A zener diode as power supply voltage limiter ?

[ricothetroll]

Hi !
I'd like to build a pedal with a +/-12V PSU, from a 9V power input. I did it successfully with a TC962 as voltage doubler (= 18V - 2xdiode drop), then a LM78L12 for +12V on one side and another TC962 as voltage inverter on the other side + LM79L12 for -12V. My problem is : if i plug my pedal with one of those Boss PSU that deliver 12V unloaded, I'll risk to fry my second TC962, as its maximum input voltage is 18V (24 - 2*0.75 = 22.5V > 18V).

To limit the input voltage, I though of a zener diode + transistor circuit like that one :

http://hyperphysics.phy-astr.gsu.edu/hbase/electronic/ietron/bufvr.gif

Are there better options for my purpose ?

Best regards.

Eric

[familyortiz]

Rico,
   It seems like although your concept will work, it's got a lot of losses in there which will certainly drain a battery quickly. The LM series of regulators are not the most efficient and you have the drop across each.
Do you need the full +/- 12 v range? The following link shows a 9v doubled by a charge pump, which seems way easier to implement and does not need further regulation to avoid frying something. Just a thought...

http://www.diystompboxes.com/smfforum/index.php?topic=64924.0

[ricothetroll]

Hi ! Thanx for answering.
Yes I do need some full +/-12V ! That's because I'll use a AD633 for voltage multiplication, +/-12V is a minimum for acceptable S/N. Battery drain isn't a problem since I'm not planning to use my circuit with batteries anyway...
Best regards.
Eric

[slacker]

Is there any reason you can't just run the inverting TC962 off the regulated 12 volts to get approximately -12volts. That simplifies the design and solves the problem.

[ricothetroll]

Hi,
Yes that's what I did on the first version of my circuit but without a regulator the DC rail is significantly polluted with the internal oscillator of the TC962. I could clearly hear it on the audio output !
Best regards
Eric

[slacker]

In that case, I would just try a resistor and zener that should be enough to protect against over voltage.
http://www.reuk.co.uk/Zener-Diode-Voltage-Regulator.htm

[R.G.]

Yes that's what I did on the first version of my circuit but without a regulator the DC rail is significantly polluted with the internal oscillator of the TC962. I could clearly hear it on the audio output !

For US$0.35 you can get a *second* regulator to power just the TC962. It's a solution which is simple and not terribly much more expensive than the zener plus transistor.

[ricothetroll]

Hi,

Well, I actually have a few problems left :

- In case I'd use a simple resistor-zener circuit, losses are HUGE when the input voltage reaches values such as 21V (2*12V - 2*Vd) ! I planned a current of maximum 50mA on that rail.

- If I use another regulator for the second TC962, I'd have to use a LDO one in order to ensure having enough voltage at the input of the LM79L12 after the TC962 (for -12V). Those barely exist for that kind of voltage and are very expensive.

- I though of another solution that I used for signal voltage limitation : the circuit called "active clamp" as mentioned in "The Art of Electronics" (Horowitz, Hill, 4.17 - Active Clamp). I used it to limit a VC to a certain value, it's an opamp with a series resistance to the inverting input, a diode in the feedback loop and the voltage a the non-inverting input defines the limit. the output is at the inverting input of the opamp. The circuit needs a buffer right after it. Do you think I could use such a circuit with a power transistor used a buffer ? Or is there even a way to design it only with transistors (to spare some room).

Best regards.

Eric

[slacker]

I think you can just use say a 7815 before the second TC962, when the voltage is less than 15 volts it will just pass through. When it goes to 20 volts it will be regulated to 15 volts which should be enough for the TC962.
You don't really care if the 7815 regualates properly you're just using it to limit the voltage.

[ricothetroll]

Hi,
I'll try it ! I hope it won't drop the voltage too much.
Thanx a lot !
Best regards.
Eric

[ayayay!]

Dammit did it again!  keeps posting to the wrong thread!  Ignore.

