Has anyone seen some 1.8V- translators that
come in a human solderable package? I want
to extend the SPI bus from the BB and use
some GPIO. But the proto at least, needs to
be assembled by hand.
Regards
cww
Wed, 24 Mar 2010, curt wuollet wrote:
Has anyone seen some 1.8V- translators that
come in a human solderable package?
Define 'human solderable' 
I have been using MAX3378EEUD which can be quite easily soldered manually.
I'am using the TXS0102 and TXS0104 assembled by hand on an home etched
board.
If speed is not an issue, you will find very useful level translators
in the venerable CD4000 CMOS family and its many successors. They come
in DIP packages and cost next to nothing,
siñ
Antti Seppanen wrote:
Wed, 24 Mar 2010, curt wuollet wrote:
Has anyone seen some 1.8V- translators that
come in a human solderable package?Define 'human solderable'
I have been using MAX3378EEUD which can be quite easily soldered manually.
Hi Antti
I guess what I should have said is: "Solderable by a typical automation
guy with the ratty old Weller iron in the toolbox". That's my target
audience. But, it is encouraging that someone considers these tiny beasts
hand solderable. My shiny new design, so far, uses all through hole parts
for this reason. But, I haven't seen anything much easier than TSSOP. I
would hate to make a kludge like soldering these to dip adapters. I need
SPI, reset and 8 CS which would be fine with the MAX chips. but replacement
on the factory floor might be an issue.
Regards
cww
I guess what I should have said is: "Solderable by a typical automation
guy with the ratty old Weller iron in the toolbox". That's my target
audience. But, it is encouraging that someone considers these tiny beasts
hand solderable.
Prior to three weeks ago, I would have said that SMT parts were out of the
question for a hand-soldered board. However, after actually trying, I found
that TSSOPs are a cinch and I can imagine that even QFNs are quite doable. At
this point, I intend of using TSSOPs over DIPs in future designs. They can be
substantially smaller and yet there is little to no added difficulty in
fabrication.
My shiny new design, so far, uses all through hole parts
for this reason. But, I haven't seen anything much easier than TSSOP. I
would hate to make a kludge like soldering these to dip adapters. I need
SPI, reset and 8 CS which would be fine with the MAX chips. but replacement
on the factory floor might be an issue.
While TSSOPs are easy to solder initially, I can't speak for ease of
replacement. Desoldering anything with more than 10 pins is tricky with a
standard iron, IMHO. But, that being said, why do you expect the level shifters
will need replacement? If the chance of incorrect voltages being applied is
that high, you probably ought to have some protective measures in the circuit
design.
Cheers,
- Ben
Has anyone seen some 1.8V- translators that come in a human solderable package? I want
I’m using Sparkfun BOB-08745 (but has only two translators per item; so I soldered 5 of them to have ten GPIO 1.8V converted to 5V).
I have been using MAX3378EEUD which can be quite easily soldered manually.
I’am using the TXS0102 and TXS0104 assembled by hand on an home etched
If speed is not an issue, you will find very useful level translators in the venerable CD4000 CMOS family and its many successors. They come
("Oh, if only the XM had a dozen of 3.3V and a dozen of 5V translators aboard...!") 
Sorry, no level translators on the -xM board. We added so much other stuff, there just wasn’t any room.
Gerald
Ben Gamari wrote:
I guess what I should have said is: "Solderable by a typical automation
guy with the ratty old Weller iron in the toolbox". That's my target
audience. But, it is encouraging that someone considers these tiny beasts
hand solderable.Prior to three weeks ago, I would have said that SMT parts were out of the
question for a hand-soldered board. However, after actually trying, I found
that TSSOPs are a cinch and I can imagine that even QFNs are quite doable. At
this point, I intend of using TSSOPs over DIPs in future designs. They can be
substantially smaller and yet there is little to no added difficulty in
fabrication.My shiny new design, so far, uses all through hole parts
for this reason. But, I haven't seen anything much easier than TSSOP. I
would hate to make a kludge like soldering these to dip adapters. I need
SPI, reset and 8 CS which would be fine with the MAX chips. but replacement
on the factory floor might be an issue.While TSSOPs are easy to solder initially, I can't speak for ease of
replacement. Desoldering anything with more than 10 pins is tricky with a
standard iron, IMHO. But, that being said, why do you expect the level shifters
will need replacement? If the chance of incorrect voltages being applied is
that high, you probably ought to have some protective measures in the circuit
design.Cheers,
- Ben
Hi Ben
The application is going to resemble a PLC with
cards that are plugged and unplugged, hopefully
with the power off. But, having been in maintenance
as a PLC tech, I have observed that some of the
respondents to a problem are not very particular
in that regard. There are a few things I can do
but, complete idiot proofing is not possible as
there seems to be an infinite variety of idiots.
I have a lively discussion about the project on
the control.com site and one of the things that
comes up frequently is that the equipment should
be repairable in house. The field of industrial
automation is one of the last bastions of hyper
proprietary zeal. Where companies will take a
serial card that we would pay maybe $10 for,
solder a unique special connector on it and sell
it for $300 as the only way to talk to their
equipment. So one of the many goals of this
project is for it to be as non-special as it
can be. And also to demonstrate that there really
is nothing magical about PLCs, justifying their
lofty pricing. This will be an _extremely_ powerful
PLC, well beyond it's peers and the cost will be
very low. It will be far more interoperable, with
standard interfaces and protocols. And it will
run Linux. And the tools will run on Linux and
my working life will improve dramatically over
dll hell and forced obsolescence. Oops, sorry:^)
Regards
cww
That's OK, I guess, but that would be the best
place for them as far as signal integrity goes.
I don't know how far you can wire 1.8V signals
before noise is a problem. I'm hoping for a
couple inches. I broke down and tore up tracks
and designed in the MAX translators. I'm sure
I'll hear about the soldering:^) Thanks to all.
Regards
cww
Gerald Coley wrote:
I'm by no means an expert, but even considering the multitudes of stupidity
that exist on this planet, it still seems like a few Zeners would go a long way
in preventing smoking level translators. Just a thought.