TPS40210 with current limiter

Hi Thierry,

Another option would be to do this with two stages instead of one. For example boost the 12V to 36V then the 36V to 90V. I do not know of any other alternative parts.

Can you share your switching frequency and schematic? I think 12V to 90V at 1A should be possible if the switching frequency is low enough.

Best Regards,
Anthony

Hi Antony,

Many thanks for your reply and sorry for my low level english.

Two stages will be to expensive for the futur production. I don't need a perfect Vout regulation but the power supply will have pulsed high current inductive load (up to 15A 500uS max at 300Hz max). 90V 1A is not a constant load but a maximum average load.

On screenshot you can see that the problem is in case of no constante load, the current rise to fast and thePWM go to minimum Toff. In this case, the coil doesn't have time to discharge and in few time go to saturation.

My switching frequency is 160KHz.. I send you my last schematic where I have try a current limiter sending Vshunt on Vfb. It is not perfect because quit difficult to adjust (and less experience on them). I have ordered TPS40210 samples for my prototype but now I think it will be better to use TPS40211 because to much power on Rshunt.

The frequency must be higher because I have to high current ripple but at this moment I have some thermal problem with the mosfet and I will change it for bigger one (D2PAK instead DPAK with less gate charge and faster switching time).

Regards

Thierry

Hi again Anthony,

The schématic doesn't work. The current is limited but not like I was espect. In fact, the system oscillate at lower frequency.

See the attached file.

Magenta: current (Vshunt)

blue: filtered current

yellow: Vfb

Green: Vdrain

You can see that my mosfet doen't switch with the minimum Toff. Also I don't understand why the PWM change so fast (to fast) and I don't know what I have to do for slow down the PWM change.

I will change the mosfet asap and try to put 500kHz frequency swicthing for have 92% max PWM.

Thierry

Hi Thierry,

I am not sure if the added circuitry to bring Vshunt to Vfb will do much. I am assuming you are referring to R8 as Rshunt and Vshunt is the voltage across it. In the screenshots you had attached was there a load on the output? Is the TPS40210 not able to start-up and regulate the output to 90V with no load?

In this design the lower reference voltage of the TPS40211 will not help. The Rshunt it refers to for the TPS40211 is when a sense resistor is connected to the FB pin regulating current through it and the the load instead of the voltage. The Rshunt in your design is the over current sense resistor and the sense resistor used for the PWM control circuitry. Both the TPS40211 and TPS40210 will have the same voltage here.

I expect a lower switching frequency will help more in a high duty cycle design. The longer off time with the lower switching frequency helps to make sure the FET can turn off. Also the FET with lower gate charge allowing faster switching time will help a lot in the high duty cycle design.

I recommend you try simulating the loop response with the TINA or PSPICE average model. The model can be found in the TPS40210 product folter. If it is unstable, this could cause the TPS40210 to go into hiccup current limit during start-up.

Best Regards,
Anthony

Hi Anthony,

I have try today with FDB86102 mosfet with 500KHz switching but it doesn't enougth. when I do pulsed load, the current still going up to overcurrent.

I think my pcb layout is also not good.

For answer to your questions. The TPS was able to start-up and regulate the 90V with constant load (120Ohm). The problem is when the 90V move. In this case the TPS try to return to fast to 90V and put duty cycle to high for the system.

The best result was when I had connect the current sense to the feedback but it was very dificult to adjust.

Tomorrow I try to put more switching frequency but at 90V step-up, I am afraid regard heating.

Best  regards

Thierry

Hi Thierry,

Thank you for explaining! I think I have a better understanding on the challenges in your design. It sounds like your load does not require the 90V output to be constant.

I suggest trying to increase your output capacitance to support the pulsed load. This way the duty cycle will not need to increase as much to support it. The output capacitors will do more of the work to supply the current. Like you mention when the output voltage dips the TPS40210 will try its best to increase it back to 90V regulation.

You may also want to try decreasing the value of R8 to increase the over-current trip point. Although this would increase the maximum peak current allowed in the inductor so you will need to be careful to not saturate it.

With your test setup, is it possible to slowly increase the peak current to find the limit of when you start tripping the hiccup over-current? This way we can find the limit of the design as is and get an idea how far off it is from your requirements. Alternatively you can also try the steady-state transient simulation model to see how it works. This can be found on the TPS40210 product page.

Best Regards,
Anthony

Hi Anthony,

Great, you have understand the target. The 90V ripple doesn't matter; could be more or less 5V.

Larger capacitor is an option but I have to put a lot of them and the pcb final will be to large for the customer enclosure (another challenge!).

The pulsed load is current up to 20A in inductive load (pulse up to 200us).

This night I had an idea. I will try to use the TPS as mosfet current limiter (shunt connected to FB)  with voltage stop using over current input with a fast start up.

I have some TPS40211 yet and this version could be better for this new design.

I will tell you the result.

Best regards

Thierry