Op-amp circuit grounding exercise questions

Weaver · Nov 8, 2018

Homework Statement

Circuit schematic of a humidistat: a type K thermocouple pair (dry and wet) feeds op-amp ADA4528 (RS 759-2086), then resistors R1 and R2 to an MBED LPC1768 Ain input; Dout drives MOSFET Q1 IRL530N switching fogger X1, with 5V in via 7805 regulator and +24V supply.

Worksheet of numbered questions on LPC1769 MBED grounding and noise, each with a blank answer box and a unit: uV, V/V, ohms, mV, uF, us, kHz, V.

Worksheet form listing grounding questions: estimate DC voltage drops in mV on MBED, Q1, A1, R1 R2 and 7805 ground wires, each with an ADC LSB conversion in bits, plus thermocouple and PCB track current questions.

Circuit schematic titled Answer Sheet 3: Circuit design, showing a Dry and Wet sensor input to op-amp A1, feeding Ain of an MBED LPC1768 with R1 and R2 resistors, plus a 7805 regulator from +24V to 5V in and a 3.3V out, with a crystal and MOSFET at the right.

2. Homework Equations
3. The Attempt at a Solution
For the first set of questions:

Handwritten answer sheet showing offset calculation (17 x 41 uV/C = 697 uV), gain A = 3.3/697u = 4.73e3, and an op-amp circuit with thermocouple, input resistor R1, feedback resistors R1 and R2, and bias current sources Ib.

Handwritten derivation on lined paper: Ib(R1 || R2) < 1 mV, then Ib = 400 x 10^-12 (R1 || R2) < 1 mV, leading to (R1 R2)/(R1 + R2) and finally R1/R2 = 4.7336 x 10^3, R1 = 4.7336 x 10^3 R2.

Handwritten working on lined paper: resistor ratio calculations with R1 = 10 MΩ and R2 = 2.2 kΩ, then an input offset voltage expression 400 x 10^-12 (R1 || R2) evaluated to 879.80 x 10^-9 V.

Handwritten working on lined paper: 0.1 ohm + 0.2 mH becomes 0.1 + j(2π×1.6 MHz×0.2m) ohm, simplified to 0.1 + j(…π) ohm, then 100 mA × that gives 10×10^-3 + j(…π) V boxed. Below, 12-bit ADC, 3.3 V max gives 3.3/2^12 = 8.0537 mV boxed.

I've worked through to part 6), at which is I encountered my first problem. I'm not entirely sure what the question is asking. Is it as if there would be a capacitor between Q1 and X1? Is the voltage to be reduced the one calculated in part 5)?

Moving on from that

Handwritten notes: 1) Qd/Imax = t_switch, with 12x10^-9 / 10x10^-3 = 1.2x10^-6 s = 1.2 us; 8) frequency = 1/time, giving 833.333 kHz; 9) left blank.

I am also unsure about part 9) and 10). We haven't really covered MOSFETs in lectures so these two questions are a bit lost on me. Do I use the frequency, current and resistance to get the average value?

Handwritten answer sheet 2 with worked calculations: item 1 gives typical length 0.1 ohm, dV drop = 0.1 x 90 x 10^-3 = 9 mV; item 2 gives 1 x 10^-3 over 0.8057 x 10^-3 = 11.17; item 3 gives 1 A x 0.1 = 100 mV; item 4 gives 100 x 10^-3 over 0.8057 x 10^-3 = 124.17; item 5 is blank.

For the second answer sheet for I haven't worked out 5), 7), 9), 12) However I am working on this right now. I will probably have questions later on this

I don't want the answers, but rather help on what methods I should be doing or perhaps pointers to resources that would help

Also, I'm not even certain what I have done is correct

I appreciate any help

Thanks in advance,

anorlunda · Nov 8, 2018

Those images under attempt at a solution are too poor quality to read. You probably won't get any answers until you post something more legible.

Weaver · Nov 8, 2018

anorlunda said:

Those images under attempt at a solution are too poor quality to read. You probably won't get any answers until you post something more legible.

Thanks for that. I'm at uni, so I don't have access to a scanner. I will update them later tonight

Handwritten answer sheet with an op-amp circuit: a thermocouple with source Vs and series resistor Rs feeds the non-inverting input, with bias current Ib+ and offset source Vos, and the inverting input has feedback Rx, Ib- and Ios, plus resistors R1 and R2 to ground.

Photo of handwritten working on lined paper: I_b(R1 || R2) < 1 mV, then 400 x 10^-12 (R1 || R2) < 1 mV, expanded to 400 x 10^-12 (R1R2)/(R1+R2) < 1 mV, leading to (1 + R1/R2) = 4.7336 x 10^3, R1/R2 = 4.7336 x 10^3, and R1 = 4.7336 x 10^3 R2.

Handwritten working on lined paper: algebra solving for resistor values R1 and R2 so an input bias error current stays under 1 nA, a chosen value of 2.2 kΩ, and a check of the resulting gain and error voltage.

Handwritten worked solution on lined paper: 1) Q_G = γ_switch × I_max, so Q_G / I_max = γ_switch, giving 12×10^-9 / 10×10^-3 = 1.2×10^-6 s = 1.2 μs. 8) Frequency = 1/time, giving 833.333 kHz. Part 9) is started with no working.

