Searching \ for '[EE]: level translation' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=level+translation
Search entire site for: 'level translation'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: level translation'
2011\01\27@041739 by Wouter van Ooijen

face picon face
What is the jellybean solution for connecting a 3.3V SPI peripheral to a 5V uC?

I know I could run the uC at 3.3V, or rely on the 3.3V levels to be sufficient to drive the uC and use some resistor-based kludge for the uC-to-peripheral direction.

But I was hoping for a dual-supply chip that could translate for instance 4 unidirectional signals in each direction. Browsing Mouser I find lots of 3.3V max translation chips, I2C level shifters (but those require external pull-ups), and even complete isolators (Si8431 seems ideal for isolating RS485). But no aparent 5V <-> 3.3V solutions. any suggestions?

--
Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2011\01\27@045013 by cdb

flavicon
face


:: What is the jellybean solution for connecting a 3.3V SPI
:: peripheral to a:: 5V uC?

Wouter,
I would go the three sets of MOSFET route plus two resistors per mosfet. One of those dual jobbies made as ballast inverter drivers would do the trick.

As an aside I'm surprised at how cheap RS are for the octal 74LVC4245A bus transceiver  0.49c each minimum order 5!

Colin

--
cdb, spam_OUTcolinTakeThisOuTspambtech-online.co.uk on 27/01/2011
Web presence: http://www.btech-online.co.uk   Hosted by:  http://www.justhost.com.au
 

2011\01\27@055154 by Wouter van Ooijen

face picon face
> I would go the three sets of MOSFET route plus two resistors per mosfet.
> One of those dual jobbies made as ballast inverter drivers would do the
> trick.

I know that solution, but I prefer one chip to a bunch of fets plus resistors.

> As an aside I'm surprised at how cheap RS are for the octal 74LVC4245A bus
> transceiver  0.49c each minimum order 5!

Now it would be nice if there were a similar chip with 4 in each direction, fixed direction would be OK.

--
Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2011\01\27@060804 by Geo

picon face
Wouter van Ooijen wrote:

> Now it would be nice if there were a similar chip with 4 in each
> direction, fixed direction would be OK.

Would the 74LVC241 work?

George Smith

2011\01\27@073656 by Wouter van Ooijen

face picon face
On 27/1/2011 12:07 PM, Geo wrote:
> Wouter van Ooijen wrote:
>
>> Now it would be nice if there were a similar chip with 4 in each
>> direction, fixed direction would be OK.
>
> Would the 74LVC241 work?

nice chip! solves the 5 - > 3v3 part.

why didn't I even see it when searching? probably because it isn't called a level shifter :(

--
Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2011\01\27@080519 by peter green

flavicon
face
Wouter van Ooijen wrote:
> On 27/1/2011 12:07 PM, Geo wrote:
>  
>> Wouter van Ooijen wrote:
>>
>>    
>>> Now it would be nice if there were a similar chip with 4 in each
>>> direction, fixed direction would be OK.
>>>      
>> Would the 74LVC241 work?
>>    
>
> nice chip! solves the 5 - > 3v3 part.
>   Change LVC to HCT in that part number to get a chip that will translate the other way.
> why didn't I even see it when searching? probably because it isn't
> called a level shifter :(
>   mmm, it's not specifically designed as a level shifter, it's just a 3.3V logic chip with 5V tolerant inputs.

Similarly HCT is a 5V logic family with 3.3V compatible inputs (they were originally designed to be TTL compatible inputs but the thresholds work fine for 3.3V cmos signals as well).

You can use other chips from those families too allowing you to do logic and level shifting at the same time

2011\01\27@081800 by RussellMc

face picon face
Going 3.3 -> 5V and just for fun

Q1 = NPN "Logic transistor" with internal base resistor.
5V in enable pullup (or use an R)

V_3V3_Out - Q1e
3V3+ - Q1b
Q1c - V5_in

Single eg SOT23 or smaller per channel.

_________

Bidirectional MAYBE MOSFET version.

MOSFET as above.
No rg needed but Vth < 3V.

For 3V -> 5V works as above.

Now make 3V = input, 5V = output, same connection as before.
Input internal pullup or R.

V5_out high - Vin_3V floats pulled high, all well.
V5_out low -> pulls V3In low via FET body diode.
FET is turned on a Vg > Vs but as Vd < Vg all is well.

Getting speed may be "problematical".
Sure to be bugs.
Pullups annoying.
Looks like fun.


2011\01\27@081916 by Per Linne

flavicon
face
Hi Wouter,

I have used the MAX3377 (there are several variants) in a couple of one off:s, to be on the safe side.

PerL

{Original Message removed}

2011\01\27@085943 by Michael Watterson

face picon face
On 27/01/2011 12:36, Wouter van Ooijen wrote:
> On 27/1/2011 12:07 PM, Geo wrote:
>> Wouter van Ooijen wrote:
>>
>>> Now it would be nice if there were a similar chip with 4 in each
>>> direction, fixed direction would be OK.
>> Would the 74LVC241 work?
> nice chip! solves the 5 ->  3v3 part.
>
> why didn't I even see it when searching? probably because it isn't
> called a level shifter :(
>
Old 74LS chips in theory (many of them) have logic 1 and 0 output levels that are more compatible with 3.3V logic than 5V cmos logic.
I think the CMOS HCT series is TTL in compatible and 5V out so would do reverse?

LS TTL levels are not 5V for logic "1" more like 2.7V?

I'd need to re-read the old data sheets :-

2011\01\27@205757 by Bob Axtell

face picon face
On 1/27/2011 3:51 AM, Wouter van Ooijen wrote:
>> I would go the three sets of MOSFET route plus two resistors per mosfet.
>> One of those dual jobbies made as ballast inverter drivers would do the
>> trick.
> I know that solution, but I prefer one chip to a bunch of fets plus
> resistors.
>
>> As an aside I'm surprised at how cheap RS are for the octal 74LVC4245A bus
>> transceiver  0.49c each minimum order 5!
> Now it would be nice if there were a similar chip with 4 in each
> direction, fixed direction would be OK.
>
There is: LM239, quad comparator, at $0.25 a device. These are reasonably fast. I am using one now
for an MMC card interface.

--Bob A

2011\01\28@024415 by Wouter van Ooijen

face picon face
> There is: LM239, quad comparator, at $0.25 a device. These are
> reasonably fast. I am using one now
> for an MMC card interface.

That's four in one device, not four in each direction, and I guess it requires pull-up resistors on the outputs, and a resistor divider to provide the reference voltage?


--
Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu

2011\01\28@190758 by Sergey Dryga

flavicon
face
Wouter van Ooijen <wouter <at> voti.nl> writes:

>
> What is the jellybean solution for connecting a 3.3V SPI peripheral to a
> 5V uC?

Something like PCA9509 maybe? It's I2C, but they probably have chips specific
for SPI as well.  Available at Mouser.com

Sergey Dryga
http://beaglerobotics.com

More... (looser matching)
- Last day of these posts
- In 2011 , 2012 only
- Today
- New search...