Supply 5V, 5mA from the 240VAC mains without SMPS

[treez]

Hi,
We need to supply 5V, 5mA from the 240VAC mains and can think of the attached four methods. We do not want an SMPS as its too expensive.
Can you think of any better way? Ie cheaper or more efficient, or less components?
(schem and LTspice sim attached)

 

[asdf44]

Zener shunt regulator is the least efficient. Always draws worst cast current.

Series regulator is next best. My preference as I’ve expressed before is to use a depletion fet which eliminates R6 to the high voltage rail and thus improves efficiency size and cost.

The zener in the fet example can be replaced with a tl431 which modulates the fet gate while sensing/regulating the output. The eliminates the need for a following regulator.

Finally the cap dropper is the lowest loss especially if you know it’s always 240.

 

[KlausST]

Hi,

I'd go for the capacitive dropper... and series voltage regulator.

You also may use some kind of a capacitive divider to limit the output voltage (before the regulator) in "no load case".

Klaus

 

[Easy peasy]

controlled HV mosfet with good sized R in the drainto dissipate most of the power - control mosfet for 5.00V out and 7mA curr limit = pretty easy ....

 

[treez]

controlled HV mosfet with good sized R in the drainto dissipate most of the power - control mosfet for 5.00V out and 7mA curr limit = pretty easy ....

Thanks, i am guessing you are referring to eg a ZR431 with cathode to gate, and ref pin " looking" at a divider on the 5V rail which would be coming off the fet source?
I would be put off using a drain resistor due to the inrush power spike in it when on'd at mains peak, though i guess you have a current limiter in there to stop that

 

[Easy peasy]

yes the mosfet ( or igbt ) has volt and current limit - the big R is to limit the Pdiss of the device and act like a fuse if it goes short - 5W zener on Vout to ensure the fuse blows ... no need for a ZR431 - a simple discrete ckt will work fine ...

 

[treez]

Thanks, use of IGBTs as linear regulators is very interesting, as they are tougher than fets and more resilient to current surges. I have lost count of the number unsucessful google searches i have done for igbt based linear regs. Also, igbt based current clamps.

 

[asdf44]

If you're not doing cap dropper then....depletion.





This shows the simple zener and a TL431 high voltage regulator topology. I've used both of these. Again the advantage is no high voltage resistor biasing the gate. Only the fet sees high voltage (bonus that it's drain tab is heatsunk to the 400V plane).

The first Vout=Vzener+Vgsoff. The second creates a well regulated 5V. An additional R on the source adds a current limit.

The LND150 is what I had a model for, DN2470K4-G is an example of a part that can supply more current.

 

[Easy peasy]

those ckts are risky in that if the fet fails - the following ckts and any isolated wdg's on any comm's transformers go poof ...!

 

[asdf44]

Fusing element to prevent single point failure works in these circuits as well as any other.

 

[FvM]

According to datasheet LND150 Idss range is 1 to 3 mA. Thus the post #8 circuits can't be guaranteed to supply more than 1 mA. "Bigger" depletion mode transistors like DN2470K4-G should work.

 

[asdf44]

(For some reason I love these high voltage linear circuits...)

Here they are with a higher current mosfet and the current limiting resistor R7/R8 which creates a Vgsoff/R8 upper current limit. In the first example it creates some Vout error, in the second example it doesn't. If you want to get clever (too clever probably) R8 can be a PTC tucked close to the fet to share heat and become a thermal shutdown.


Though finally the best choice is probably Microchip's 450V LR8 regulator with a 50 or 100V zener in series to add additional voltage safety margin if needed.

https://www.microchip.com/wwwproducts/en/LR8#additional-features

 

[treez]

Thanks asdf44, your RHS example above is absolutely great. It looks so good that id like to simulate it to see it going.,...but ltspice doent do dep mode fets.
Though one point is that depletion mode fets at more than a few hundred volts seem a little lowly stocked and a little pricey

- - - Updated - - -

Also, it a high voltage dc supply is suddenly applied as a step input.....then that dep mode fet is just a through connection, and so the tlv431 is going to have to quickly pull down its cathode in order to stop the high voltage getting through to the low voltage circuitry being supplied......though i suppose the action of the current coming through the R8 is going to instantly pull the gate down anyway....it just seems too good to be true.

 

[asdf44]

Add these lines to standard.mos (probably in C:\Users\youruser\Documents\LTspiceXVII\lib\cmp) to add these to your fet library.

.MODEL DN2540 NMOS (LEVEL=3 RS=1.05 NSUB=5.0E14 DELTA=0.1 KAPPA=0.20 TPG=1 CGDO=3.1716E-10 RD=11 VTO=-1.50 VMAX=1.0E7 ETA=0.0223089 NFS=6.6E10 TOX=725E-10 LD=1.698E-9 UO=862.425 XJ=6.4666E-7 THETA=1.0E-5 CGSO=2.50E-9 L=4.0E-6 W=59E-3 mfg=Depletion)
.MODEL LND150 NMOS (LEVEL=3 RS=150.00 NSUB=5.0E13 DELTA=0.1 KAPPA=1.O TPG=1 CGDO=2.1716E-12 RD=40.0 VTO=-2.0 VMAX=1.0E8 ETA=0.1 NFS=6.6E10 TOX=1.0E-7 LD=1.698E-9 UO=862.425 XJ=6.4666E-7 THETA=1.0E-5 CGSO=5.09E-10 L=10.0E-6 W=600E-6)
.MODEL LND250 NMOS (LEVEL=3 RS=150.00 NSUB=5.0E13 DELTA=0.1 KAPPA=1.O TPG=1 CGDO=2.1716E-12 RD=40.0 VTO=-2.0 VMAX=1.0E8 ETA=0.1 NFS=6.6E10 TOX=1.0E-7 LD=1.698E-9 UO=862.425 XJ=6.4666E-7 THETA=1.0E-5 CGSO=5.09E-10 L=10.0E-6 W=600E-6)

If you implement the TL431 variant I'd consider adding a TVS clamp on the gate and/or source to prevent any startup (and esd) funny business. Stability is also a bit tricky to analyze (though no gain since its a follower topology). When I added a 400V startup step it set the TL431 circuit ringing. A compensation C and output C on the TL431 quieted it down. After doing that I see no overshoot on an input transient on either circuit.

The zener circuit on the left has no stability concerns so you can stick C anywhere you want to control startup/transient behavior. Add a following regulator for output accuracy and it's about the same complexity as the TL431 variant with additions.

I see 38,000 DN2470K4-GCT-ND in stock at digikey US at 80 cents quantity one. Cheap and available by my standards anyway.

Finally these circuits work fine with enhancement fets if you pull the gate up to the drain instead of source. But while the depletion circuits work consistently from about 10V and up a drain pull-up probably won't supply minimum zener/cathode currents until higher voltages.