[ricothetroll]

Hi,
I found that discrete LDO Regulator schematic on my HD (found it somewhere on the internet but can't remember where) :
http://www.wuala.com/ricothetroll/public/Low%20Voltage%20Drop%20Regulator.png

While doing SPICE simulations I noticed that the voltage drop is really low for moderate loads : 300mV at 97mA (Vi=10V, Vo=9.7V), then the output voltage falls dramatically. I also noticed that when I reverse the input voltage, the output is nearly nulled (-4.7mV) and the current consumption is also very low (about 60uA for a 82R load). So I thought I could use it also as a reverse polarity protection !

The thing is I don't understand how that schematic works. Could someone explain it to me ? Maybe I could then modify it for bigger loads, etc...

Thanx in advance.

Best regards.

Eric

[R.G.]

The thing is I don't understand how that schematic works. Could someone explain it to me ? Maybe I could then modify it for bigger loads, etc...

Q3 is a pass transistor. When its base is pulled below its emitter by more than 0.5V, it starts conducting from emitter to collector.

R1 and Q1 pull current through the base of Q3 turning it on. R4 turns on Q1. If this was all the circuit you had, no regulation would happen.

Zener diode D1 and the base-emitter of Q2 form a voltage reference. No current flows through them until the output voltage exceeds the zener diode voltage plus the base-emitter turn-on voltage of Q1. When the output voltage reaches this threshold, current begins flowing in Q2's collector.

This current comes from the current available to Q1's base, turning it of. This decreases the current through Q3's base, decreasing the current through Q3, and lowering the output voltage because the load pulls down the output voltage.

So as the input voltage increases, the output rises just a little less than the input until the output voltage reaches the zener plus base-emitter voltage, then it stops rising. It regulates at that level.

You can increase current by using a PNP darlington for Q3, or changing it to a P-channel MOSFET. You can change voltage by changing the zener voltage.

[PRR]

> voltage drop is really low for moderate loads : 300mV at 97mA (Vi=10V, Vo=9.7V), then the output voltage falls dramatically.

Don't SPICE stuff you don't understand well enough to work-out on a napkin.

The 2N3906 can flow a LOT more than 97mA.

When Q3 2N3906 is flowing 97mA, what is its base current? Collector current divided by current gain. Current gain of the '06 is, say, maybe 100. 97/100= 1mA base current. Where does this 1mA come from? R1 which is 15K. What is the maximum voltage on R1? Your 10V, minus about 0.7Vbe, minus maybe 0.1V for Q1's Vce = 10V max across 15K is 0.6mA.

So we find that the current-gain of this particular (or simulated) 2N3906 is 97mA/0.6mA= 150, which is indeed a reasonable value.

So the current limit is R1 and the hFE of Q3.

If you make R1 smaller, Q3 can have more base current. For (bad) example: Make R1 equal 100 ohms. It can flow almost 10V/100 = 100mA of base current to Q3. If Q3 gain is still 150, then Q3 current could be 100mA*150= 15,000mA or 15 Amps!. In fact the '3906 will melt long before that. Perhaps you can find a value lower than 15K yet higher than 100 ohms which is effective and not dangerous.

> I also noticed that when I reverse the input voltage, the output is nearly nulled (-4.7mV) and the current consumption is also very low (about 60uA for a 82R load). So I thought I could use it also as a reverse polarity protection !

Do NOT trust SPICE for extreme abnormal conditions!! Yes, for small reverse voltage, Q3 emitter junction will block. But the E-B junction breaks-down about 7V, and you may have 10V. Most SPICE models do not duplicate real transistors in this way.

[ricothetroll]

Thanx a lot ! I'll have to take some time to understand your posts completely 
You guys are the best teachers I've ever had 

[ricothetroll]

Ok I'm getting it ! So Q2 is here to "choke" Q1 by lowering its base current whenever the output voltage exceeds the zener voltage and thus makes current flow through it. That's clever ! One last thing I do not get is the role of R2 ? Is it supposed to keep a minimum amount of current flowing into Q3's base ?
Best regards.
Eric

[PRR]

> the role of R2 ? Is it supposed to keep a minimum amount of current flowing into Q3's base ?

It's going the wrong way to flow current INto Q3 base.

It might set a lower limit on Q1 current, or bleed Q3 leakage, but it is much too high-value.

Overall it seems to be a student excercise, not a ready-to-go plan.