MOSFET as switch

[Cantplaypiano]

I have some 7-segment LED displays which require ~9 volts to run. I want to control them using a TTL digital logic circuit which works perfectly on it's own when 5-volt displays are used. I need a way to power the 9 V displays from 5 V logic via a BCD-to-7seg decoder chip. I anticipate having to use a 9 or 12 V main supply, a voltage regulator to go down to 5 V for the logic, and a MOSFET to take the 5V signal and push the larger voltage through the LED. I can't seem to find a good MOSFET to use though, and if anyone has a suggestion as to what particular transistor to use or a better idea, I would appreciate it.

 

[itsmoked]

Hello Cantplaypiano;

Personally i'd use standard transistors (bipolar) 2N3904 NPN
or 2N3906 PNPs.

But if you really want a FET look at the VQ1004J or the very popular VN2222L.

 

[IRstuff]

Push? How?

Why not simply use an open-collector buffer to drive the segments?

TTFN

 

[Cantplaypiano]

My concern about using a bipolar transistor is that the logic won't push enough current. Granted the signaling current is on the order of microamps, and I have some 2N3904s lying around, I'll see what I can do.

thanks all

 

[itsmoked]

I use 10mA to switch entire displays(multiplexed)with 2N3904's all the time. Unless your LEDs are real pigs, 2N3904's should suffice.

 

[IRstuff]

Well, of course regular logic won't. That's why you need open collect logic.

You should be using a pnp, unless you want to run two supplies. You could run a single supply with npn, but you'd need either a bootstrap capacitor or you'd need a zener clamp on the base drive

TTFN

 

[Cantplaypiano]

My plan was to use a voltage regulator so that I would have a 9 or 12 volt supply and 5 volts regulated off of that, so two power supplies isn't a problem.

The OC buffer is sounding like a better idea now that I've tried breadboarding it with no success. The breadboard setup is pretty clunky though since I don't have a single board large enough to fit the display. I'll look into it and see what I can come up with.

 

[ScottyUK]

Can'tplaypiano,

The VN10KM is a useful small MOSFET in a TO92 package. Costs a lot more than a small signal BJT, and doesn't offer much benefit other than slightly easier drive requirements. Have you considered the ULN200x series of driver ICs? Ideal for interfacing logic to heavier current. From memory I think the full PN of the one I'm thinking of is ULN2004, but I'm at home and my catalogues are at work. I'm pretty sure there are a few other ICs in the series.



----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[itsmoked]

That's a nice part.. use it all the time!

http://rocky.digikey.com/WebLib/Texas Instruments/Web data/ULN2001A-4A,ULQ2003A-4A.pdf

 

[IRstuff]

That wasn't my point. If you run an npn or n-channel MOSFET and expect the emitter or source to be at 9V, the base or collector will need to be higher than 9V. In the case of MOSFET, it could be as much as 3V higher. In the case of a Darlington, you'll need a minimum of two diode drops to drive the base.

If you use p-channel or pnp, then you can use active low to drive the transistor without needing a supply higher than the Ve or Vss of the transistor.

TTFN

 

[ScottyUK]

I sse your point now IR. It depends on whether the 7-seg displays are common anode or common cathode. The OP didn't state which, and I don't know which is the prevalent type these days. If they're common cathode then it should be possible to use a low side switch, in which case the circuit would not need the higher supply rail. If they're common anode then you are right about the drive requirements.

Something else to check is the source/sink capability of the logic family used. TTL can sink more than it can source: about 16mA vs 2mA for 74LS, for example. If you are driving BJTs into saturation this could be significant.



----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[itsmoked]

Cantplaypiano never said whether he has common anode or cathode displays. If the LEDs are common Anode he could easily use just NPNs or ULN2004s driven by the logic to run 9V LED elements.

If the displays are common Cathode you are correct he would need to follow your suggestion.

 

[itsmoked]

Ah Scotty.. you scooped me!

 

[ScottyUK]

ItSmoked,

Dirty minds think alike... oops, that's not right... great minds think alike!



----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[Cantplaypiano]

The displays are common anode, and crave 8-10 V at a max of 25 mA. Using the 74LS192 to sink 16 mA *should* give a bright enough display. After a quick test setup, 9.4 V at 15.2 mA is a little on the dim side, but it would suffice. I'm going to at least 20 mA though, which is very easily visable.

Looking at the datasheet for the ULN2004, it appears to hold alot of promise. It should be extremely convienient considering there are seven pairs on the chip and seven segments in a display.

Using a 12 V supply instead of a 9 V supply isn't a hardship for me, since it will be regulated to 5V for the TTL logic anyhow.

 

[itsmoked]

Just watch the TOTAL current in those 2004s Each channel is a lot but the device has a total limit too.

If you are running the display from a PIC you *could* run it directly from the PICs pins. (same total limit concerns tho)

 

[Yashu]

Check out the TI TPIC6 series open drain power logic... or... even the SN7406/7

 

[Yashu]

or even... TI TPIC2810