LM2903 power supply ramp rate vs output

Nick Dai

Other Parts Discussed in Thread: LM2903

Hi Team,

Could you help check below question from the customer about LM2903D?

When VCC power up from 0V to 12V with 0.5V/ms ramp rate in below circuit(Rp=150k,Rn=8.4k),there is a pulse at the output during power up.

If change the Rp=15kohm and Rn=48kohm,then there is no pulse during power up with same ramp rate as below.

When Rp=150k,Rn=8.4k,there will be a pulse during power up if the ramp rate slow than 1V/ms. There is no pulse when power up ramp rate>2V/ms.

Could you help check why there is the pulse during power up in different condition? And why the output pulse is relate with power up ramp rate and the resister value?

Best Regards,

Nick Dai

over 7 years ago

0 Ron Michallick over 7 years ago

TI__Guru*** 139956 points

Nick,

If there was no input current, then both inputs would be the same and the output state would be unknown. Some device would be output high, others low based on each devices VIO (input offset error).

The difference in resistance multiplied by input bias current will be an input voltage.
For the standard grade of LM2903, provide at least 7mV to the input to get a known output high or low.

0 Nick Dai over 7 years ago in reply to Ron Michallick

TI__Expert 4240 points

Hi Ron,

Thanks for your feedback.

Best Regards,
Nick Dai

0 Nick Dai over 7 years ago in reply to Ron Michallick

TI__Expert 4240 points

Hi Ron,

Could you help check why the power up ramp rate will impact the output of LM2903 as below test result?

CH1(blue)---2IN+ PIN

CH2(red)---2IN- PIN

CH3(green)---VCC PIN

CH4(purple)---2OUT PIN

When VCC ramp up with 1V/ms rate,there is a big spur at output during VCC power up.

When VCC ramp up with 2V/ms rate,the output is nearly flat during power up.

Best Regards,

Nick Dai

0 Ron Michallick over 7 years ago in reply to Nick Dai

TI__Guru*** 139956 points

NIck,

How many devices (samples) do this?

I think the input doesn't have even delta voltage to ensure high or low. So a device that has same VIO as the tiny input could act differently

0 Nick Dai over 7 years ago in reply to Ron Michallick

TI__Expert 4240 points

Hi Ron,

Thanks for your feedback.

The power supply ramp up fast on customer's old products,so there is no high output glitch issue.

On their new project,the power supply ramp up speed is slow,and they found nearly all devices have such high output glitch.

Could you help check is there any way can remove such high output glitch with slow power supply ramp up speed.

Best Regards,

Nick Dai

0 Ron Michallick over 7 years ago in reply to Nick Dai

TI__Guru*** 139956 points

Nick,

I can try to replicate in the lab., but first I need to see the pins 5 and 6 on a 10mV/division scale
They both look like zero to me. Do you have a partial schematic?

0 Nick Dai over 7 years ago in reply to Ron Michallick

TI__Expert 4240 points

Hi Ron,

Please check below figure for small signal.

Figure 1,2 and 3 are slow ramp up rate.

Figure 4,5 and 6 are fast ramp up rate.

The schematic is in below:

Best Regards,

Nick Dai

0 Ron Michallick over 7 years ago in reply to Nick Dai

TI__Guru*** 139956 points

Nick,

The issue is as the first post has shown. The input bias current coupled with the ground resistances creates an input voltage differential that cause the output high. If VCC ramps faster then the extra capacitance at pin 5 delays the input bias differential enough so that the Vref turns on first before pin 5 can charge up.