Low Voltage Relay

[Kaluriel]

I need to power an OLED display with 14V, however it must be powered up after some commands have been sent. However my logic is running on 2.5V which isn't enough to power any coil relays I can find.
The maximum voltage on the open-drain pins of the PIC i'm using only allows a maximum of 5V, so I was wondering if there is any other way to switch on another circuit at low voltages, else I'll just replace the PIC I'm using.

[bilko]

Try googling 'logic level mosfet'

[johnmx]

Use a small N-mosfet, preferably with low rds(on) and low gate-threshold voltage (e.g. Si2312BDS). Connect the gate to the pin of the PIC with a small resistor, 100Ohm will do. For power MOSFETs you should use a mosfet driver instead of a resistor.

[Tony R]

maybe you shouldn't use a coil. they make solid state relays. I'm sure you can find one that will work for what you need.

[Kaluriel]

thanks, I'll look into all of these

[Mechatrommer]

SSR that i've seen is only for high volt/mains. any pointers to lo level cmos/mcu ssr stuff?

[vk6zgo]

Opto-coupler driven from the PIC output.
Use the transistor inbuilt into the opto to drive a discrete transistor to operate the relay.---complete isolation.
The relay supply can be the same 14volt supply you are using for the display.
VK6ZGO

[Zero999]

Why not just use a traditional BJT?

It seems like the most cost effective option.

[johnmx]

Why not just use a traditional BJT?

It seems like the most cost effective option.

Just one NPN will raise the Ground level of the LCD by 0.2V. But he can use 2 transistors, one NPN for the input stage and one PNP to drive the LCD.
The power loss will be higher compared to just one N-MOSFET.

[Zero999]

Just one NPN will raise the Ground level of the LCD by 0.2V.

Why does that matter? 0.2V is 0.143% of the supply voltage.

But he can use 2 transistors, one NPN for the input stage and one PNP to drive the LCD.

Yes, that would work assuming he needs to switch the positive supply.

The power loss will be higher compared to just one N-MOSFET.

An N-MOSFET can't be used to switch the positive supply without increasing the gate voltage above the supply.

For lower voltage losses, he could use a P-channel MOSFET and an NPN BJT. The MOSFET doesn't need to be logic level since the supply voltage is 14V which will be more than enough for a standard MOSFET.

Another option is to use an SCR but that's only acceptable if one shot operation is desired, i.e. the display doesn't need to be turned off.

[johnmx]

Just one NPN will raise the Ground level of the LCD by 0.2V.

Why does that matter? 0.2V is 0.143% of the supply voltage.

It’s 1.43% of 14V, but if the LCD logic is powered with the same 2.5V and this supply shares the same ground pin of the 14V, then 0.2V means 8% of the supply.

The power loss will be higher compared to just one N-MOSFET.

An N-MOSFET can't be used to switch the positive supply without increasing the gate voltage above the supply.

Of course, the N-MOSFET is to drive the low-side. If you read my first post I said that he could use a low gate-threshold voltage MOSFET connected directly to the pin of the PIC. You know that this is only possible in that situation.

[Zero999]

It’s 1.43% of 14V

Yes, you're right, i make a mistake with my decimal place, though 0.2V of 14V is still negligible and you can get BJTs with lower saturation losses than that, especially if you de-rate.

but if the LCD logic is powered with the same 2.5V and this supply shares the same ground pin of the 14V, then 0.2V means 8% of the supply.

I'm not sure if you've through that through. If the display supply shares a ground as the logic then you can't switch the low side and control it from 2.5V. When the display is off, it's ground suddenly jumps to 14V, feeding power into all of the logic outputs connected to its inputs, cooking the protection diodes, so the voltage drop being 8% of the logic supply is irrelevant.

[johnmx]

..If the display supply shares a ground as the logic then you can't switch the low side and control it from 2.5V. When the display is off, it's ground suddenly jumps to 14V, feeding power into all of the logic outputs connected to its inputs, cooking the protection diodes...

Well seen. I missed that.

[mikeselectricstuff]

Why would you use a relay..?
P-channel FET switching the high side to the display, NPN transistor pulling the MOSFET gate low to turn on, pullup to hold gate off. Easy.

[vk6zgo]

I'm not sure if my opto idea would work,the PIC might not be able to source enough current to turn the LED part on.

An alternative would be a couple of  BC108s or similar connected as a Darlington pair with the second one operating the relay.
The input impedance of a Darlington is high,as the first section is really an emitter follower.You can put a fairly high value resistor(I used 82k Ohm ) between the PIC output & the base of the first transistor to reduce loading even further.

If you use a real relay,all the problems about earth level shifting,etc go away.

Grab a few components & try a few of everybody's suggested configurations,or model them,it's a lot quicker than asking people on a forum.

VK6ZGO

[Zero999]

Why would you use a relay..?
P-channel FET switching the high side to the display, NPN transistor pulling the MOSFET gate low to turn on, pullup to hold gate off. Easy.

Seems like a sensible idea to me.

Are the turn on/off times important?

If all that's needed is low speed switching, a voltage doubler circuit can be used to shift the gate of an N-MOSFET above the positive supply. It can be done with minimal extra components, assuming you're using an MCU which can easily send a square wave to an output pin.

[mikeselectricstuff]

Why is everyone making such a meal out of such a simple problem?
This is all you need :

[Zero999]

Why is everyone making such a meal out of such a simple problem?

I was tempted to draw that but I thought your description was so good it wasn't necessary.

I don't see the need for a P-MOSFET when PNP BJTs are much cheaper.

[mikeselectricstuff]

Why is everyone making such a meal out of such a simple problem?

I was tempted to draw that but I thought your description was so good it wasn't necessary.

I don't see the need for a P-MOSFET when PNP BJTs are much cheaper.

Depends how much current the OLED draws and how fussy it is about supply voltage. A suitable SOT-23 P-MOSFET is about £0.10 so not a big deal compared to the cost of the OLED.
Another advantage is where it is necessary to control risetime, as this  could be done by adding a resistor in series with the gate.

Of course chances are the 14V supply is coming from a regulator of some sort which may have a shutdown facility which could be used for switching.

[Lawsen]

You could try a small optoelectronic relay.  Motorola has AAC142/43 or AAC242/43 in an 8 pins DIP.  The voltage turn on is 10V and turn off at 16 V.  That might be too high for your application.  It is packaged in a small 8 pins DIP.  . 

http://loci.asicadv.com/aac14x24x/?utm_source=supplyFrame&utm_medium=Ads

In electronics trade school, I have seen some larger solid state relay, but the controller voltage is as high as an electromechanical relay for being compatible and easier to sell them. 

Lawsen