LM2901: LM2901PWR

Hi Shritesh,

200kHz corresponds to a period time of 5µs. The response time of LM2901 is 1µs. So, the LM2901 doesn't seem to be fast enough for you application? Hard to say without seeing the full schematic and having more details on the applciation.

Kai

Hi Shritesh,

As Kai said, you would really be pushing the limits of the LM2901 speed. It's never good to push the limits of comparator speed. Speed could potentially vary ±30% across lots.

The output would not look pretty..and would be somewhat non-symmetrical. I would use a faster comparator.

What are the input voltages and supply voltages? What is the output driving?

Thanks a lot Kai for your reply. My application is for CLLLC Converter. I have attached circuit and waveform in attachment. I want to turn ON switches Q5 and Q8 when current I2 crosses +5A and turn them OFF once current falls below +5A. Also, I want to turn ON switches Q6 and Q7 when current I2 falls to -5A and turn off these switches once current rises above -5A. Sensed current is also shown in waveforms, 5A corresponds to 1.59V and -5A corresponds to 1.4V.

Thanks a lot Paul for your suggestions. I will use faster comparators. For my application, Input voltage to OPAMP is quasi-sinusoidal signal as shown in waveform (I2_sensed) and power supply to OPAMP is +5V and GND. Kindly let me know if comparator circuit shown in application note SNOA999 will be suitable for my application. I want to detect +1.59V and +1.4V levels in sensed circuit. I want accuracy to be within 10%, meaning I should be able to turn ON switches Q5 and Q8 when current I2 is between +4.5A and +5.5A.

Hi Suritesh,

What is the ground reference? How is the +0.75 to +2.5V being created? How is the current being sensed?

If this is truly a single supply application (GND and +5V), and the input is swinging +0.75 to +2.25 around +1.5V above ground, then you could use pretty much any high speed (<200ns) comparator. Your input overdrives are pretty high (>100mV), so the comparator should be operating at it's fastest speed.

Do you need a single or dual? I would recommend a push-pull output for best output symmetry.

Since you are requiring some accuracy, I would recommend looking at the TLV9032, as it has 100ns prop delay, a max offset of 1.5mV, and also does notT have built-in hysteresis, so it will trigger right at the crossing. Your signal is fast enough, and large enough that I don't think you would need hysteresis.

Other possibilities would be the TLV3201/2 (40ns), TLV3501/2 (4.5ns), TLV3601 (2.5ns), though these high speed devices do have a mV or two of built-in hysteresis (your threshold would shift  ±1 to 2mV on the positive and negative slopes).

Just keep in mind that the faster the comparator, the faster the crossing is reacted to, and you have more time to turn on the switch (which will probably have nF gate capacitance which can add delays).

Hi Paul,

Thanks a lot for detailed explanation.

To answer to your question "How I am getting input swinging from +0.75V to +2.25V",
Current sensor output is varying between -0.75V to +0.75V (corresponding to -40A and +40A resp.).
Output of this current sensor is fed to OPAMP TLV4376 with level shift of +1.5V. So, output
from TLV4376 is varying between +0.75V to +2.25V (corresponding to -40A and +40A).
For current sensor supply and OPAMP supply, I am using same ground.

Dual package OPAMP is okay for me, so I have chosen precision comparator TLV9032 for analysis.
With TLV9032, I have quickly carried out simulation in TINA (attached simulation file)

Expected output from OPAMP U1 is +Ve pulse (+5V) for switches Q5 and Q8 when VG1 is more than 1.594 V (corresponding to +5A).

Expected output from OPAMP U2 is +ve pulse (+5V) for switches Q6 and Q7 when VG1 is less than 1.406 V (corresponding to -5A).

I have carried out all the calculation of resistor values keeping in mind that OPAMP is ideal (attached calculation doc).

But, due to non-idealities of Comparator, I am getting pulse for Q5 & Q8 when VG1 is between 1.829 V to 1.24 V while I am getting pulse for Q6 & Q7
when VG1 is between 1.169 V to 1.67 V. This is very huge error for my application.

Please suggest me, how do I tweak resistors values to get accurate output by considering OPAMP/Comparator non-idealities.

Thanks,

Shritesh

Level_Detection_circuit.docxTLV9032_Level_Detection_5A_-5A.TSC

Thanks a lot Paul for suggestions.