The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
From: Chuck Simmons
Organization: You jest.
X-Mailer: Mozilla 4.61 [en] (X11; U; Linux 2.0.33 i586)
Subject: Re: Shifting TTL to 5V or 3.3V CMOS levels
Date: Fri, 27 Sep 2002 20:40:50 GMT
NNTP-Posting-Date: Fri, 27 Sep 2002 13:40:50 PDT
> In article <3D9486FE.2529F574@webaccess.net>, Chuck Simmons
> > markp wrote:
> > >
> > > "Sir Charles W. Shults III" wrote in message
> > > news:Sq_k9.6977$O8.email@example.com...
> > > > I can think of a hundred ways to do this- start with a simple voltage
> > > > divider. You can use a pair of resistors to cut the logic levels down to
> > > 3.3
> > > > volts easily. Now think "switchable". MOSFET on the upper leg perhaps?
> > > Or a
> > > > simple driver?
> > > > There are even chips that will to it easily enough and some are old
> > > style
> > > > CMOS, but I don't know what your switching times are going to be..
> > > >
> > > > Cheers!
> > > >
> > > > Chip Shults
> > > > My robotics, space and CGI web page - http://home.cfl.rr.com/aichip
> > >
> > > Thanks, but there are 16 (actually 17) of them, and I need to translate from
> > > TTL to 5V CMOS or 3.3V CMOS outputs on all, so adding discrete devices for
> > > each signal on the output is not really practical (I might as well use a
> > > level shifting buffer which I don't want to do). The resistor divider may
> > > also work for really slow things but the resistance needed to keep the power
> > > down would slug the line too much.
> > >
> > > The logic devices such as HCT translate to 5V CMOS directly from TTL, but
> > > are specified 4.5V - 5.5V for Vcc, so can't use them for 3.3V (Actually, it
> > > may be possible with some families to run at 3.3V but the input thresholds
> > > are then not specified in the datasheets). Switching times needed are around
> > > 60ns max so not too restrictive. Most CMOS type devices are specified
> > > 0.3*Vdd for Vil and 0.7*Vdd for Vih, which puts it at 3.5V for 5V Vcc, and I
> > > only have TTL levels coming in!
> > >
> > > There might be a logic family out there specified for this, or even a small
> > > PLD with selectable I/O voltage..
> > >
> > > Mark.
> > A lot of 3.3 volt CMOS devices are 5 volt tolerant. In one iteration of
> > a board design (a large board with mostly 3.3 volt parts) I have 3.3
> > volt devices, one 3.6 volt device and a good many TTL and 74F devices.
> > The 3.3 volt stuff all was designed with 5 volt tolerant inputs so I had
> > no problem. 3.3 volt CMOS drives TTL inputs just fine and 5 volt
> > tolerant inputs really are. The mix is not ideal but the interface is
> > IDE (5 volt) and the supplies are +5 and +12. The IDE is driven with a
> > 3.3 volt part and meets specification.
> > Chuck
> What are your clock speeds? If you are going down in voltage, use a 5v
> tolerant part like recommended before in the thread. If you are going up,
> look out and email me, please. This can be a pain.
We are running pretty slow. We do have another part (3.3 volt) that does
ATA100 but we are content on my project with ATAPI at 33MHz. The clocks
are synthesized and are rather arbitrary. The part I mostly work with
uses 40MHz for a base clock. I have a lot of mixed signal which limits
clocks. For example, I run my A/D SAR at 20MHz for a leisurely 0.6usec
10 bit conversion. The board is large. I have more trouble with distance
than anything. I can drive a 40MHz clock about 6 inches. Beyond that
(and I don't need it) I will have to do transmission lines. The 3.3 volt
sags a lot with distance (cannot drive other 3.3 volt parts). Series
terminators at the driving end can do wonders (see the ATA100
specification for an example - the whole damned bus).
... The times have been,
That, when the brains were out,
the man would die. ... Macbeth
Chuck Simmons firstname.lastname@example.org
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup