The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> Simulating Current Reference using Switched Capacitor Resistor
https://designers-guide.org/forum/YaBB.pl?num=1514548287

Message started by sanjayr on Dec 29^th, 2017, 3:51am

Title: Simulating Current Reference using Switched Capacitor Resistor
Post by sanjayr on Dec 29^th, 2017, 3:51am

I need to simulate the current reference. Resistor is implemented using switched capacitor resistor.

I simulated the circuit (and startup) with passive resistor; it is working. When I replaced it with switcap resistor and tried to simulate using Cadence PSS with more tstab time; I could not get the required current. Current waveform looks like a ramp wave. How to simulate this circuit using PSS?

Regards,
Sanjay

Title: Re: Simulating Current Reference using Switched Capacitor Resistor
Post by Ken Kundert on Dec 30^th, 2017, 11:28pm

Why do you think you PSS results are incorrect?

-Ken

Title: Re: Simulating Current Reference using Switched Capacitor Resistor
Post by R.kumar on Jan 2^nd, 2018, 2:22am

The waveform will have a ramp due to the SC operation. However, the average current should be equal to what you have designed for.

Title: Re: Simulating Current Reference using Switched Capacitor Resistor
Post by sanjayr on Jan 2^nd, 2018, 11:25pm

Dear Mr. Kundert,
I attach here the waveforms (of 50 nA current reference) for with passive resistor and Switcap resistor. Both results are not matching. I do not know that Switcap current reference result is correct. What could be the espected result?

Thanks,
Sanjay

Title: Re: Simulating Current Reference using Switched Capacitor Resistor
Post by Tako on Jan 3^rd, 2018, 12:10pm

@sanjayr

Your simulations results seem to be correct. Your current is waveform current when you use SC. You may modify the capacitance value and the waveform control signal (its frequency) and see whether you get expected results.

How can we help you?

Title: Re: Simulating Current Reference using Switched Capacitor Resistor
Post by sanjayr on Jan 3^rd, 2018, 10:44pm

I have designed a Switcap Resistor for 1.01 MΩ. Values are: fclk = 100 kHz, C = 9.9 pF. You are saying that current reference waveform using switcap seems to be correct; but the average current is only 31 nA. The expected current is 50 nA.

I tried different clock frequency and capacitance values for 1.01 MΩ. But the average values are lower. How to rectify the problem?

Thanks,
Sanjay

Title: Re: Simulating Current Reference using Switched Capacitor Resistor
Post by Tako on Jan 4^th, 2018, 1:48am

Do not keep to the R = 1.01 MΩ, if it does not give you proper current. Try other values. Also run longer simulations e.g. 100 x Tclk, where Tclk = 1/fclk to see whether there is no change in the average value of the current over time.

Title: Re: Simulating Current Reference using Switched Capacitor Resistor
Post by R.kumar on Jan 4^th, 2018, 9:01am

The difference could be attributed to charge leakage in the Cap. Since your current is in nA range the charge leakage can have significant impact on the average current. Try simulating with an ideal cap and ideal switch.

Title: Re: Simulating Current Reference using Switched Capacitor Resistor
Post by Ken Kundert on Jan 4^th, 2018, 10:55am

It is strange that when the capacitor is connected its current seems to be settling to 67nA. I would expect it to be trending to zero. Perhaps it is because the voltage on this circuit is a ramp as you suggested earlier.

What happens if you greatly increase the clock period? Is it possible that this circuit is just exceptionally slow?

-Ken

Title: Re: Simulating Current Reference using Switched Capacitor Resistor
Post by R.kumar on Jan 5^th, 2018, 1:22am

Assuming that the clk has 50% dutycycle, to get an average current of 50nA you should have 100nA when S1 is closed. But your waveform shows only 67nA. This could be due to two reasons
1. some of the current is shunted though S2. Check the off-resistance of S2. This should be much greater than the effective impedance of the Cap.
and/or
2. The on-resiatnce of S1 is large enough to limit the current. In this case your degeneration resistance is more than 1.01MOhm.