LM5170-Q1: Current controller saturation not reflected on MOSFET gate signals

Stratos Chatzinikolaou

Part Number: LM5170-Q1
Other Parts Discussed in Thread: LM5170

Hello,

I am using the LM5170 controller in my project to control a bi-directional boost converter. Since I have not a dedicated voltage source connected at the HV port I have connected a constant 14V to the VIN pin of the chip. Also these 14 V are connected directly to the pin of the ramp generator network resistor where the VINX pin should be connected according to the datasheet (the other pin of the resistor is connected to RAMP pin of the TI chip). The board operates fine up to a certain current reference, however above that the current controlled seems to saturate, which I have noticed by probing the COMP pins of the chip. The voltage of these pins i.e. the Vea error signal is ~ 3V higher than the peak of the generated ramp (so it's higher than the peak ramp current even when considering the 1V offset). However the duty cycle of the MOSFET gate signals is not at its maximum value (as calculated from eq, 16). My question is how is it possible to have the controller saturated but the gate signal not?Is there an internal limit to the gate driver duty cycle when the controller saturates? Let me know if you require more information of how I setup the TI chip.

Thanks,

Stratos

over 5 years ago

0 Garrett Roecker over 5 years ago

TI__Mastermind 21720 points

Hi Stratos,

Thanks for reaching out with your question and for considering the LM5170 for you application.

Can you please try connecting the LM5170 VIN pin back to HV terminal? Also the VINX pin should be connected to the RC of the RAMP pin as shown in the datasheet.The RAMP pin RC should always be connected to the VINX so that the pins do not exceed the absolute maximum voltage rating.

To operate with a supply connected to the HV terminal please short out the breaker MOSFETs (Q5, Q6, Q8, Q9) and populating J3. This will allow the LM5170 to start-up and operate as a pure boost.

Please let me know if you have any questions.

Thanks,

Garrett

0 Stratos Chatzinikolaou over 5 years ago in reply to Garrett Roecker

Prodigy 75 points

Hi Garrett,

thanks for taking the time to reply. I'll follow your advice and connect the VINX pin to the RC of the ramp (I can see why not doing that can result to an overvoltage by charging the RC cap). From your answer it seems that you assume I am using a TI developer board, in case this is not clear from my first post I am using a board that I made which is including the TI chip for controlling the converter currents. Starting up is not an issue as I have deactivated the protection function. Regarding my original question, is there a reason that the current controller would saturate but the duty cycle of the MOSFET gate signals will not reach their max (or min) value?

0 Stratos Chatzinikolaou over 5 years ago in reply to Garrett Roecker

Prodigy 75 points

Hi Garrett,

is there an update to my question? If it helps, I can provide you with a TINA model that illustrates the issue.

Thanks in advance,

Stratos

0 Garrett Roecker over 5 years ago in reply to Stratos Chatzinikolaou

TI__Mastermind 21720 points

Hi Stratos,

Sorry, I though that you were using the development board for the LM5170. Can you please post your schematic?

Thanks,

Garrett

0 Stratos Chatzinikolaou over 5 years ago in reply to Garrett Roecker

Prodigy 75 points

Hi Garrett, I've attached a tina model with a configuration based on my design. In this simulation the controller saturates with the duty cycle being ~60%. test_sim.TSC

0 Stratos Chatzinikolaou over 5 years ago in reply to Garrett Roecker

Prodigy 75 points

hi Garrett,

any thoughts on my issue after looking at the simulation? Would be nice to understand what is limiting the controller because if I'm not able to overcome this I will not be able to use the LM5170 chip in my design. Also would be happy to continue this conversation offline if you can point me to the someone who might be able to help debugging this.

0 Garrett Roecker over 5 years ago in reply to Stratos Chatzinikolaou

TI__Mastermind 21720 points

Hi Stratos,

I am looking at the simulation right now. I will send you a direct message to discuss this in more detail.

Thanks,

Garrett

0 Youhao Xi over 5 years ago in reply to Garrett Roecker

TI__Guru* 78400 points

As you guys are discussion through the PMS, can we close the thread here? You can always re-open this thread by following a few post.

Thanks,

Youhao

Power management

Power management forum

LM5170-Q1: Current controller saturation not reflected on MOSFET gate signals