a load of harmonics that are multiples of 50hz. The 275 Hz source (a Kenwood TS-590 transceiver) does need even need to be powered up for the same noise to occur. Just having a decent terminated co-ax cable connected to the input starts the instability off. I tried a 50 Ohm resistor to ground on the middle pin of the input BNC with no change. When fed a sine wave from the 590 all is fine with a clean and nice looking square wave output Can I stop this happening and what do i need to try please? Thanks! --- End quote --- When you put a 1M0 resistor from the output of a 4000-series gate to its input you bias the gate into the middle of its linear region with about 60 dB of gain. It will happily amplify any noise it picks up at the input if the source impedance is low (like 50 Ohms). So putting a 50 \$\Omega\$ dummy load on the input or any 50 \$\Omega\$ terminated coax will do this. I think you need to increase the impedance from pin 1 of the gate to the input cap. Try a 1K - 10K res. This will lower the gain of the gate. Then with a 50 \$\Omega\$ source it should be more stable when there is no input signal but still work to square up the sine wave when its there. Hope this helps. |
| Navigation |
| Message Index |
| Next page |