RS485 tranceivers?

Yep - then why wouldn't they make tranceivers this range too? I mean the differential drivers can be made as 3.3V and 5V...

/A

most logic has this range.

Actually, most rs422/485 drivers use these logic levels.

Well, I think the rs (or actually EIA) 422 specification specifies what the complementary (they're usually NOT differential) voltage levels are supposed to be.

As far as 3.3 and 5 V levels go, almost any device designed to work at a nominal 3.3V will be specified to work down to 3V (usually the fine print will say 3.3V +/- 10%). And any CMOS device which works at 3.3V AND

5V will work everywhere in between, as long as it is interfaced to devices connected to the same rail.

Maybe you could post a specific part number and a little more information about your application. Then you might get some meaningful feedback about whether your application is likely to work.

You could also try asking the vendor. Some of them actually have good applications engineers who get paid to answer your questions.

--Mac

Yeah - but they're either specified for 3.3V OR 5V (it's the supply I'm referring to). I'm wondering why the device I need (3-5V supply range) does not exsist....

/A

The problem is they are EITHER 3.3V or 5V types.

Well, at the moment the MAX3072E is the more promising. Similar types from TI and a lot of pincompatible 5V types have also been investigated

The application involves a lot of units supplied through a ribbon cable. Thus, I'd like to save the power, cost and complexity of having to output excess power and dissipating this in a LDO. If I had 3-5.5V types the say

0-1V drop over the cable could easily be tolerated....

Cheers, Anders

So what you are really saying is that you want a chip manufacturer to design a device capable of operating over a 3.0 - 5.0 volt dc supply range without compromising your slew rate or other operating specifications, just so that you can have the luxury of a low system cost design?

Supplying the receiver dc operating voltage over a long buss is not good design practice, UNLESS you provide local regulation/filtering/decoupling at every receiver drop-off point along the buss. This is just good design practice and it should not be compromised. You are trying to do it on the cheap it seems, and IC manufacturers don't usually go to such lengths to cover the inadequacies inherent in low cost designs. It would mean they would have to include on-chip regulation thus pushing the design complexity and production cost above alternative devices from other manufacturers, thus making their devices unviable for most oem equipment manufacturers.

Please let them compromise slewrate! As long it's nog gonna be fast... ;-)

Yessss! This is the way of the business...

There's a difference between filtering/decoupling and linear regulation (or even switching - which would be the "proper" solution)

You'd be amazed! Look at all the new products comming out only to lower cost....

I think you missed my point. My uC goes 1.8-5.5V. Standard CMOS logic 3-6+V. I just don't understand the problems involved in making RS485 drivers for

3-5.5V. I mean they can make 3.3V types with 6V maximum rating. And 5.5V types....

Enough said. Yes, its gonna be cheap - so kick me! For now I'll have to design in a LDO and waste some cents and some power from the drop. Bad for me (complexity, stock, price), bad for my customers (price, power) and bad for the environment (power, components)!

Be happy! Cheers, Anders

Hmm. I wonder if you could just use a 5V type at 3.3 V? What baud rate are you running at? If it is slow I bet it would work, particularly if you are not going over long cables, and are only interacting with devices of your own design. I would definitely try asking an applications engineer from the relevant company.

Good luck.

--Mac