# 100MHz Schmitt Oscillator

Discussion in 'General Electronics Discussion' started by Hoobbyy_Engineer, May 21, 2017.

1. ### Hoobbyy_Engineer

13
0
May 21, 2017

#### Attached Files:

• ###### 1846108_4.pdf
File size:
261.4 KB
Views:
49
Last edited by a moderator: May 21, 2017
2. ### Audioguru

3,481
740
Sep 24, 2016
8mHz is a very low frequency. 8MHz is a radio frequency. "m" is thousandth, "M" is Millions. A huge difference.

3. ### Alec_t

3,055
833
Jul 7, 2015
I doubt it. Have you allowed for capacitor charge time and the pulse edge rise and fall times?

4. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,501
2,841
Jan 21, 2010
I would suggest that it is possible.

The propagation delay quoted is for an output capacitance significantly greater than the input capacitance of the gate.

Not so long ago I managed to get 84MHz out of a much lower spec device.

Having said that, there is a difference between getting it to oscillate at this frequency and being able to control the frequency. The capacitance of both the input and the load will contribute to the frequency, different gates within the same package may vary somewhat, and your construction method will have a great input (I'd use a dead bug method).

Last edited: May 21, 2017
5. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,501
2,841
Jan 21, 2010
Oh, another thing you need to remember is that propagation delay is for each transition. You will have one delay for high to low and another for low to high.

However, you might find this of interest.

6. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,501
2,841
Jan 21, 2010
At very high frequencies the output is current limited and the edge case has the output swinging between the input hysteresis values. The output also looks a lot like a sine wave. Depending on whether the rise/fall times or the propagation delay dominates, the actual output will tend to look more triangular or more square.

7. ### Hoobbyy_Engineer

13
0
May 21, 2017
Thanks for all the replies. Looks like it is not that easy to make the oscillator.
So to avoid complications I think it would be better to choose LTC6905 oscillator

File size:
170.7 KB
Views:
61
8. ### (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,501
2,841
Jan 21, 2010
Yep. It's not particularly cheap, but it should deliver a stable and relatively accurate frequency.

Hoobbyy_Engineer likes this.