Handwritten answer sheet titled "Answer Sheet 2" with numbered working: line 1 gives typical wire resistance of 0.1 ohm and voltage drop 0.1 x 90 x 10^-3 = 9 mV; line 2 shows 1 x 10^-3 over 0.0897 x 10^-3 = 11.17; line 3 shows 1 A x 0.1 = 100 mV; line 4 shows 100 x 10^-3 over 0.9097 x 10^-3 = 124.17; item 5 is blank.

Weaver · Nov 11, 2018

Conor_B said:

Thanks for that. I'm at uni, so I don't have access to a scanner. I will update them later tonight
]

I've decided to make smaller threads asking about the problems i have been having more specifically with typed solutions (my scanner broke) so that it is easier to follow. I would delete this thread, but I don't know how

https://www.physicsforums.com/threads/calculate-current-following-in-op-amp-to-ground.959831/

https://www.physicsforums.com/threads/decoupling-capacitor-for-pcb-track.959829/

https://www.physicsforums.com/threads/ground-current-for-voltage-regulator.959832/

CWatters · Nov 11, 2018

It was a mistake to have several threads running on the same problem.

Regarding the capacitor on X1 Q1... they are talking about a decoupling capacitor you need to add. It would be on the 24V rail. See my reply in your other thread where I mention ripple currents.

Weaver · Nov 11, 2018

CWatters said:

It was a mistake to have several threads running on the same problem.

Regarding the capacitor on X1 Q1... they are talking about a decoupling capacitor you need to add. It would be on the 24V rail. See my reply in your other thread where I mention ripple currents.

Apologies for that. I thought it might help break up the question. I won't do it again. Thanks for replying in the other forum

CWatters · Nov 12, 2018

I haven't looked at your working for 1-6 properly but..

For 9. It asks you to calculate the voltage drop down a typical PCB trace due to the 10mA FET gate current. Use the frequency you calculated in 8.

EDIT: In case it's not obvious this current flows in and out of the MBED power/ground pins on it's way to/from the FET gate.

For 10: Similar to Q6

Weaver · Nov 12, 2018

CWatters said:

I haven't looked at your working for 1-6 properly but..

For 9. It asks you to calculate the voltage drop down a typical PCB trace due to the 10mA FET gate current. Use the frequency you calculated in 8.

EDIT: In case it's not obvious this current flows in and out of the MBED power/ground pins on it's way to/from the FET gate.

For 10: Similar to Q6

Thank you very much! I get and I think I have those out figured. When it asks average AC, does that mean RMS?
I’ve followed through with the rest of the questions. The last one I have a problem with the op amp ground current

Should there be a value in a data sheet for this? Or is it calculated?

If it’s calculated, at worse case, the op amp will have 0V at it’s In(+) so no voltage appears at the output. So at this point it is just the supply current into the op amp?

CWatters · Nov 12, 2018

Conor_B said:

Thank you very much! I get and I think I have those out figured. When it asks average AC, does that mean RMS?

That's what I would answer. Not sure why they ask for that as it's the peak to peak that more important.

Edit: Earlier they suggest you model the output as a 10mA current source feeding the gate capacitance of the FET. So perhaps think carefully about the waveform and how you do the rms calculation.

CWatters · Nov 12, 2018

Conor_B said:

The last one I have a problem with the op amp ground current

Should there be a value in a data sheet for this? Or is it calculated?

If it’s calculated, at worse case, the op amp will have 0V at it’s In(+) so no voltage appears at the output. So at this point it is just the supply current into the op amp?

I agree.

Weaver · Nov 12, 2018

CWatters said:

That's what I would answer. Not sure why they ask for that as it's the peak to peak that more important.

Edit: Earlier they suggest you model the output as a 10mA current source feeding the gate capacitance of the FET. So perhaps think carefully about the waveform and how you do the rms calculation.

Right, so the waveform is generated from the resistance and inductance of the the PCB with the ripple current produced from the MOSEFT. In this case is the Vripple close to a square
waveform as it is the MOSFET switching?

(By the way, I really appreciate all the help you've given)

CWatters · Nov 13, 2018

No isn't this Q9 where they ask about the MPEG average AC current? Eg due to the FET gate current not the source/drain current.

~~What I was hinting at is that a constant current (10mA) driving a capacitor (FET gate) produces a triangle wave~~

CWatters · Nov 13, 2018

Actually forget that. The gate voltage will be a triangle wave but the ripple on the supply voltage will be a square wave as it's the constant 10mA * PCB trace impedance.

Weaver · Nov 13, 2018

CWatters said:

Actually forget that. The gate voltage will be a triangle wave but the ripple on the supply voltage will be a square wave as it's the constant 10mA * PCB trace impedance.

Okay, so as it's a square wave, the Vrms is just the peak voltage in this case?

CWatters · Nov 13, 2018

Yes that would be my answer for that question.

Weaver · Nov 13, 2018

CWatters said:

Yes that would be my answer for that question.

Great! Thank you for all the help you've given. Definitely helped improve my understanding! Really appreciate it

Op-amp circuit grounding exercise questions

Homework Statement

Discussion

Similar threads