Switching 12v high side

Thread Starter

Wolframore

Joined Jan 21, 2019
2,609
Does anyone have a trick to switching 12vdc high-side using a P channel from a 3.3 or 5v logic signal without adding another switching layer? I have some weight and size requirements and trying to switch multiple channels. Low side would work, but not best solution since then you have hot rail. Looking for the least amount of components and low current.
 

Thread Starter

Wolframore

Joined Jan 21, 2019
2,609
I have little control over the current switched but anywhere from 10mA to 100mA in most usage but I’m sure someone will try to switch much higher. @bertus that is a great driver... I’ll look into it. It might be a good buffer if I need more.

Thanks @danadak ill look into those but I have a feeling they will break the board budget.

The circuits I’m working on currently need to be small as possible and strict weight requirements so as easy it would be to add a low side to switch the highside now my board grows...

Thank you I think I know what I have to find!
 

Thread Starter

Wolframore

Joined Jan 21, 2019
2,609
@bertus is there a reason why most of these chips are designed around transistors and darlington instead of CMOS or mosfets? I can't see why they couldn't design a chip that would pass 1 or 2 amps total between 4-8 channels based on FET...
 

Thread Starter

Wolframore

Joined Jan 21, 2019
2,609
Ugh it’s the dumb things folks do. Trying to dumb proof it. But I’m finding highside switching in this situation challenging to keep it small and compact.
 

Thread Starter

Wolframore

Joined Jan 21, 2019
2,609
It’s a series of boards I’m doing for controlling LEDs, we had a guy plug in power to an output and take everything out. First one was CMOS but the next ones will be uC for more complex stuff. Everything needs to be small as possible, efficient and basically free. So far I’m just doing a lot of work but not seeing any break even. It’s keeping me productive and it’s been fun but this market has little or no potential.

I’ve ordered a test board with low side switching so I can get back to programming.

I’m considering going full open source in the future but not just yet.
 

TeeKay6

Joined Apr 20, 2019
573
It’s a series of boards I’m doing for controlling LEDs, we had a guy plug in power to an output and take everything out. First one was CMOS but the next ones will be uC for more complex stuff. Everything needs to be small as possible, efficient and basically free. So far I’m just doing a lot of work but not seeing any break even. It’s keeping me productive and it’s been fun but this market has little or no potential.

I’ve ordered a test board with low side switching so I can get back to programming.

I’m considering going full open source in the future but not just yet.
@Wolframore
Here's the extra switch you didn't want...
View attachment 184848
 

Thread Starter

Wolframore

Joined Jan 21, 2019
2,609
Thanks TeeKay, it a clever way of making a non inverting transistor (common base) to switch the high side if I could get that in a sot23 package I’m in. I was thinking N channel to switch the P channel high side. That way invert it to high being on which always makes it easier for me to keep track of. But the trade off is all those resistors and 2 transistors taking all that real estate.
 

Thread Starter

Wolframore

Joined Jan 21, 2019
2,609
The second examples might work. I have to start adding resistors and see how much 4 channels grows. I was using 0805 but can go to 0603. I don’t want to go smaller populating by hand.
 

Thread Starter

Wolframore

Joined Jan 21, 2019
2,609
@Bordodynov thank you, I will research simple load switches. This is exactly what I was looking for. Your example uses a pnp I was hoping for mosfet but I haven't started to calculate the expected switching costs. It looks like they can get very complex and well designed but for my need I need small and simple and your suggestion is a very good one.
 

Thread Starter

Wolframore

Joined Jan 21, 2019
2,609
Ok I really want the load switch based on matched pair of mosfets. I expect PWM in the 1kHz or slower so I can have low switch current well under a mA. I wish they made a load switch using PFET and NFET. I don’t even care about Rds in this situation.
 

TeeKay6

Joined Apr 20, 2019
573
Ok I really want the load switch based on matched pair of mosfets. I expect PWM in the 1kHz or slower so I can have low switch current well under a mA. I wish they made a load switch using PFET and NFET. I don’t even care about Rds in this situation.
@Wolframore
Have you looked at gate driver chips? Logic control. Moderate Ron. High input voltage limits.
 
Top