PCF8574T

I want to extend the digital outputs of arduino nano with pcf8574t,
and then have the possibility to activate and deactivate relays.
How can this idea be implemented? :slight_smile: :slight_smile: :slight_smile:

Might be better to use a TPIC6B595. It requires three pins to control, but you can then chain more using the same three pins.

The PCF8574 is fine as well. Here is a nice library for it: https://github.com/RobTillaart/Arduino/tree/master/libraries/PCF8574

The advantage of the TPIC 595 variants is that they have much higher output current drive, sufficient to drive a relay directly instead of needing another IC or transistor to drive the relay coil. They're much like a 595 with the outputs tied to a ULN2803. Check the current required by the relay to determine if you need the TPIC6A595 or can make do with TPIC6B595. Do not forget the diodes across the coil to clamp the back emf when you turn the relay off.

What relays. Maybe OP has relay modules. In which case all the advice could be wrong.

OP, read the "how to post" sticky before asking questions, so we don't have to guess. Leo..

DrAzzy: The advantage of the TPIC 595 variants is that they have much higher output current drive, sufficient to drive a relay directly instead of needing another IC or transistor to drive the relay coil. They're much like a 595 with the outputs tied to a ULN2803.

Except that the ULN2x03 cannot supply as much current to multiple outputs at once, and loses something like 1½ V which is quite significant when using 5 V relays.

FETs rule! :sunglasses:

DrAzzy: Check the current required by the relay to determine if you need the TPIC6A595 or can make do with TPIC6B595.

However it is, the TPIC6B595 is much more readily available on eBay and such.

DrAzzy: Do not forget the diodes across the coil to clamp the back EMF when you turn the relay off.

Not required when using the TPIC chips - that functionality is built-in.


Following previous discussions, it occurs to me to muse how they make the TPICs. Clearly the "A" version is the premium part and it makes no sense to deliberately make ones with lesser specification. Presumably, batches are made and tested for Rds at given test voltages. They would then be marked according to that specification (it is the only one that actually matters), but if the manufacturing process is "up to scratch", they would likely actually mark down the lower end of those still qualifying for "A" grade, in order to match market demand.

Wouldn't they?

Paul__B: Following previous discussions, it occurs to me to muse how they make the TPICs. Clearly the "A" version is the premium part and it makes no sense to deliberately make ones with lesser specification. Presumably, batches are made and tested for Rds at given test voltages. They would then be marked according to that specification (it is the only one that actually matters), but if the manufacturing process is "up to scratch", they would likely actually mark down the lower end of those still qualifying for "A" grade, in order to match market demand.

Wouldn't they?

I think it's more likely that the TPIC6B's are deliberately made to a lower specification - with a smaller (hence die is smaller, hence cheaper) MOSFET. This is a dead simple device and it's hard to imagine that their process control is that lousy that they're getting the low spec modules that way.

So that would be the 6Bs and 6Cs then as a second run.

I see your point - a 4 to 1 range would reasonably encompass two die sizes, but surely not more. :astonished:

Might be hard to manufacture to exact specs, so there must be some testing/selecting going on. But who cares. Just use the part that's easy to get and suits your needs. I mostly use the TPIC6C59*6* in smd pakage. Leo..