Maker Pro
Maker Pro

A 1.5V to (1.5V -> 6.5V) level shifter, and vice versa

A

alan

Jan 1, 1970
0
Basically, my boss wants me to build a testing system for an IC, which
has an absolute maximum rating of 1.5V to 6.5V VCC.

So my boss wants me to apply some digital signals to the IC and read
some feedback (at up to 20MHz clock). With a variable VCC from 1.5V
to 6.5V (and other features besides, such as being low power). And I
need to be able to apply 1.5V to 6.5V voltage at high, based on my
current VCC.

However the only level shifters I've found are either 1.5V to 5.5V
supply voltages or 4V to very high supply voltages. Nothing quite
fits the 1.5V to 6.5V range.

The best I've managed to come up with is to use a bunch of
(mechanical) relays to select between two level shifters, one for 1.5V
to 5.5V range, the other for 5.5V to 6.5V range. I'm not sure if
solid-state relays will work, because I'm almost sure that the
propagation delay of actual metal is much faster than propagation
delay of semiconductor.

Is this the best solution? Might I have missed a level shifter
capable of reaching the required range?

My controlling device is an FPGA, which can have a VOH of 1.2V, 1.5V,
3.0V and 3.3V (1.5V and 3.3V being the preferred, because our FPGA
prototyping board has only those voltages). The maximum recommended
VIH for the FPGA is 4.1V, with the minimum obviously varying according
to the I/O VCC I select.

Thanks in advance!
 
B

Brendan Gillatt

Jan 1, 1970
0
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Basically, my boss wants me to build a testing system for an IC, which
has an absolute maximum rating of 1.5V to 6.5V VCC.

So my boss wants me to apply some digital signals to the IC and read
some feedback (at up to 20MHz clock). With a variable VCC from 1.5V
to 6.5V (and other features besides, such as being low power). And I
need to be able to apply 1.5V to 6.5V voltage at high, based on my
current VCC.

However the only level shifters I've found are either 1.5V to 5.5V
supply voltages or 4V to very high supply voltages. Nothing quite
fits the 1.5V to 6.5V range.

The best I've managed to come up with is to use a bunch of
(mechanical) relays to select between two level shifters, one for 1.5V
to 5.5V range, the other for 5.5V to 6.5V range. I'm not sure if
solid-state relays will work, because I'm almost sure that the
propagation delay of actual metal is much faster than propagation
delay of semiconductor.

Is this the best solution? Might I have missed a level shifter
capable of reaching the required range?

My controlling device is an FPGA, which can have a VOH of 1.2V, 1.5V,
3.0V and 3.3V (1.5V and 3.3V being the preferred, because our FPGA
prototyping board has only those voltages). The maximum recommended
VIH for the FPGA is 4.1V, with the minimum obviously varying according
to the I/O VCC I select.

Thanks in advance!

I would suggest you use the 1.5v boost converter and a 4.7v zener to
limit the voltage.

- --
Brendan Gillatt
brendan {at} brendangillatt {dot} co {dot} uk
http://www.brendangillatt.co.uk
PGP Key: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xBACD7433
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (MingW32)

iD8DBQFHAQk3kA9dCbrNdDMRAq+iAKDf98JQLfUMl7ZhTOQ1J+jOvCN52gCgq3vQ
tQFzVUNYCRktYNPlDVh5Q1k=
=PzFr
-----END PGP SIGNATURE-----
 
L

Lostgallifreyan

Jan 1, 1970
0
Sometimes an input might be driven with a voltage reaching or exceeding the
supply voltage, If you have to test this condition, you need two variables,
your supply, and your pullup voltage.

Once you have those it's easy with the method I copy from an old post of
mine (idea raided from a 'Best of EDN' book I found).

Try an n-channel JFET, gate to common ground, pull-up voltage through
resistor on the drain, input to the source. Anything over 1 to 2V input
should be fine. Without having seen your circuit I can't be sure, but this
might all be doable with a single FET per input. I'm not sure how low your
pullup voltage can go but if the FET is satisfied, you could offset the
final voltage the inputs see by passing through one or more signal diodes
between FET drain and input. You'd then have a continuously variable range
to test by adjusting the main pullup supplie's voltage.
 
L

Lostgallifreyan

Jan 1, 1970
0
pullup supplie's voltage.

Note to self: If I ever make a typo that tasteless again I'll shoot myself,
I swear. >:)
 
Top