Series Zener Instability 1

[cmb042]

My regards circuit jockeys,

I have a sensor device that is lowered into a well. Its power cable also serves as its data transmission line. The voltage supplied to the circuit is regulated by two zeners in series. This creates two voltage rails, one to power the sensor, the other lower voltage to power a multivibrator that shapes the output of the sensor into square pulses coupled to the line. I want to increase the voltage of the second zener, to raise the voltage of the pulses, but when I do the signal gets obfuscated in noise.

I have this gut feeling that the series zeners are the cause of this problem, but I can't think of why this is happening. The word leverage comes to mind, I read over zeners in my old textbook, but it didn't give me any new ideas. Can anybody explain this to me?

 

[MacGyverS2000]

What's the line voltage at the sensor, and what is the combined voltages of the two zeners?

Dan - Owner

http://www.Hi-TecDesigns.com

 

[OperaHouse]

What comes to mind is the currents has increased to the multivibrator circuit and you don't have enough headroom to regulate. Zeners are fairly poor at regulating over wide ranges. Think of using an amplified zener like a LM431.

 

[cmb042]

So how to you calculate the headroom of a zener? It is a single solid state junction, so I can't think of anything that would delay its response. It should be practically instantaneous. Even datasheets for zeners seem to dwell on temperature properties. I want to understand why it is not working before I look for something better.

120V DC line
32V total to the sensor with 5.6V to the multiV, works fine
24V total with 12 to the multiV, still working
24V total with 18 volts to the multiV, choppy distored signal.

 

[IRstuff]

The notion that zeners "regulate" is, I think, a bad concept for circuit design. They are basically diodes operated in reverse breakdown, and their stability is only maintained by running a very low current through their bias network.

If you are trying to drive a long cable, you need lots of current, which is anathema to a zener circuit.

As suggested by Operahouse, the LM341:

http://www.datasheetcatalog.com/datasheets_pdf/L/M/4/3/LM431.shtml

behaves like a zener, but with a substantially higher current shunting capability, roughly 10 times what a bare zener can do.

TTFN

FAQ731-376

 

[btrueblood]

Could you not use a 4-20 mA signal instead, or a three-wire cable?

 

[itsmoked]

What is your multivibrator? Your problems seem to magnify just around the highest reasonable voltage threshold for any common multivibrator. Could you be exceeding the multivibrator's maximum voltage on the way to trying for a larger signal?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[cmb042]

The multivibrator is rated for 20V. I don't think it's the problem, least not directly. I was thinking maybe some sort of feedback loop, the pulse from the multivibrator changes the voltage lvl the zener sees, some interaction between them bounces it around... There is a 10uF cap parallel with the second zener, maybe I should try one on the first zener as well.

 

[BobM3]

How constant is your 120 V DC? Is is rectified 115 VAC? Are the zeners getting hot?

 

[IRstuff]

The ONLY reason a zener voltage might change is because of excessive current.

TTFN

FAQ731-376

 

[ScottyUK]

Are you saying that this circuit is a two-wire design and that the multivibrator actually modulates the same line that powers it? Sounds like a great way to create feedback and instability.

Time for a look at the circuit - you can upload it to engineering.com as part of your reply.


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[Skogsgurra]

Another case of dabbling in someone else's field of expertise. I guess that you are using corporate money - or government's. Then spend it wisely and get a good consultant to help you.

An analogy: If I need, for example, a shaft with bearings for a measuring wheel, I go to a shop to get it done. I do not start my old lathe and produce a substandard shaft.

And, I agree that putting two zeners in series and not even know if you are into avalanche region all the time sounds like a very bad idea. There are standard solutions to most telemetry situations.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[cmb042]

I am not concerned with power, the whole thing pulls about 42mA. The zeners are rated more than 1W. While I agree that there are better voltage regulators out there this is not my design. I'm trying to tweak it not reinvent it.

Circuit is posted for those that want a look at it.

 

[IRstuff]

It would help if you would bother to post the circuit.

The point about the current is NOT the zener power capacity, but about your circuit stealing bias current from the zener. If your zener is only barely biased into breakdown, then ANY current reduction will cause a drastic voltage change.

TTFN

FAQ731-376

 

[OperaHouse]

Add a regular diode in series with D4. D4 allows current to go back when output pulses low. Certainly C4 needs to be much larger and an additional cap is recommended on the next zener.