the signal I need to generate it will be 100khz but its on time 1 micro second.Using a PLL to generate a 1MHz signal from your 100 kHz signal would let you generate the desired pulse accurately.
How accurate is your 100kHz clock, how accurate do you need the width of the 1μs pulse to be, and how much delay can you stand between the rising edge of the clock and the rising edge of the 1μs pulse?I have clk signal with frequency 100khz I need to generate another pulse at the positive edge of the clk for 1 micro second on and 9 micro second off ,can I use one shot timer or there is another idea more accurate.?
a capacitor + a resistor coupling to the 100KHz clock will do.I need to generate another pulse at the positive edge of the clk for 1 micro second on and 9 micro second off
100khz+-1% and the 1 us=-10% ,and I need 1us every clk and is not important how accurate from the rising edge.How accurate is your 100kHz clock, how accurate do you need the width of the 1μs pulse to be, and how much delay can you stand between the rising edge of the clock and the rising edge of the 1μs pulse?
1us+-10%Need to know the pulse width accuracy you need, otherwise we are just guessing.
A 555 one-shot timer would be an obvious choice except they don't work well down to a 1μs period.1us+-10%

A 555 one-shot timer would be an obvious choice except they don't work well down to a 1μs period.
Below is the LTspice simulation of a 74HC123 one-shot timer which should do what you want.
The 74HC is fast enough to give a reliable 1μs pulse.
In the actual circuit you may have to tweak the value of R1 to get an exact 1μs pulse width (you could change R1 to a 5kΩ pot in series with a 2.5kΩ resistor if you want it to be adjustable).
Note that the unused inputs on the 74HC123 should be tied to ground.
View attachment 118322