Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

chargepump in RFID Tag

Status
Not open for further replies.

shaikss

Full Member level 4
Full Member level 4
Joined
Jun 18, 2007
Messages
229
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,298
Activity points
3,319
Hello,

Normally array of caps are used for Charge pumps.
Can you please explain how the attached charge pump works?
Its not a series or cascade connection. I am confused in analysing this circuit.

Please help me.

Thanks,
Shaikss
 

Attachments

  • charge_pump.gif
    charge_pump.gif
    51.6 KB · Views: 114

I think it's just a variation of the Dickson charge pump: instead of cascading 2 voltage doublers, it uses 2 in parallel, stacking their voltages.
 

I think it's just a variation of the Dickson charge pump: instead of cascading 2 voltage doublers, it uses 2 in parallel, stacking their voltages.

Is there any advantage going for parallel (voltage stacking) instead of cascading.
How does it works actually?
 

Is there any advantage going for parallel (voltage stacking) instead of cascading.

I'd think the decision (serial or parallel) depends on the source impedance. Try and simulate it!
 

I'd think the decision (serial or parallel) depends on the source impedance. Try and simulate it!

Hi,

I had seen some strange behaviour when I simulated chargepump (dickson) constructed using MOSFETs.

What is the switching frequency to be used? How should we select Cap value?

I have chosen 500KHz as the clock frequency and cap is 1pf.
I tried b varying the clock frequency from 250KHz to 1GHz. This charge pump is used for low power RFID applications@ 865MHz.
I am aware that there are -ve implications when we go for higher frequency. but I experimented to see the behaviour.
Moreover, I changed the width of the MOSFET from 240nm to 1um.

Strangely, I saw that as the width is increased, output is decreased. I assumed that as width of MOSFET is increased, more will be the drain current and so output voltage must be higher. Surprisingly, I found the reverse, wherein, output voltage started decreasing as the width is increased. Why is it happening so?

I changed the capacitor value from 0.01pf to 1uf. Here also, I agree that cap should be small in order to see faster response time. But still, I experimented to see the results. To my surprise, I found that as cap is increased, response time is decreased. I don't understand why is it happening so.

Can you please explain the logic behind this?

Thanks!
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top