Quad push pull driver with FETs as replacement for L293

[0xdeadbeef]

I'm currently tinkering a semi-private sensor simulator with four channels and wanted to add a push pull driver to the output just to create a 5V output signal and also to be able to drive small loads (like 100mA) per channel.
I didn't put much thought into this and assumed using a standard L293 (NE) push-pull driver would be a good idea.

Now the prototype is working but I realized I should have read the L293 datasheet better, as the output level is more like 0.5 to 5V (also a bit below that) instead of 0 to 5V. While this is not really a big issue, I would rather have a low level closer to 0V. Besides, according to the datasheet , the "low" output voltage could be as high as 1.8V at higher currents, which would be of course too high.

Anyway, as I assume that the relatively high "low" voltage is caused by the fact that the driver uses bipolar transistors and I would assume that a push-pull driver using FETs should have a low output voltage. So I looked for a quad push pull drivers with (internal) MOSFETs, but all I found way the Microchip TC4467/TC4468/TC4469. Which would be ok for my purposes, but it's a bit exotic and I don't really need the dual inputs. I could also find a lot of 8fold bus drivers, but they can only drive something like 20mA per channel while my target was something closer to 100mA.

The problem might be also that when searching for "quad push pull driver fet", you find a lot of drivers for  FETs, but not drivers with  FETs.

So any hints for a cheap quad push-pull driver with internal FETs that is easily available in some nice DIP package and can drive 100mA per channel? Or is the TC446x really already my best choice?

[SeanB]

Even mosfet output stages are going to struggle to reach 0V at any significant current. You might look at running it off a split rail, like 7V and -2V, which will give a full range easily. Only works if you have the supplies available, or a mains supply, or a switcher where you can add either a few extra turns of wire and a diode/capacitor for the negative rail and adjust the output voltage a little.

[0xdeadbeef]

But I'm not talking about significant currents, but something in the <= 100mA range.
For a MOSFET based push-pull driver, the low voltage should be more or less only based on the RDS in on state. So it should be pretty much exactly 0V without load and even at 100mA, the voltage should be in the few mV range (e.g. 100e-3 [A]*0.5 [Ohm] = 0.05 [V]).

[TerminalJack505]

The requirement that the package be DIP is the real trick.  The only part that I can think of is the TC427.  This is a two-channel driver (so you'd need two) with rail-to-rail output.  They call it a MOSFET driver since it is ideal for driving MOSFET gates but you don't have to use it for that.

From the datasheet...

1.5A Dual High-Speed Power MOSFET Drivers

The TC426/TC427/TC428 are dual CMOS high-speed drivers. A TTL/CMOS input voltage level is translated into a rail-to-rail output voltage level swing. The CMOS output is within 25 mV of ground or positive supply.

[0xdeadbeef]

Hm, yeah Microchip again. I looked at their page and set the filter only on the quad outputs but I saw they had quite a few dual outputs as well.
Then again, there doesn't seem to be much difference in the low output voltage between the dual and quad drivers.
Funny thing is that in the TC446x datasheet, they give the high voltage as VDD-0.025V and the low output voltage as 0.15V, but only because they use a load current of 100µA for the highside and 10mA for the lowside.
In the TC42x datasheet no load currents are given, but the output resistance seems to be more or less the same, so I guess they just gave both numbers for 100µA.
By the way, when looking at the output resistance diagram, it looks as if the resistance for the low side will be something like 18Ohm with a 5V supply, so that's more than 10 times higher than I expected, but yeah well. At least in open load condition, I should get the the ground level and I wanted to limit the output current anyway with a 50Ohm resistor.

Still I somehow can't believe that none of the big brand like TI or IR has a push-pull driver with lower output resistance. But I guess I can live with either solution.