Trigger for a CMOS Flip Flop

You don't need a fast rise time. What you need is a clean, debounced signal. I would change the N.O. push button with a N.O./N.C. push button (a.k.a. form C contacts) and send the two signals to the input of a S/R latch that will then drive your single asynchronous counter.

Can't change the button (it drives other stuff too), and the cd40175 had a max clock rise/fall of 15 microseconds. Now maybe I could use a 4093 Schmitt trigger for debounce and to give a fast clock to the 40175.

http://www.labbookpages.co.uk/electronics/debounce.html

https://howtomechatronics.com/how-it-works/electrical-engineering/schmitt-trigger/

A Schmidt trigger, it has hysteresis, is the classic debounce method​.

https://pubweb.eng.utah.edu/~cs5780/debouncing.pdf​_

Why do you think the manual push button will have a slower rise time than 15 microseconds?

I tried just the button as the clock and the flip flop went nuts with the bouncing. Then tried a typical RC debounce circuit, but the flip flop can't take the slow rise/fall of that approach. So will now try debounce plus Schmitt as Rixter and rwilliams suggested. That link by rwilliams is very interesting.

Schmitt trigger is more or less the go-to standard to de bounce and clean up a noisy signal.Been in use since at least the 1970's. I think it will solve your problem.

You should use a RC lowpass followed by a schmitt trigger. The lowpass must debounce your pushbutton bounce to a lower level as the hysteresis difference. The bounce time of a pushbutton may be very long, in worst case the whole pushing time which is user dependent.