I am building an analog to digital converter of the SAR type. I've got most of the circuit designed and tested. All that's left is the sample and hold and comparator circuits. I need to purchase a bunch of transistors for my project. This will be an 8-bit SAR ADC running at 12 MHz, which translates to 1.5 MHz for the sample and hold circuit. The transistors need to be sized correctly, but data sheets don't give the gate width's for the transistors. If I understand correctly, I need to be looking at the TpHL and TpLH. I could use some tips on how to do the search via Mouser or Digikey for the transistors that will work for my project. The comparator circuit I plan to use is shown in the picture. This is a dynamic two-stage latched comparator.

Thanks for your help,
Mike

 

It will be very hard to find small-signal discrete MOSFET because they are not very common. Also with discrete MOSFET you will need to add resistors in to the circuit to set the current in the desirer range. Also MOSFET will have large spread between his main parameters. And this is why we don't build this typo of a circuit with a discrete device. Try BSS84 and BSS138P.

 

Why are you trying to build an 8-bit ADC from discrete components? Is that an explicit requirement of the assignment?

 

Yeah I am supposed to build most of the circuit with discrete components. My instructor does allow some nondiscrete components though. I just have to ask him about specific components. Also, I need to be able to breadboard this circuit. I plan to design a circuit board using SMD's too once I get the breadboard circuit working correctly. Do you have suggestions for nondiscrete components my instructor might ok. I asked him about using DG212 for the switch in the sample and hold and he said ok for that. Also, can you suggest nondiscrete components for the two amplifiers to use in the sample and hold circuit? You can just give me a part name and I'll search for them myself.

The digital to analog part of the circuit doesn't have to be discrete and I already got the IC for that.

Thanks for your help,
Mike

 

If I understand correctly, I need to be looking at the TpHL and TpLH.
Mike

Not sure if you know what that means exactly, so just gonna give a quick post.

TpLH= Time propagation(sp?) LOW to HIGH. In my words, the time delay required to switch from a 0-1.

TpHL= time propagation HIGH to LOW. Same thing in reverse.

Also keep in mind, setup and hold times.

As far as what to use for amplification, transistors. Research & experiment with different combinations.

EDIT: (now realize you want non-discrete) EDIT2: (Just confused now. Ignore, lol)

 

You are right that I didn't fully understand those two terms.

Alright, let me clarify, some parts of the circuit don't have to be discrete, but most of it must be discrete. I asked my instructor about using DG212 for the switch in the sample and hold circuit and he said I could use that. My understanding is that this isn't a discrete component, so the line between discrete and nondiscrete seems to be blurred.

I need to buy through hole components so I can breadboard it for testing. When I search for a mosfet transistor via Mouser.com I can narrow my search by filtering, but I don't see an entry for capacitance. With my understanding of this circuit, I need a low capacitance transistor so there isn't too much leakage from the holding capacitor. Also, I know that the input impedance has to be high and the output impedance has to be low. Please correct me if I'm wrong.

Thanks a lot,
Mike

 

Ok, I'd like to share these few words with you

[Pulled from Wiki]

Interconnect capacitance
Traditionally, switching time was roughly proportional to the gate capacitance of gates. However, with transistors becoming smaller and more transistors being placed on the chip, interconnect capacitance (the capacitance of the metal-layer connections between different parts of the chip) is becoming a large percentage of capacitance.[41] [42] Signals have to travel through the interconnect, which leads to increased delay and lower performance.

--This is something, you may have to experiment with, as I have very little hands-on experience with FET's (which is why I pulled that from wiki), but in your application, look for a data sheet accompanied with the component of interest. I do know there are ways to take that unwanted capacitance and use it towards your advantage in your circuit, well maybe not advantage but turn it into something good. These 2 links may help, I ran across:

1) http://bwrcs.eecs.berkeley.edu/Classes/icdesign/ee141_f10/Lectures/Lecture11-MOS_Cap_Delay-6up.pdf
2 http://www.doe.carleton.ca/~len/477W2003/LectureNotes/January_13_2003_2up.pdf

What you are describing in your recent post is an ideal FET. High input impedance and low output impedance, those 2 characteristics are fundamental for any sort of amplification. Do you understand why this is? If not, i can proceed into further detail.

From what I can tell you are on the right track, and your over thinking a few things possibly. Get back to us when you can and we will try and get you squared away.