Depletion MOSFET for supply switch?

[DigiTan]

I'm in need of advice on what type of MOSFET I should use to switch off current from two battery supplies.

Basically, I've got this plan for a 6V regulator. The regulator output feeds to the supply of a graphing calculator that would normally use AAA batteries. I want to augment the original batteries with a second bank of AA cells. I've done this by linking both battery groups in parallel through a Schottky barrier pair to prevent backward current from the strongest battery set.

Then I added a regulator/TVS so that a wall adapater can be used as a third source. This also runs in parallel with the batteries and uses the same Schottyky barrier.

**broken link removed**


To maximize battery life, I want to completely shut of current from both battery groups whenever the adapater source is present. I used a relay in my previous version, however the switching time caused transients that could reset the calculator under certain conditions.

I want to replace the relay entirely. I've heard of depletion-mode MOSFETs being compared to normally-open switches, and this is exactly the type of behavior I need.

My questions are: will a depletion MOSFET in series with Battery supply 1 and 2 provide the switching I want without the need for a negative voltage supply? And will it have a very low transconductance when it is supposed to allow current from the batteries? I'm thinking the MOSFETs would go in series with D3 and D4 on the diagram?

 

[Russlk]

A depletion mode MOSFET is like a normally CLOSED switch. Mouser has some but they are all high voltage, high resistance (50 ohms). You would be better off to use a P-type Mosfet and turn it off when the walwart is connected.

 

[Roff]

The simplest method is to make sure the regulator output voltage is slightly higher than the battery voltage. A 1N4148 diode in series with the GND leg of the regulator will give you about 6.5V out.
If you are concerned about diode leakage, or you don't like the method above, you can add a P-channel MOSFET as shown below (where you suggested). Then the regulator voltage doesn't have to be higher than the battery voltage.
I think a depletion-mode MOS would also work here, but I believe PMOS FETs are more readily available.
You should move the 2000uF cap to the output. I didn't see it when I was adding the PMOS.
BTW, is there a reason for R2?

 

[DigiTan]

Oh, that's right I meant to say "normally closed" instead of "normally open."

I'm liking both methods so far. I might go with the PMOS for now for a smaller Volt swing when the adapter goes (in)active.

R2 was originally 2.7V a zener diode (now gone). I'm wanting the use a decorative LED but I never decided if I want I want 2.7V orange, 3.2V blue, or the color-shifting type with the built-in µcontroller. I'll get rid of R2 and calculate R1 as soon as I choose an LED and find out the I/V spec for it.

 

[Dr.EM]

Don't know if i've got this right, but I thought you needed a suitable sized cap located after a bridge rectifier to give proper DC, the diagram only appears to show the bypass cap for the regulator :?:
Off the point a bit, sorry

 

[Roff]

Don't know if i've got this right, but I thought you needed a suitable sized cap located after a bridge rectifier to give proper DC, the diagram only appears to show the bypass cap for the regulator :?:
Off the point a bit, sorry

Good catch, doctor.

 

[DigiTan]

Oh yeah. I guess I should've mentioned it. Input voltage on JP1 is assumed to be regulated DC. The FWR was added so both center-positive and center-negative adapters could be used. I guess a smoothing cap still wouldn't hurt. I'd add whatever will fit in the enclosure.

 

[DigiTan]

Alright, another consern I had with this setup is the inrush current due to C1. Since my fuse is rated 1/4 Amp, will the inrush run a risk of blowing the fuse? The fuse is a miniature radial lead type like the kind **broken link removed**. All I know is this is not the slow-blow fuse type.

Also, I was thinking of using a TVS in parallel with the 33V zener diode for voltage spike protection. I've never used a TVS before so can anyone recommend any certain type for this application?

 

[Roff]

What is the voltage at the input to your regulator (after the bridge rectifier)?

 

[DigiTan]

The 78L06 had a 1.7V dropout so that voltage will range from 7.7V to 30V.

 

[Roff]

The 78L06 had a 1.7V dropout so that voltage will range from 7.7V to 30V.

You said above:

Input voltage on JP1 is assumed to be regulated DC.

Did you just mean filtered? 7.7V to 30V doesn't sound very regulated - unless you are planning to have a variety of sources, each being regulated.

 

[DigiTan]

Yep, its a variety. I want to use any available adapter that fits.

 

[Roff]

A fast-acting fuse seems unnecessarily restrictive for a circuit that has to charge a 2000uF cap. Any reason you can't use a slo-blo?

 

[DigiTan]

I guess I could try one if I can get it in the small size. If not, I could probably use a higher fuse rating. The 78L06 provides its own short-circuit protection at the output so it won't be as dependent on F1 as I originally thought.

 

[Roff]

I guess I could try one if I can get it in the small size. If not, I could probably use a higher fuse rating. The 78L06 provides its own short-circuit protection at the output so it won't be as dependent on F1 as I originally thought.

The fuse is good for overvoltage protection, and for shorts in front of the regulator. **broken link removed** appears to have some time lag fuses in that package.