TLC2272: The risk of adding the 1nf Capacitor in the V+ and V-

Zhihong Huang

Part Number: TLC2272
Other Parts Discussed in Thread: TLV2772

Hi team,

My customer now use the TLC2272 to do the current-sensing from the compressor of the Air conditioner and the frequency of the current detected in EM_SENSE is several tens of KHz.

1. he found that there is the noise from the other part of the system which cause the 400mV(p-p),10k output noise in the EM_IH pin and effect output signal of the TLV2272, and it is not allowed to do something to the source of the output noise because of the timeline of the project.

2.So he put the input capacitor C165 between the V+ and V- to decrease the noise. After he added the C165, he found that the output noise in the EM_IH pin is decreased to near 50mV which doesn't effect the output signal and cause the unstable system.

3. Now to keep the noise near the 50mV, the customer has to add the 1nF capacitor(C165) in the V+ and V-, so he want to know if he use this circuit to detect the current of other frequency(approximately 45Hz ~ 100KHz), does this capacitor cause other risks？

4.Could you please help me evaluate the risk or give me some advice about what kind of experiment that the customer can do to identify the risk?

This case is quite emergent and I hope that you could give the advice as soon as possible.

Best Regards

Wesley Huang

wesley-huang@ti.com

over 4 years ago

0 kai klaas69 over 4 years ago

Guru 140200 points

Hi Zhihong,

TLC2272 or TLV2272 or TLV2772?

Kai

0 Zhihong Huang over 4 years ago in reply to kai klaas69

TI__Intellectual 1550 points

Hi Kai,

TLC2272, the label in the schematic is wrong.

Best Regards

Wesley Huang

wesley-huang@ti.com

0 Michael Steffes over 4 years ago in reply to Zhihong Huang

Guru 47485 points

Morning Wesley,

The units for the RC are unclear, can you send this TINA file back with units corrected -

TLC2272 test ckt.TSC

0 Zhihong Huang over 4 years ago in reply to Michael Steffes

TI__Intellectual 1550 points

Hi Micheal,

Thanks for your reply, here is my TINA file.

Best Regards

Wesley Huang

wesley-huang@ti.com TLC2272 test back ckt.TSC

0 Zhihong Huang over 4 years ago in reply to Zhihong Huang

TI__Intellectual 1550 points

4162.TLC2272 test back ckt.TSC

0 Michael Steffes over 4 years ago in reply to Zhihong Huang

Guru 47485 points

That helps, but,

1. Are you sure that 1.8k is not 18k?

2. I can't find 300pF to gnd on the inverting node?

0 Zhihong Huang over 4 years ago in reply to Michael Steffes

TI__Intellectual 1550 points

Hi Michael,

1. 1.8k is 1.8k.

2. The 300pF is C174.

3. Please help me analysis that as soon as possible. thanks!

Best Regards

Wesley Huang

wesley-huang@ti.com

0 Michael Steffes over 4 years ago in reply to Zhihong Huang

Guru 47485 points

Yes, I see that now, The output of this stage also goes over the V+ input of the other stage I think, so there will be some direct parasitic C on that output as well. say 20pF for now between device input C and layout trace.

So first run a DC operating point, output looks to be sitting at 4.07V which looks correct for the circuit - that what they want? Or is the input sitting on some DC level that adjusts this? that 127mV probe is the V+ input

Looks to be stable with our without the 1nF, running closed loop the response does not change much with the 1nF. The problem statement is confusing - sounds like there is a 10kHz noise that they are trying to notch out, this circuit does not simulate to do that so not sure why it is making any difference to add that 1nF across the inputs - it is ok to do that is seems, but the sims are not apparently capturing the full effect he is seeing.

0 Zhihong Huang over 4 years ago in reply to Michael Steffes

TI__Intellectual 1550 points

Hi Micheal,

Thanks for your analysis.

I will check for the customer about whether they decrease the 10K noise by the 1nF Capacitor.
The Gain set in the circuit is near 30dB. So if the customer want to sense the current whose frequency is over 4.7kHz, the output could be wrong from the simulation result in the TINA.

Do I have the right analysis?
What kind of experiment that we could advice the customer to do for evaluating the system?
Could you summary your conclusion and the next action for the customer generally?

Best Regards

Wesley Huang

wesley-huang@ti.com

0 Michael Steffes over 4 years ago in reply to Zhihong Huang

Guru 47485 points

I was just trying to clarify the questions - to go further, perhaps the product group would pick this up. Good Luck,

0 Tamara M Alani over 4 years ago in reply to Michael Steffes

TI__Genius 10850 points

Thank you, Michael, for your help.

Hello Wesley,

Thanks for your post. Please accept delayed response as many of us are on holiday break. We will support you shortly.

-Tamara

0 kai klaas69 over 4 years ago in reply to Zhihong Huang

Guru 140200 points

Hi Wesley,

as always, scope plots would be very helpful. First, a proper GND has to be chosen. This can be pin 4 of TLC2272 or the GND terminal of C178. The GND terminal of decoupling capacitor of TLC2272 can also be used which is hopefully connected closest to the TLC2272. Then, the "2.5V_EM" signal, pin 5 of TLC2272, P40 and P327 should be probed. Do not touch P327 directly with the scope probe but insert an isolation resistor of 47...100R which you directly solder to P327. Also, avoid pigtailing of ground wire of scope probe by using a proper scope probe with ground spring as shown here:

Kai

0 Thomas Kuehl over 4 years ago in reply to Zhihong Huang

TI__Guru* 93105 points

Hi Wesley,

Michael provided a very complete simulation of the TLC2272 circuit with and without the 1 nF capacitor across the op amp inputs. The simulation indicates that the circuit is stable in both cases. I ran an additional analysis to check the phase margin and see about 87 degrees indicating the circuit should be unconditionally stable with the addition of the 1 nF capacitor across the inputs.

This circuit has 20 to 30 dB of gain at tens of kHz so if there is wideband noise coming from the source it will be amplified and passed on to the EM_IH output.

Regards, Thomas

Precision Amplifiers Applications Engineering

TLC2272_Cap_in_stab_01.TSC

Amplifiers

Amplifiers forum

TLC2272: The risk of adding the 1nf Capacitor in the V+ and V-