Re: Gaasfet switch driver needed

On Thu, 26 Oct 2006 22:37:13 GMT, no_email@xxxxxxxxxxxxx (Charles
Luke) wrote:

I need a switch driver that can switch in 1 to 3 nsec. Complementary
drive outputs, 0 to -5 or -8 volts. Driving a SPDT Gaasfet switch.

I've tried differential A-to-D drivers but they are not fast enough.
Input to the driver is from a cmos asic. Also looked at comparators
with ECL output but they don't have the voltage swing.

Impellimax makes a driver but they want mucho$.

Other constraints. Battery powered(low powered). Small size.

I'm looking at a discrete driver or biasing the Gaasfet switch.

Can bias the Gaasfet switch for a 5V ground and use 0 to 5V switch
outputs for the driver. This will be at 4.5GHz. Anyone ever done this?
What are the ramifications? Lost isolation? Bad RL? If I go this way,
is CPWaveguide the way to go to minimize via inductance?

Charlie Luke
Luke Engineering, Inc.


Shifting the supplies would sure be easiest. I can't see why that
would affect the RF, since you're probably AC coupling that in/out
anyhow. Now you could use fast CMOS gates as drivers, and have zero
static power dissipation. Maybe a TinyLogic flipflop, or a tiny dual
XOR, as a diff driver?

If you wanted to stay with the negative supply, an LVDS-to-TTL
receiver chip would be the basis for a decent driver. Say a DS90C402
powered by ground and -5, with a little level shifting on the input.
500 ps edges!

Or just shift a 5-volt CMOS swing down 5 volts, to 0 and -5, with a
bypassed zener or bandgap reference and a pulldown resistor... the
GaAs gate input probably needs very little current.

I've used the Hittite GaAs switches at speeds like this, but the ones
I use are TTL compatible when run from +5 and ground. No surprises,
except that whan they say "DC to xx GHz" they don't really mean DC...
check the S-params for hints of low frequency funnies.

CPW is fine for stuff like this. Avoid signal vias if possible, but
4.5 GHz isn't *really* fast.

Whose switch are you using? Why do you need to switch so fast?

John

.

Relevant Pages

  • [git patches] IDE updates (part 4)
    ... ide: add IDE_HFLAG_BOOTABLE host flag ... alim15x3: always tune PIO ... Fix siimage driver accessing beyond array boundary ... There is a 25/33MHz switch in configuration ...
    (Linux-Kernel)
  • Re: [PATCH] atmel_serial: update the powersave handler to match serial core
    ... I think that a driver can do the request to a the gpio layer (may can be implemented ... pin change interrupt. ... For waking up, we need to switch on the main oscillator, wait until it ...
    (Linux-Kernel)
  • RE: RapidIO - general questions
    ... we sometimes find that, although a switch may ... can train your code to avoid that port or try it later. ... Subject: RapidIO - general questions ... Aux driver - basic driver, ...
    (Linux-Kernel)
  • Re: [PATCH][RESEND] New type of DTV2000H TV Card
    ... But I had installed GPG and signed my patch, maybe this is what I should have done earlier. ... This piece of code used to switch the input of RF signal to "Air Antenna" ... if this patch (and driver) would be able to switch RF ... inputs in DVB-T mode somehow. ...
    (Linux-Kernel)
  • Apparent flow control problem between Gigaswift cards and Cisco C atalyst 6500 switch
    ... I have an E220R running Solaris 9 with a fiber gig card with the ce driver. ... When I hook it to one of our Cisco Catalyst 6500 switches, ... driver, also with a catalyst 6500 switch (but not the same one, which rules ...
    (SunManagers)