TPS40090: pwm signal discountinued

Hello,

Sorry but let me better understand. Do you mean you forced the COMP pin by external voltage 1.4V to 2V? And it is working in this case? The PWM signal are normal as expected?

Regards

Thanks for your replay！

Yes,FB & DROOP connect to REF directily。COMP forced by external voltage generated by Opamp circuit with limit between 0.5 V~ 2.9 V。

When didn't EN ucc27223 ，FB ＞1.4V PWM signal begin to generate, oscilloscope's channel 1 = PWM1 ,channel 2 = PWM2，The multimeter shows the voltage at FB point

when FB voltage is 1.8V ，PWM reached it's maximum duty cycle

when EN ucc27223 ,The PWM signal begins to become discontinuous

then i change channel 2 = SS，The results show that some kind of failsafe seems to have been triggered

I think there is something wrong with my CS1~4 circuits?

I used a 1.5 milliohm sampling resistor in conjunction with an instrumentation amplifier to obtain the current value for each phase. The sampling resistor is placed after the inductor. The REF terminal of the amplifier is connected to CSCN and set to 0.7V through a resistor divider circuit. The sampling gain is currently 1.5 mV/A and the ILIM voltage is currently set to 0.460 V.a 1kΩ resistor is added between instrumentation amplifier's output and CSX in。

I think i found the problem，the inductor didn't match my load。

1 uH ichoose，400 khz each phase，when VDCin=5V ，duty cycle at 50%。1 Ω load。

in this condition ，inductor current drop to negative ，like DCM mode。

It seems I need to use a lower load resistor and a higher current setting to ensure continuous inductor current.

iIn addition, I found that the ucc27223 of some channels will lock B and turn off the upper and lower MOS tubes after being PWM turned on repeatedly in this situation

Will this situation trigger the chip's fault protection measures?

Thank you in advance for your time.‌ 

Hello,

great you made some progress! I was off last week. Can you please make it clear what is the issue you are facing now?

Regards

delighted to receive your response!

i wonder if csx output negative because some phases' inductor current drop to zero and then become negative。Is that situation will trigger some protection？

Hello,

what happens exactly? Is it shutting down and need to recycle the power to turn it on again?

Regards

Hi，here is some wave

channel 1：Vin 

channel 2：TPS40090 PWM 1

channel 3：TPS40090 COMP

channel 4：TPS40090 SS

Rload=0.5 Ω ，Apply Vdcdc_in ≈3.10V ，it's work at maximum Duty ,PWM signal continued

When increase Vdcdc_in  at some point，PWM signal began to discontinued，The voltage on the SS terminal starts to charge repeatedly, it looks like some kind of fault protection is triggered。Because I have forced the FB terminal voltage = REF, the over-voltage and under-voltage protection will not be triggered. So the only possibility is that the overcurrent protection is triggered?

Then i check the CSCN voltage ，It seems that some problems have been discovered

channel 2：TPS40090 CSCN

channel 4：High-side gate voltage of PWM4

When shutdown Vdcdc_in【VDCDC_in is only connected to the upper-side input of each phase】 ，Only control voltage，CSCN = 0.486 V

When Vdcdc_in=3 V，CSCN change to  1.04 V

Then I checked my CSCN circuit. I used a voltage divider resistor to generate a reference voltage of 1.5 V, as shown in the figure below, but it seems that the CSCN pin of TPS40090 has a large sink current, causing the actual voltage to not reach 1.5 V and has fluctuations.
So CSCN needs to be connected to a low internal resistance voltage source? Is it possible to adjust the voltage dividing resistor to the hundred ohm level?

I don't know if this is the reason for PWM discontinuity.

Hello,

I'm not sure what is the impedance on CSCN pin. From your experiment, the 1.5V reference voltage, from the voltage divider, is changing means the CSCN has low impedance. I think changing the voltage divider to hundreds Ohm will make the 1.5 ref more stable.

Regards

Hi，thanks for your reply!

maybe a 1.5v AA battery is a good choice for stable cscn :-）

I'll try it and see if the problem solved