1. Home
  2. Electronics Design
  3. 74LS47 substitution

74LS47 substitution

Is there a CMOS-type substitute for the 74LS47 with "open collector" type outputs suitable for sinking around 20mA?

That uses less quiescent power than the 47's ~10mA. Supply voltage either 5 or 3.3 volts, depending. It doesn't really matter if the chip itself does a conversion from BCD to the appropriate outputs, as that mapping can be done in software, but the chip would need to have at least a 3 bit address space.

The '47 has an input that's suitable for dynamic dimming of the loads by applying PWM, it would be cool if the replacement had something like that.

I think JL posted something along these lines in response to another question of mine, but I can't find the thread.

Reply to
bitrex
Loading thread data ...

A TPIC6595 might work. The data sheet says it needs 5 volts, but there is an appnote that says it's OK at 3.3. We use them at 3.3.

The input is SPI serial, and the /G pin can do dimming.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

Thanks! Will look into it...

--


----Android NewsGroup Reader---- 
http://usenet.sinaapp.com/
Reply to
bitrex

The CD74AC238 appears to be rated for 50 mA through a pin, 200 mA continuous to ground, at 1.5 to 5.5 volts supply, so that might also be a possibility.

The edge case would be the "8"...

--


----Android NewsGroup Reader---- 
http://usenet.sinaapp.com/
Reply to
bitrex

Do you need open drain? HC or AC logic will sink 20 mA, probably better with Vcc=5.

The 6595 is a beast, designed for a lot of current.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

I don't think so. The 74LS47 outputs don't pull down much lower than 0.3 volts.

While it looks like most HC logic can sink 20 mA, the problem is it seems most are rated for only 50mA continuous through GND, and the SOIC-16 packages are rated for only 500mW continuous dissipation.

The CD74AC238 can sink 200mA to ground, and is rated for 1.1 watts dissipation at 30 degrees, so even if I have to derate it a bit for

140mA passing through the device when all segments are lit, it seems it should be okay.
Reply to
bitrex

How about the ULN2001 family, 500 mA / pin?

--

-TV
Reply to
Tauno Voipio

can sink about 20 mA/pin with a total of 200 mA for VCC and GND. Of course the outputs are not open collector. If you need OC and more

multiplexing.

--
Dipl.-Inform(FH) Peter Heitzer, peter.heitzer@rz.uni-regensburg.de
Reply to
Peter Heitzer

Yeah, there's going to be a uC involved one way or the other...I'm trying to figure if it would be better to just use an AVR with more pins vs an 8 pin AVR and an external switch IC.

An ATTiny85 that costs under a dollar in quantity plus an external switch that costs 25 cents in quantity is a better value than something like a Mega168 that costs $1.65 in quantity, as the board size would be pretty much fixed by larger components either way.

There are other weird AVR devices that seem to use a "Tiny" core but have more pins, but I'm taking advantage of the Arduino toolchain and AFAIK they're not supported, and I don't really want to take the time to write an implementation.

Reply to
bitrex

It's a good chip, it doesn't look like it draws any current whatsoever unless the Darlington outputs are switched on...only 25 cents in quantities of 100 surface mount.

But it needs one address line for each output, and I was hoping to cut costs by using an el cheapo 8 pin AVR and an el cheapo driver, instead of shelling out for an AVR with more pins but also more features and code space which I don't need for this. If I wanted to use 7 pins it would make more sense to just use a uC with more pins, as a poster mentions below I think even the AVRs in SMT can handle 20mA per pin and

200mA to ground (though driving the outputs of the display directly from the processor close to the limits makes me feel a little uncomfortable.)
Reply to
bitrex

The other issue with using the uC pins to drive the display directly is that there aren't enough hardware PWM channels to dim every segment common anode. If the display I had to use was common cathode I could simply implement dimming by using an external pass transistor on the cathode and PWMing that to ground and driving the segments with a "high" logic voltage, but unfortunately I don't have that option here, so I think I'd have to implement some scheme using a filter and regulator on the supply and manipulating its reference, but it would likely have to be an LDO, which would probably cost around the same as an appropriate driver IC in a SOIC-16.

Reply to
bitrex

Actually, if I decided to go that route I could probably just chop the supply with PWM and a two transistor switch like so:

formatting link

Reply to
bitrex

Oops, I made a mistake in my power dissipation calculation, what a dummy. It's only the voltage drop across the pass device that matters, so even a 500mW dissipation should be good.

Reply to
bitrex

It looks like the best "bang for the buck" right now is to simply use an AVR with a "Tiny" core that has 12 IO pins instead of 6, and then PWM the supply with a two transistor switch.

I didn't realize that AVRs could sink as much current to/from the supplies as they appear to be able to.

The ATTiny84 has two PWM channels and is just over a dollar at Mouser, a PNP and NPN for the switch is basically free.

Reply to
bitrex

The only supported "Tiny" device is the Tiny85, but the Tiny2313 is not very different to the Tiny85.

formatting link
mentions the 2313. So it should be possible to use the 2313 with the toolchain. The 2313 costs about the same as the 85 but gives you 18 usable pins vs 6 pins on the Tiny85.

--
Dipl.-Inform(FH) Peter Heitzer, peter.heitzer@rz.uni-regensburg.de
Reply to
Peter Heitzer

Looks like setting it up for the Tiny84 is straightforward these days:

formatting link

The Tiny84 only has 12 pins, but even that is more than I need. I'd prefer the larger code space, 8k vs 2k for the 2313. They're both about the same price in quantity at Mouser.

Reply to
bitrex

That makes sense. If you program the uC in C 2 KiB can be used up easyly if you happen to use "printf" or its friends. You also could use linear code more easyly if you need speed for some tasks.

--
Dipl.-Inform(FH) Peter Heitzer, peter.heitzer@rz.uni-regensburg.de
Reply to
Peter Heitzer

There are about a million I2C/SPI LED drivers out there that have built in PWM dimming. Have you looked at any?

--

Rick C
Reply to
rickman

I have a slight pathology in that I like using C++, even for "small" projects. avr-g++ compiles pretty efficient code, so long as one avoids using a ton of "virtual" methods, and initializes all memory required for the system at startup and then doesn't use the "new" operator to do anything but place or copy objects onto pre-allocated memory.

I think the moderate increase in code size is worth the advantages gained in portability/maintainability.

Reply to
bitrex

I think the OP is looking for a part costing less than a dollar. The drivers I found are rather costly.

--
Dipl.-Inform(FH) Peter Heitzer, peter.heitzer@rz.uni-regensburg.de
Reply to
Peter Heitzer

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.