This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPS40210 not operating when added an additional capacitor for backup

Other Parts Discussed in Thread: TPS40210, TPS54160

I am facing a serious issue in my TPS40210DGQR Boost circuit. Please find the attached schematic.

45V BOOST - 18V BUCK CKT.pdf

In the Schematic the TPS40210DGQR (U20) Boost converts the battery voltage 12V to 45V. The 45V is used to 2 purpose. The 1st purpose is close back a big magnetic solenoid and to fire a small solenoid. In 2nd purpose this 45V is buck converted to 18V using Buck Converter IC TPS54160DGQ (U6). The 18V is used to generate 5V using TPS54226PWP DC/DC Converter (not included in attached schematic) used for digital circuit with maximum load of 300mA. Also this 18V is used to generate 3.3V using TPS715A33DRVR LDO (not included in attached schematic) for powering the Microcontroller (MSP430F5438A).

 

The circuit operation is that when microcontroller turns on the CHARGE_45V signal the battery voltage is connected to the Boost Converter. The Boost Converter should boost convert the battery voltage to 45V and charge the C36 (2200uF) Capacitor. Also this 45V is connected to Buck Converter which converts this voltage to 18V and then 18V is further converter to a 5V and 3.3V. The Micro controller will switch off battery power to Boost converter when the C36 capacitor is charged to 45V and then the MCU controls Open and close solenoid. The Boost regulator does not need to operate solenoid it only require to charge the capacitor to 45V. 

 

The entire circuit is operating fine when only 2 capacitors (C33 & C35) are connected in 18V rail. For getting more backup time I have added an additional capacitor (C90) in the 18V (marked in the attached schematic) to preserve battery power so that micro controller in Low power mode can operate with capacitor power itself for at least 30 Sec. When the 3rd capacitor (C90) is added for some operation circuits works fine and after some time it stops working (the capacitor is getting charged only upto 20V). When I removed the 3rd Capacitor it works fine again.

 

I have checked the gate waveform of the Boost regulator with 2 capacitors and 3 capacitors. In both cases the mosfet operates in ~ 90% and drain waveform operates in ~10%. Wave form attached as Jpeg. 

  

Fig 1: Channel 3: Mosfet Drain Waveform Channel 4: Mosfet Gate Waveform.

Fig 2: Channel 2: Mosfet Gate Waveform.

 As per my understanding the dutycycle should be less than 75%. Maybe in 2 cases the boost regulator in operating not correctly due to miscalculated component values. Please help me to validate the Boost and buck circuit component selection and Please help me to solve this issue.. 

 

Also attaching 105V Boost circuit used to Open the solenoid. Please review the circuit components too.

105V Boost Circuit.pdf

  • Please anyone reply.. Its very urgent..
  • It is the weekend.  I'll get someone to look at this on Monday.

  • Just to verify, your actual problem is with the TPS54160 converter rather than the TPS40210, correct?
  • Hi John,

    Thanks for your response.
    No the actual problem is in TPS40210 only because the output is not getting boosted above 20V after some operations.
    I already have a huge output capacitor on the TPS40210 (2200uF) for backup and 470uF * 2 on output TPS54160 which is in series.
    With the above output capacitors current design is working. But for additional backup when I have added one more capacitor on TPS54160 output (470uF * 3) the TPS40210 itself is not operating. It is boosting the battery voltage only upto 20V while I require 45V.

    Thanks..

    Yasin
  • adding an additional 470 uF output capacitor requires an additional 2.64 A of output current, so with 0.01uF SS cap, it requires 7.93 A of output current from the TPS54160 just to charge up the thee output caps during slow start time. You are undoubtedly reaching current limit with the TPS54160 and probably with the TPS40210 as well. Can you check it?
  • Hi John/All,

    Sorry for writing this long question. But it is so critical project so I'm troubling u all.

    I'll check the SS pin waveform of the 2 converters wrt output and will update. But the TPS40210 output itself is not getting boosted above 20V.

    From the datasheet using Css equation, the tss is calculated as 3.2ms. For one 470uf cap the charging current required is Iccharg = ((470uF)*18V)/3.2ms = 2.64A. So for charging 3 capacitors the current required is Iccharg = ((470uF*3)*18V)/3.2ms = 7.93A. I understand this. But I don't understand how this is related to current limit. How TPS54160 is sensing the current and limiting.

