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.

ask for voltage independent delay cell

Status
Not open for further replies.

pianomania

Member level 5
Joined
Jun 15, 2006
Messages
84
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
1,906

as the pic shows , i design a delay cell by long channel length MOS, but the delay time is variant by VLV .
Is any possible we can design a simple circiut by only simple modified inverter and it won't be infleuence by Vcc ?

It is very important , since we need to appy at deadtime control on power supply application.

Ask for Help ....pleeeeze
 

ieropsaltic

Full Member level 4
Joined
Sep 25, 2006
Messages
199
Helped
64
Reputation
128
Reaction score
18
Trophy points
1,298
Activity points
2,595
One possible method would be using current sources at the sources of both the NMOS and the PMOS (of equal currents of course) and using the NMOS and PMOS devices to switch the currents. But this is not a simple solution as you'll need to implement a current source.
 

dick_freebird

Advanced Member level 5
Joined
Mar 4, 2008
Messages
7,242
Helped
2,118
Reputation
4,240
Reaction score
1,971
Trophy points
1,393
Location
USA
Activity points
58,069
Yes, you's want to move to a current-starved inverter
design with a master reference whose tempco and
supply_voltco produces a flat delay profile -at the
point of interest- (logic delay dead flat, plus driver taper
chain delay not-flat, would still not be uniform; your
logic delay might need to have a counter-slope for a
net flat result).

So following this train of thought, perhaps you only need
to have one delay-comp block and let the rest of the chain
be simple. That's a better outcome for routability and
such, if you can make it work.

Or, you could make a deadband (anti-shoot-through?)
scheme that is not "ballistic" but based on sensing the
transition further down the chain on the opposite leg,
then you will adapt to the situation without embedding
so much guessing or assumptions about tolerances and
environmentals.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top