SN74LVC1G123: Use SN74LVC1G123 to generate constant On-time pulse with external triggered square wave signal

Yi Huang

Part Number: SN74LVC1G123
Other Parts Discussed in Thread: SN74123

Hi All, I am trying to use SN74LVC1G123 as a constant On-time generator to convert an input square wave signal (~50% duty cycle) to the pulse signal with the same frequency and Constant On-time (programmed by R_EXT and C_EXT). My input signal comes from a logic inverter, and the steady state frequency can be adjusted between 1MHz to 5MHz. I used to use SN74123 in a similar project and got a pretty constant On-time by using proper R_EXT and C_EXT.

However, right now I am using SN74LVC1G123 to save some PCB space, and I found that the On-time is constant (~120ns) when the input signal's frequency is above 2.5MHz. When the input frequency is between 1MHz to 2.5MHz, the On-time has a variation, and drops from 140ns at 1MHz to 120ns at 2.5MHz. The external components I am using is REXT=5.76K, and CEXT=10pF, VCC is a regulated 5.2V. Both R and C are

I do not know why the On-time can have variation in the lower frequency range. Since my target is to achieve a relatively constant On-time in the entire frequency range, can someone tell me if there is a quick fix by using some external components? Or if I have a use a similar product, can you suggest one with the same package and pinout?

Thank you!

over 7 years ago

0 ShreyasRao over 7 years ago

TI__Guru 57390 points

Hi Yi Huang ,

We will look at this and respond back on Monday.

0 Yi Huang over 7 years ago in reply to ShreyasRao

Prodigy 210 points

Hi Shreyas, do you have a solution now?

0 ShreyasRao over 7 years ago in reply to Yi Huang

TI__Guru 57390 points

Hi Yi huang ,

I believe you will have to account for the passive components variation across process,temp,voltage corners. The timing is dependent on the passives more than the device itself. I see a spec on the ds about the variation and for lower cap values ,the variation increases. The k value is not specified in the graph for 10pf range.

0 Yi Huang over 7 years ago in reply to ShreyasRao

Prodigy 210 points

Hi Shreyas,

If I understand you correctly, if I can increase C value, decrease R value, and make sure their values are accurate, I suppose to have constant On-time in each cycle in my frequency range. Is this correct?

0 ShreyasRao over 7 years ago in reply to Yi Huang

TI__Guru 57390 points

Hi Yi ,

I would not expect it to be perfectly constant since the passive will have tolerance, however, it will be minimize the inaccuracies.

Please let me know what you find out.

0 Yi Huang over 7 years ago in reply to ShreyasRao

Prodigy 210 points

Hi Shreyas, would you tell me do you have any other monostable multivibrator that has the same package as SN74LVC1G123 but without schmitt trigger at its "B" input? For example, I want some parts like SN74123 (a monostable multivibrator without schmitt trigger) but has the same package as SN74LVC1G123 (8-VFSOP).

0 Emrys Maier over 7 years ago in reply to Yi Huang

TI__Guru* 96681 points

Hi Yi,

Why would you want to eliminate the Schmitt-trigger input? A normal CMOS input will have far more strict input requirements without the benefits of the ST.

From the system standpoint, I can't see a benefit to eliminating a ST input.

Logic

Logic forum

SN74LVC1G123: Use SN74LVC1G123 to generate constant On-time pulse with external triggered square wave signal