    What would be safe value for Css that will allow slower charging of the SSpin cap so the output also charges with lower current. I need to charge this cap within 100ms so for safer side if I select 25ms and calculate my Css cap as 78nF (Select 82nF) then can I solve this problem? Then the Iccharg current will be ~1A. Is there any recommended Css capacitor value range.

    I have some doubts with inductor calculation of the both the converters

    1. In TPS54160 Datasheet I found the equation in page no. 29 Lomin = ((Vinmax - Vout)/(Iout * Kind)) * (Vout/(Vinmax * fsw)) = 12uH. While in the circuit the inductor value is 330uH. I have considered the output current as 7.93A with 10nF Css cap charging at 3.2ms. When I tried to calculate back the frequency required for inductor used I got 20Khz. So there is mismatch between the inductor selected and the frequency selected.

    Thinking on an other way I substituted the Inductor value (330uH) and frequency used (562Khz set by Rt = 205k) on the above equation and calculated the output current I got 0.29A = ((45.5 - 18)/(330uH*0.2)) * (18/(45.5*562Khz)). If I substitute this output current on the tss equation tss = (Cout * Vout) / Icchg = ((470uF*3)*18)/0.29 = 88ms. The required tss time to charge the 3 capacitors is 88ms. Is my understanding correct? But the Css 10nF will get charged in 3.2ms but at time the output capacitors will be charged only upto 0.65V calculated using the above equation Vout = (3.2ms * 0.29A) / (470uF * 3) = 0.65V. Is this correct? So Vsense is not following the Vss with a 45mV offset. So is this the reason of issue? But then why is it working for 2 Capacitors which require 58ms to charge?

    2. In TPS40210 Boost regulator when considering the output capacitors to determine the output capacitors whether the buck regulator output capacitor charging current also need to be considered. if we change the Css of the Buck regulator to 82nF to reduce the output current requirement to 1A then the input current requirement will be ~500mA. So to calculate the maximum output current of the TPS40210 whether we need to be consider all the capacitors like (0.1uF(C72)+10uF(C73)+100uF(C31)+2200uF(C36)+470uF*3(C33,C35,C90)) or we need to consider only the current required to charge (0.1uF(C72)+10uF(C73)+100uF(C31)+2200uF(C36)) and then add the buck converter input current requirement of ~500mA.

    3. I tried to validate the boost design considering the output capacitor as 3720.1uF. I calculated the Icchg with tss of 50 ms set using Css capacitor 1uF using formula Icchg = ((Cout*Vout)/tss) = ((3720.1uF*45)/50mS) = 3.35A. Is it ok. Whether it will create overcurrent limit. How to know when the overcurrent limit will be enabled. Whether I need to change the tss to 100ms so that the Icchg will be 1.674A.

    4. I calculated Ioutcrit using equation no. 4 and calculated Ioutcrit = 11mA with current inductor 330uH and frequency selected 298KHz (Rrc = 549k, Crc = 100pF). So I consider the Continuous conduction mode equation to calculate the Dmin and Dmax. The Dmin = ((Vout - Vinmax + Vd)) / (Vout+Vd) = (45-14+0.5)/(45+0.5) = 69.23% and in same way Dmax = 82.4%. Is it correct? I used this Dmin and Dmax to calculate the Iripplemax considering the peak to peak ripple as 30% and Ioutmax as 2A(tss=100ms, Css = 2uF(is it ok?)) and calculated Iripplemax = 1.95A using equation no. 34 from datasheet.

    5. On the Design example the ILrippple is selected as 30%. Is it ideal for my design too. Can I select the same? The equation no. 35 on TPS40210 datasheet to calculate Lmin = (Vinmax/Iripplemax) * Dmin * (1/fsw) = (14/1.95)*0.6923*(1/298KhZ) = 16uH. In circuit I have used 330uH. If calculate back the frequency required for the Inductor value used I got fsw = (14*0.6923)/(1.95*330uH) = 15Khz. So in boost circuit also there is a mismatch of Inductor and frequency. If I recalculate Iripple using the used inductor value and used frequency I got Iripple = 99mA. If calculate back the Iout max using the Iripple I got Ioutmax = 0.102A. using this Ioutmax when I calculated back the tss required for charging is tss = (Cout*Vout)/0.102A = 1.64s. But we have designed tss for 50mS but at 50mS the Vout will be at 1.37V (Vout = ((50mS*102mA) / 3720.1uF). Whether my understanding is correct? Whether that is a possible cause for the issue?

