Query regarding Delay element in the circuit.

[shaikss]

Folks,

I have a query regarding the attached circuit.
In order to design a rectifier, I was trying different structures.

So, I started with the attached circuit without delay elements.
I have used a cap at the NMOS, so that the cap will gets charged. But I don't want the cap to discharge its energy.
When I don't use delay elements, the capacitor has discharging path and so my output was not good

When I use delay of 0ns or 0.1ns or whatever value, delay element is working as a isolator.
I see reasonable output even at 180mV. But the settling time is still high.

Can you please let me know what can be done so that cap doesn't discharge?

Thanks

 

[yungman]

I don't follow what you are trying to do. By looking at the schematic, I can spot a few problems.

1) If you delay line is labeled 500pS delay, this is a very short delay. The delay line must be 50 ohm impedance. You don't have any termination. That is not going to work as you are going to have reflection.
2)You put a cap at the source of the NMOS and you tie the gate and drain together, it does not make sense.
3) I don't know what you are trying to do with the PMOS as you tie the source and gate together! The FET never turn on.

I don't understand why you try to design a rectifier with this, I think you should describe what you want to get out of this circuit and there must be a much better way to implement a circuit.

 

[shaikss]

Thanks. I want to do rectification of lower voltage.
If I use FW bridge rectifier, vol drop is 1.4V minimally and even if I use MOSFET, there is 0.45V drop. I can't go for Schottky diodes as they have larger leakage currents. Native transistors are ruled out because of their cost.
I thought that inclusion of cap, makes the cap to charge and that voltage is presented to the output.

What do u suggest me to do? How should I proceed further?

 

[yungman]

It's getting late here. I'll think about it tomorrow. Do you have any high speed requirement that you use a delay line like this? Or is it just rectifier with very low drop out that you are looking for?

What is you input signal amplitude? You want to have a DC envelope following the peak of the signal like a peak detector?

I am trying to understand what you want.

 

[shaikss]

It's getting late here. I'll think about it tomorrow. Do you have any high speed requirement that you use a delay line like this? Or is it just rectifier with very low drop out that you are looking for?

What is you input signal amplitude? You want to have a DC envelope following the peak of the signal like a peak detector?

I am trying to understand what you want.

Thanks Yungman for your quick reply.
I don't have any high speed requirement.
My input ac freq is 865MHz.
I am designing a rectifier which operates at very low voltages like 100mV, which means low drop at the devices.
Signal amplitude ranges between 100mV to 1.2V.Whatever the signal I receive from an antenna is rectified and is given to digital blocks and memory device like EEPROM. at the same time, the received signal is then demodulated using envelope detector.

Hope you understood what's my requirement.
If you have anything in your mind, pls mail me.

Thanks!

 

[yungman]

So you are just doing a simple demodulator or peak detector for AMPLITUDE modulation?

I take that you have other power supply that you can use? If so, there are plenty of circuit in the existing demodulator that can do the job for you.

If you don't have any external supply you can use and you want a zero threshold rectifier, that is going to be hard.

 

[yungman]

I just assume you don't have external supply so you cannot use any amplifier to amplifier the signal before rectifying.

I can't think of any passive rectifier that don't have a turn on threshold. I assume your antenna has output impedance of say 50ohm. One way to boost the signal is to use a step up transformer. Mini Circuit has a lot of these type of transformer. You can use a 1:4 step up to increase voltage by 4 times so you can use a schottky rectifier. ( I don't know what you mean by too much leakage). Problem is the output impedance is going to be step up the the square of 4 which is 16. So the output impedance after stepping up is 800ohm which is quite high. But I think you can still use schottky rectifier. One problem I can see is the reflection back to the antenna as the impedance of the rectifier is not linear and hard to put exactly a 800 ohm load.

If you want more help, you need to get more information what do you expect out of it, the characteristic of the antenna etc.

 

[shaikss]

Thanks Yungman.
As you guessed, I am not using external supply. Whatever the subsequent digital blocks are present, they are driven by the rectified dc voltage.

The input power is in terms of 30-40 microwatts and even less.
I am keen on rectifier with low threshold voltages.
If we use schottky diode, it may cause problems in layout on CMOS. So, schottky is not preferred choice. The best choice is to use MOSes with zero threshold voltage or very minimal threshold voltage drop.

Any help on this?

 

[yungman]

So you are putting in a CMOS IC. As I mentioned before, your best bet is to use a transformer to boost the voltage up first.

Maybe you can use a higher ratio step up transformer to boost the voltage even higher and then just use the regular rectifier to get the DC. Look at Mini Circuit to see whether you can find a transformer.