Re: [uCsimm] Are the uCsimm I/O pins 5V Tolerant?

From: Tom Walsh (tom@mytoys.com)
Date: Thu Feb 03 2000 - 18:44:47 EST


John Ananny wrote:

> Hi all,
>
> I've been lurking for a few months on this list, but would like to
> contribute my $0.02 to this 5V-tolerance discussion.
>

    This is bad, you shouldn't lurk... Why not join right in and make
some guesstimates like some of us are doing? ;^)

>
> I don't have the data sheets in front of me, but I'm 98% confident
> that the EZ's inputs are clamped to the rail with an input protection
> diode. That means that you'll have two conflicting drives on the
> pin. The diode wants to keep the input pin no higher than one forward
> diode drop above the EZ's supply rail, but whatever output is driving
> the input wants to keep the input pin at 5V

> This is bad.
>

    Agreed.

>
> At the very best, you get a valid logic high and nothing overheats at
> steady-state. At the worst, a wirebond somewhere gives because of
> heating due to overcurrent, either at the input protection diode or in
> the output drive of the previous stage. In any case, you squander
> power because of the overcurrent condition.
>
> IMHO, absolute maximum specs aren't good design targets.
>
> A simple option for prototyping is to use 5V LS buffers. The output
> high of a 5V LS device is only actively driven to approximately 3.5V,
> so have your 5V HC device drive a 5V LS device, and have the 5V LS
> device drive the 3.3V HC inputs to the EZ. In the other direction,
> it's not quite so clean, as you need a pullup on the output of the LS
> buffer to get it all the way up to a valid input high for the 5V HC.
> This does give you an opportunity to make a response time/power
> efficiency tradeoff in your design. For a production device, adding
> LS buffers likely wouldn't be cost-effective, but it does the trick
> for a quick prototype.

    I would stay away from LS as when a device is new it can be
'healthy' enough to go to 4.2 volts VoH. But you are correct, if you
are interfacing 3.3v logic to +5 logic, use LS (or about anything but
HC). If you really need to kludge a +5 to 3.3v interface between LS and
3.3 logic, why not build a resistor totem and pick off at the lower
resistor, example:

                +5
                   |
                   \
                   /
                   \ 1K
                   /
                   |
                   |
 TTL --->O
                   |
                   |
                   \
                   /
                   \ 1K
                   /
                   |
                   |
                  O---------> DragonBall input
                   |
                   |
                   \
                   / 2K
                   \
                   /
                   |
                   |
              -----
              / / /

This would give you about 2.5 volts to 0v swing and won't damage the TTL
gate as its output impedance is about 400 ohms?

TomW

>
>
> Horowitz and Hill (2nd ed.) section 9.03 has further details on
> interfacing logic families and levels.
>
> Best regards,
>
> John Ananny
>
>
>
>
>
>
>
> At 10:48 AM 2/3/00 -0700, you wrote:
>
>> Hi Tom,
>>
>> I thought I would keep you informed as to the answer to my
>> question...
>>
>> So far the response from Motorola has been conspicuously quiet!
>> However, upon reviewing the 68EZ328 user's manual, it does imply
>> that the inputs can accept 5V inputs. The reason I say this is on
>> page 18-1...
>>
>> Supply Voltage (Vcc) -0.3 to 7.0V
>> Input Voltage (Vin) -0.3 to 7.0V
>>
>> Normally, Vin is referenced to Vcc and not an absolute value. If I
>> look at other data sheets for other devices, such as the 74LVX parts
>> (5V tolerant parts) as you mentioned, they define Vcc and Vin the
>> same way (absolute). If I look at 3.3 V only parts, I see that Vin
>> is referenced to Vcc... So, it would seem that the 68EZ328 is a 5V
>> tolerant device... I sure wish Motorola would place a definitive
>> statement in their user's manual /data sheet so as to remove any
>> doubt!
>>
>> Anyway, I have sent an additional e-mail to a FAE that has access to
>> Motorola tech people. I am hoping to hear from him over the next
>> few days to see what he has found out. As soon as I find out, I
>> will post the info to the list.
>>
>> Cheers,
>>
>> Sam
>>
>>
>> > ----- Original Message -----
>> > From: Tom Walsh
>> > To: ucsimm@uClinux.com
>> > Sent: Wednesday, January 26, 2000 6:49 PM
>> > Subject: Re: [uCsimm] Are the uCsimm I/O pins 5V Tolerant?
>> >
>> > Tom Walsh wrote:
>> >
>> > I miskeyed, use the 74LVX family
>> >
>> > TomW
>> >
>> >
>> >> Sam Saprunoff wrote:
>> >>
>> >> > Hello everyone,
>> >> >
>> >> > I was curious... Are the uCsimm I/O pins 5V Tolerant? I have
>> >> > been unable to
>> >> > verify this from searching through the uCsimm manual (Vdd -0.5
>> >> > to +4.0,
>> >> > Vin -0.5 to +7.0), nor the Motorola MC68EZ328 users manual
>> >> > (Vdd/Vin -0.3 to
>> >> > +7.0).
>> >> >
>> >> > Thanks in advance!
>> >>
>> >> No, they are not. You must use a logic family such as the 74LV
>> >> which also provides for voltage level shifting. Nothing in the
>> >> 68EZ328 User Manual says anything either way regarding the
>> >> voltages on the i/o (and other) pins, as it usually happens,
>> >> attaching a voltage higher than that of the the specified supply
>> >> of A processor, generally results in damage to the pin through
>> >> long term use.
>> >>
>> >> I would certainly welcome being wrong on this issue, but as yet,
>> >> I have read no definitive documentation that specifically
>> >> addresses the issue of TTL family interfacing. Rule of thumb? If
>> >> they don't spec it, don't believe it!
>> >>
>> >> Regards,
>> >>
>> >>
>> >> TomW
>> >>
>> >>
>> >>
>> >> --
>> >> Tom Walsh - WN3L - Embedded Systems Consultant - tom over_at mytoys(.)com
>> >> 'www.mytoys.com', 'www.openhardware.net', 'www.cyberiansoftware.com'
>> >> "Windows? No thanks, I have work to do..."
>> >>
>> >>
>> >
>
> John Ananny ananny@interval.comInterval
> Research Corporation (650) 842-6039 This message resent
> by the ucsimm@uclinux.com list server http://www.uClinux.com/

--
Tom Walsh - WN3L - Embedded Systems Consultant - tom over_at mytoys(.)com
'www.mytoys.com', 'www.openhardware.net', 'www.cyberiansoftware.com'
"Windows? No thanks, I have work to do..."

This message resent by the ucsimm@uclinux.com list server http://www.uClinux.com/



This archive was generated by hypermail 2b30 : Sun Apr 07 2002 - 00:01:34 EST