Transistor Needed

On a sunny day (Thu, 18 Mar 2021 00:21:51 -0700 (PDT)) it happened " snipped-for-privacy@gmail.com" snipped-for-privacy@gmail.com wrote in snipped-for-privacy@googlegroups.com:

You may want to use a MOSFET

FDV301 is one of our basic gumdrops. With 3.3 volts on the gate, Rds-on is about 4 ohms. We pay about 4 cents.

There are nfet parts with much lower Rds-on too. IRLML6344 is around

Lower Ron will have bigger capacitances.

The Digikey search engine would find a lot more.

A little cmos logic gate might be an option.

Alternatively, a 2SD2114K (NPN, beta ~ 1200, V_CEsat ~ 50 mV @ 120 mA).

Cheers

Phil Hobbs

That's interesting. -12 on the base!

I'd force in a bit more base current if he wants low Vc at 120 mA.

You may find something suitable here:

El 18/03/2021 a las 8:21, snipped-for-privacy@gmail.com escribió:

The NTR4501 has 70 mohm with Vgs = 3.3V, Id = 3.2 A and it comes in SOT-23.

torsdag den 18. marts 2021 kl. 08.21.54 UTC+1 skrev snipped-for-privacy@gmail.com:

Thanks Lasse, I think that must be the explanation, I had been wondering how the OP could have got such a high "saturation" voltage. Countless penny cheap bipolar gum drops would have no problem getting Vcesat far below 600mV at 120mA.

piglet

I tried some PBSS305NZ (and PZ) recently to get a near rail-to-rail output for a low voltage class B line driver. Worked fine at normal temperatures, but became very leaky at high temperatures, which may not matter to you but did to me. I had my suspicions, but it was worth a try.

As it was clear they weren't suitable, I didn't do a thorough test, but it seems that much above 125'C and the Silicon transmutes to Germanium. Well, sort of.

It's my fave higher-current cap multiplier transistor.

Cheers

Phil Hobbs

Have you done a closed-loop cap multiplier, to get slow DC regulation?

I'd expect that Vbe changes with temperature could make 1/f noise that a slow loop could fix.

Sure. You can even put one inside the feedback loop of a positive switcher--you have to use split feedback (AC from switcher o/p, DC from cap mult o/p) and watch out carefully for the resulting sneak path--a series RC in parallel with the cap multiplier. (An op amp is a help.)

This doesn't work with inverting bucks, unfortunately--the feedback is upside down.

You can also do things with a TLV431 on the base of the pass transistor, but that's much noisier.

When I need to do that I generally use an op amp and a handy reference, such as one in a DAC, or else an LM329B. (TI discontinued them but AD still makes the old LTC version.) The TO-92 package helps avoid voltage shifts due to die stress.

Cheers

Phil Hobbs

(Just had a gig come in to do noninvasive *in utero* measurement of fetal blood oxygen. Fun!)

Thanks. I've tried several, and none, including this one, perform as well as a BC817. In this case, the S/D voltage is less than the C/E of the BC817, but the Drain current is less. Perhaps I will stick with the BC817. It works OK. Not great, but well enough, I suppose.

The FDV301 will work, but at 4 Ohms the S/D voltage is over 500mV. The BC817 beats that.

I am showing the leakage with the IRLML6344 is way too high.

The frequency is only 3KHz, so a modestly high capacitance is not a big problem.

I tried that, but for some reason it keeps croaking on me.

I am not sure what you mean by "high side", but no. 'Standard open collector with the base grounded. I started with a 2N2222, and went from there. After I posted this originally, I tried a BC817, which gives me less than 300mV.

On a sunny day (Fri, 19 Mar 2021 19:24:22 -0700 (PDT)) it happened " snipped-for-privacy@gmail.com" snipped-for-privacy@gmail.com wrote in snipped-for-privacy@googlegroups.com:

According to Ohms law, if the SD voltage is then lower the load gets more, and more current will flow. You must be doung something wrong.

Circuit?

• | load | transistor | ///

The 3K base resistor seems way too high. If supply is 3.3V then the resulting 870uA base drive will not saturate the transistor for collector currents above a couple tens milliamps. For saturation switching 120mA the base drive should be at least 3mA (pref more) so try changing the 3K resistor to 330-750 ohms.

piglet