CD74HC123M monostable not working (no output)

[Flyback]

Hi,
The attached is a synchronised PFC controller using UCC28070A.
The Monostable (CD74HC123M) is not working, ie, its not providing narrow pulses to the UCC28070A. Do you know why?

The UCC28070A now seems to have powered up OK, and the UCC28086D is correctly providing a square wave to the monostable.

At first power up, the UCC28070A Vcc pin was stuck at around the turn-on threshold of some 10.4V. ….And it had 0V on its VREF pin indicating it hadnt woken up. I then changed R7 to a signal diode (anode to CD74..) and then the UCC280780A powered up fine (nice 6V on its VREF pin). However, even though a Square wave (~25% duty cycle) is correctly being sent to the CD74, the square wave is not going back down to 0V at the CD74 side of R3…..it is being held up at around 4.2V. -So the waveform seen at the “1B” pin of the CD74HC123M is just going continuously from 5V to 4.2V.

This must mean that the “1B” pin of the CD74HC123M is broken, because its supposed to be an input, and shouldn’t be able to source voltage.
Do you agree that the CD74HC123M must have somehow gotten damaged, eg, ESD, or perhaps something to do with R7 when it was in the first place, a 100R resistor? ..Also wondering about differences in supply assertion time between the 12V and 5V supplies. UCC28086D and CD74HC123M are supplied by the same 5V rail. UCC28070A is supplied by a 12V rail. (5V rail actually gets derived from the 12V rail so 12V asserts first)

CD74HC123M datasheet:

https://www.ti.com/lit/ds/symlink/cd74hc123.pdf?HQS=dis-mous-null-mousermode-dsf-pf-null-wwe&ts=1659373327659&ref_url=https%253A%252F%252Far.mouser.com%252F

UCC28070A datasheet:

https://www.ti.com/lit/ds/symlink/ucc28070a.pdf?ts=1659390653180

 

[alec_t]

post deleted

 

[Tony Stewart]

Replace, U3 dual 1shot is damaged with latchup due to load voltage > Vdd+0.3 Replace R7 with a low pF signal diode such as 1N4148. Not sure if the output is ESD protected internally but 100 Ohms might trigger it with sufficient input capacitance from 12V to input and might trigger a latchup condition on U3. You need to ensure there is no back driving of voltage on output so you can ensure this and verify. by isolating track open with a cut and scope it.

 

[Flyback]

Thankyou Tony Stewart.......you are a genius!....superbly helpful. I just took off R7, and replaced the CD74HC123M, and it is now giving exactly the right output..........so the next thing is to connect it back to the UCC28070A.....as you say, the first time i powered up, a voltage must have coupled through the UCC28070A pin capacitance-to-VCC, and blew up the CD74.....so , yes, i will use a IN4148 in there (in place of R7).....must admit i would still like some resistance aswell.....but the track goes on an inner layer after R7, so not so easy to add an R.......maybe i can do a "wigwam" type thing with a 1N4148 and a 1k resistor in series between the vias either side of the R7 pads.....
Thanks again.

Must admit, i woudl also now like a signal diode on the CD74 output pin, (just in case it rings below ground) but i'll struggle for room.

 

[Tony Stewart]

Ringing is a phenomena when the driver impedance is much lower than the long inductive track ( or probe ground) (higher impedance ) and a small pF load. Since this is prevented with the diode in series conducting only when the driver is high with R pull down. for future reference. e.g. 25 Ohm 3.3V technology driving 200 ohm track with rise time faster than 1/4 of the delay time, then ringing is possible and can be reduced/prevented by adding around 100 Ohms or reducing track impedance to match closer to source CMOS. Vol/Iol

 

[Flyback]

Hi,
Just put 1k and a 1n4148 instead of R7, (anode to CD74), and it now doesnt work......or should i say, when i probe at the mid point between 1k and 1n4148, it pulses for a second or two after power-up, then just goes flat at ~5V.

So it now goes CD74.......1K......1N4148.......UCC28070A

So i am wondering about this now.....maybe i just need to change R7 for 10k?

*********************still not working*********************
However, with 10k instead of R7, the CD74 doesnt work at all.......just get a flat 5v5 output from it. ...So in fact now i am wondering if the UCC28070A is damaged.
Any ideas why the CD74HC123M isnt working?

 

[Tony Stewart]

I should have read the datasheet for input impedance and how it works and I was thinking about recent sailing instead. THE RDM resistance is more complicated than I thought RDM resistance 30k min 330 kΩ max and Rdm= is defined under 7.3.2, & 7.3.3

 

[Flyback]

Now disconnected R7 from ucc28070a RDM pin, and it (CD74HC123M) still doesnt work.....though every now and again, it does work, for about a second or two after power up....then the CD74 output goes flat at 5V.
********************************************
Now just changed the CD74HC123M timing capacitor to fresh one, same value....now it works but only when not connected to the UCC8070A.