    6. Using the Iripplemax (1.95A) and Iout max 2A I calculated the Inductor current rating required using equation no. 38 and I got ILrms = sqrt(((Ioutmax/(1-Dmax))2+((1/2)*Iripplevinmin)2)) = 11.38Aand calculated ILpeak as 11.5A with equan no. 39. Is this correct? I am using Inductor with 2.1A saturation current. Whether this also is causing problem. I need to find a inductor with 12A rating, I hope. Or whether I can reduce the Inductor ripple current and select a lower current rated inductor. What is recommendation to select the Inductor ripple current?

    7. Using equan no. 45 when I tried to calculate the output capacitor the I got Cout value as 88uF. Cout = 8*(((Iout*Dmax)/Voutripple)*(1/fsw)) = 8*(((2A*0.824)/500mV)*(1/298KHz)) = 88uF. If calculate back the frequency required back by providing the Cout I got fsw = 7Khz. Which is not in the recommended range of the controller. Is it ok? How it can be solved?

    8. From the equan no. 7 the RISNS can be calculated as RISNS = VISNS/((IOUT/(1-D))+(Iripple/2)) = 150mV/((2A/(1-0.736))+(1.75/2)) = 17.7mR. But If I use the next part in the same equan i.e; RISNS = VISNS/((IOUT/(1-D))+((D*Vin)/(2*fsw*L))) = 19.68mR. If I use the eq. no. 10 RISNSmax = (Vvdd*L*fsw)/(60*(Vout+Vd-Vin) = ( 12*330uH*298Khz)/(60*(45+0.5-12)) = 0.587R. if I use the eq. no. 49
    Risns < Visns(oc)min/(1.1*(IL(peak)+Idrive)) < 9.09mR and if i use eq. no. 30 I got Risns < (Vinmax*L*fsw)/(60*(Vout+Vfd-Vinmax)) < (14*330uH*298KhZ)/(60*(45+0.5-14) = 0.728R. Which is correct? What value I should select for Risns? It is used as 10mR in the design currently. Is it ok?

    This design is created earlier by a engineer who left long back. In between on the revison history the it is seen that the Inductor value is changed from 47uH to 330uH but no proper reason is given. I only added the additional capacitor for backup. Whether I need to do a complete redesign? The capacitor cant be changed as I require the same for backup. Would you please suggest me a design for the same?

    Thank you. Its very critical project application so I want to clear all my doubts with experts.

    Regards,

    Yasin.
  • Hi All,

    Please reply.. Anyone please help me..

    Thanks in advance..

    Yasin..
  • Where do you see reference to 330uH inductor? The datasheet example uses 10 uH.
  • If you do not have access to the original design notes, I would be inclined to start over and recalculate everything. I see the 330uH is in your schematic not the datasheet. It seems unreasonably high. 30% ripple current is the rule of thumb. You should be able to use it. You will need to account for your total capacitive load for your start up condition. If you post your completed new design equations and new schematic, we can review it. Also, I would be inclined to go ahead and check the deign using the pspice models available on the web.
  • Hi John,

    I was working on it and created an excel sheet with all the design calculations and values.

    Please review the sheet for 105V Boost , 45V Boost and 18V Buck converter.

    Also the inductor was changed during an initial design change and I am not aware about its reason.

    Please help me validate this design.

    New Design Suggestion.xlsx

  • Hi John / All,

    Awaiting your valuable reply.

    Kindly please review my new design suggestion excel sheet.

    Thanks,

    Yasin

  • Hi John/ All,

    Please validate my design sheet. Whether my calculation and new values selected are correct.
    Please help. I have ordered the components and will check myself also.
    But if I get a expert confirmation I'll be more confident to present the change to my customer.

    Thanks.

    Yasin.