Boost convertor TPS40210

Sunil,

Could you send me your schematic schematic to look over to see why there might be some issue with operation?

Something does not seem correct as the output voltage should achieve well more than 15V if the FB pin is shorted to GND as this should force the controller into a maximum duty cycle condition.

I checked on the BAT 46W datasheet and this appears to be an under-rated device for this application.  The diode for this application should be a minimum 1A continuous 5A peak forward current diode.

Hi

Please find the attached auto generated schematic. The part no of diode and its discription is not matching.

The diode I used was MUR 460 and the MOSFET was also different IRF3205.

The shunt / current sense was 10milli Ohm and the inductor used was 1uH , 15 milli Ohm.

Hope with this info you would be able to figure out where I have gone wrong.

Thanks

Sunil,

I'm not sure where the issue is occuring, but a 1uH inductor is way too small for this application.  I am calculating more than 12A of ripple current in the inductor, which is tripping OCP each switching cycle.  The inductor should be a 10-15uH inductor to limit the ripple current to approximately 1A pk-pk.  The inductor will also need to support about 5x the peak load current.  With your current design (15ms soft-start time) your start-up current (current to charge 5x100uF output capacitance to 48V) will require 1.6A which would drive a peak current of 8A.

I would recommend looking for a > 8A thermal > 9A Saturation current inductor with a value between 10uH and 15u.

Hi Peter,

Thanks for the quick reply.

The inductor value was as suggested in the auto generated schematic. I had used online version of switchpro.

So What you are suggesting is i should go for 10 to 15uH inductor any suggestion for the diode  and what should be inductors resistance.

Thanks 

Sunil

The thermal spec (minimum thermal current of 8.5Arms) will drive a maximum DCR as this is the heating element of the inductor.

For a diode, I would recommend looking for a 2-3A average schokkey diode.

So MUR460 should be ok for the diode .

Inductor I dont understand.

Thanks 

So I should be using above 10uH inductor with low DCR ??.

But then The Switch pro gave different values for these.

Is there any other mistake in the schematic.

Ideally what should be the feedback value to be expected at the FB pin.

Thanks & Regards,

Sunil

Yes, I would recommend an inductor larger than 10uH with a low DCR capable of supporting more than 9A without satuation and 8.5Arms at an acceptable thermal rise.

The voltage at the FB pin should be a constant 0.7V.  This pin is the inverting input to an amplifier.  The non-inverting input is an internal 700mV reference voltage and the output is the COMP pin.  I have not had the time to calculate recommended components for the Rs and Cs around the feedback pin, if you would like to check them, I would recommend using the equations in the datasheet.  Switcher Pro should be using the these equations, but I am not certain why it is calculating a 1uH inductor.

I used Switcher Pro, and while I did end up with a 1uH inductor, I got a very different design than you got.

I forced the same 5x100uF output capacitor that you had.

Switcher Pro is selecting the BAT46W because this device is showing as a 10A diode in the switcher pro database.  Unfortunately this appears to be a case of GIGO (Garbage In / Garbage Out) where the data we recieved from the manufacturer (Diodes Inc) contained an error.  I will see if it can be corrected.

I've attached my switcher pro results for your design.

Hi  Peter ,

I will change the inductor and give the feed back,

Mean while can you confirm

1.if 480KHz  measured at RC pin (1)  is okey

2. Voltage at BP pin (9) is 8.13 V

3. Can PWRPAD  be open.

When I ground the feed back what should the waveform at Gate drive be . will it be 480 Khz with 50 % or more duty cycle.

I am quite new to all these. Sorry for all the question.

Thanks & Regards,

Sunil

Hi,

After changing the inductor to 10uH also, no change, voltage boost to only 15V on shorting the feedback to ground.

Thanks,

Sunil

please e-mail me a description of your test set-up and any waveforms you are capaturing.

At the least, I will need:

The switching node (Drain of the MOSFET)

The Input Voltage (VDD pin voltage)

The Soft-Start voltage (SS)

The Output Voltage (Vout)

Anything else you have would also be appreciated.

MUR460 is not  designed  for low voltage boost use, it is ultrafast 4A 600V diode (the MUR is abbreviation from Motorola (r) Ultrafast Rectifier )

(most ONS Schottkys have amperes and volts in their typenumbers, but for ultrafasts its amperes and dekavolts(tens of volts and typenumbers begin with MBR for Moto Barrier Rectifiers or NBR for oNs Barrier Rectifier) they sure have also other typenumbers, but these are of either Moto or Ons original types)

it is far better suited for Power factor corrector switching to 400V..500V or flyback converters with 100-300V output.

far better diodes from the same manufacturer are:

MBRS360T3G (3A 60V SMC)

MBRD660 (6A 60V Dpak)

SS26T3 (2A 60V SMB) might be quite warm in your application

the only suitable ONS 4A diode is MBRS4201 (4A 200V SMC) but it is overkill for 48V boost and often ideal for 48V transformer isolated converters secondary.

or if you like leaded axial devices, the older MBR360G is also viable candidate.

Hi,

But can this wrong choice of diode fail the system. what I find is the gate drive generated at around 10Khz is restarted at 10Hz frequency and so the system build only 15 volts that too when the FB pin is shorted to ground.

Thanks

Sunil

thanks

Can open PWRPAD cause any problem other than heating .

A floating Power Pad / Thermal tab can introduce noise to the IC.

Excessive power dissipation can result in tripping over-temperature protection.

If you are seeing switching at 10kHz, what is the frequency of the ramp voltage on the RC pin?

The frequency at RC pin (1) is found to be about 480Khz.

What I saw at GDRV (pin 8) is that the frequency shown by oscilloscope is alternating between 10kHZ & 10Hz.

Thanks

frequency at RC pin and frequency at GDRV pin should be the same unless the control loop is operating in error and the control voltage (COMP) is non-constant.

Per my conversation request, can you e-mail me directly?