Timescope,

It's a variable wirewound potentiometer in series with the zinc lead to the bonding system.
It regulates the zinc current, therefore adjusting the netative voltage on the bonding system.
The digital voltmeter is connected between the silver electrode and the bonding system.
Just for fun, here's our home's solar page:

https://enlighten.enphaseenergy.com/public/systems/jMNQ131848

We installed 12 240 watt panels with Enphase inverters a few months ago and after a teething process, we're supplying our house with most of its electrical needs in the Pacific North WET.

I added a DC bypass system so not if, but WHEN the grid goes down, we have backup power via a bid lead acid bank, a sine wave inverter and a transfer switch.
The Enphase inverters instantly shut down as they feed the grid, but the DC side feeds a standalone inverter.

Only problem, I have to reset all the clocks as the inverter has a delayed start.
Oh, the agony !! lol

 

Isn't an electrolytic effect (like the problem) related to current strength?

Why not regulate current (like Bill said) but keep it very low. The voltage should then adjust automatically to the right value to suit system conductivity, but always keep a very low current to reduce anode corrosion etc.

 

Señor Mechanico,

Could you please draw a picture of the electrochemical cell? Indicate the point where you are sensing voltage and where you want to the output of the control system.

This question makes me feel like my worlds are colliding - but that is a story for a later date.

 

Isn't an electrolytic effect (like the problem) related to current strength?

Why not regulate current (like Bill said) but keep it very low. The voltage should then adjust automatically to the right value to suit system conductivity, but always keep a very low current to reduce anode corrosion etc.

From what I've read, you need a minimum voltage on the bronze to protect it.
The current required to do that depends on the conductivity of the water.
Salt water will conduct more than fresh water, and zinc won't generate enough voltage to protect in fresh so magnesium anodes are used in that case.

The minimum amount of current (with the needed voltage on the bronze) is ideal.
Any more than that and zinc is being wasted, not to mention the paint problem.

I'd love to find out I'm getting overly protective of my underwater metal, but the folding propeller on the boat was a mess of electrolysis when I bought it.
Over 1000 dollars later, I'm not ready to do that again.
On our previous boat, I had a system with a series resistor pot from the zinc to the bonding system as I've described and it worked great, but that was in offshore tropical conditions where the salinity didn't vary much.

Re the diagram, see post #3.

Steve

 

senormechanico

Is this the idea ? Please make any corrections.

Timescope

 

That's it in a nutshell.

The difference in voltage between the thruhulls and the silver electrode is the control for the variable resistance (the 250 ohm resistor).
As the negative voltage on the thruhulls drops, the circuit causes the resistance to drop, adding zinc current which raises the negative voltage.
A negative feedback loop with a gain of 1 ?

In practice, the resistance probably varies between 80 to 150, although I haven't actually measured it.
I've just had to chase the proper voltage by adjusting the pot which is a multi turn affair.

Steve

 

Having clarified the connections, I'll leave you in the good hands of those members that responded to your post to come up with something.

Regards,
Timescope.

 

Here is my first shot. I hope the feedback is the proper sense to keep the voltage constant. If not, we can make a few mods to change the feedback sense.
The circuit basically switches the Zinc anode voltage off and on, such that the average voltage between the silver electrode and the thruhulls (GND) averages to 0.55v (adjustable 0.5v to 0.625v with R4).

 

Below is my first shot at a circuit to do what you want.

The two resistors in series with the bronze and zinc voltages are rough values to simulate the resistance of the electrodes.

My one confusion is still about polarities. My interpretation of the Galvanic Table is that bronze is more negative than silver and zinc is more negative than bronze so that's how I represented them.

The voltage between the silver and bronze electrodes is measured by U1, configured as a differential amplifier. This voltage is compared with the reference voltage from U3, adjusted by pot U4 to the value you want.

U2 controls the gate current of M1 so that the differential voltage stays at the U4 value.

M1 can be any N-MOSFET with a Vth threshold voltage of >2V minimum (not logic level types).

The actual circuit may need some tweaking since I'm not sure how the real electrodes behave.

I didn't show the 0.1μF decoupling cap which should be connected across the op amp's package power pins. There's only one op amp package since the LM324 is four op amps in one package.

 

So we have two "first shots" fired fairly close together, one switching and one linear.

Ron, I noticed you have the reference connected directly to the power. I believe you need a resistor is series.

 

Thanks guys!

One thing I need to watch for is RFI as I have a HAM rig aboard.
In the first schematic, I suppose putting a couple of caps across the FET would average out the output voltage and suppress any generated RF noise.
Does that sound reasonable?

Steve

 

Thanks guys!

One thing I need to watch for is RFI as I have a HAM rig aboard.
In the first schematic, I suppose putting a couple of caps across the FET would average out the output voltage and suppress any generated RF noise.
Does that sound reasonable?

That would help. But since the rig only transmits for short periods of time I wouldn't expect it so be a significant problem in any case.

 

So we have two "first shots" fired fairly close together, one switching and one linear.

Ron, I noticed you have the reference connected directly to the power. I believe you need a resistor is series.

Oops! Thanks for catching that. I have a habit of drawing lines where components will go, and then dropping the parts onto the lines. I forgot that resistor.
I have edited the original schematic.

 

It will take some time for me to round up all the parts between work and other commitments.
Believe me, I'm impressed with and grateful for your help and will post results as soon as I can.

Steve

 

Thanks guys!

One thing I need to watch for is RFI as I have a HAM rig aboard.
In the first schematic, I suppose putting a couple of caps across the FET would average out the output voltage and suppress any generated RF noise.
Does that sound reasonable?

Steve

I don't think a cap will reduce the generated EMI, but I think the hysteresis will make the switching transients be few and far between.

 

Thanks.
I'll address the EMI if it becomes a problem if it indeed even happens.
There have been other noises already hunted down such as my MPPT solar controller.
It makes a perfect (although weak) silent carrier on channel 16 !

Steve

 

All parts I didn't already have are on order. WOO HOO !!
Now all I need is finding the the time to build once they arrive.

Steve

 

I'm having second thoughts. I don't think my circuit will work as drawn. You should try Carl's. I'll work on mine some more.
I hope you ordered parts for his circuit.

EDIT: Now I'm having third thoughts.
The grounded thruhulls make the problem very confusing to me, but I think maybe I did it right. If you were planning to build my circuit, I suppose you should go ahead and try it.
Sorry for the uncertainty and confusion.

 

Ron,
Your circuit was first in line (by two minutes) so that's the one I'm trying first.
I think it will work.

Basically, the zinc will supply the current for the thruhulls even if they are at battery negative.

I ordered all the stuff I needed from a guy on Ebay.
I had to buy 5 of each, ten of each etc, but the whole thing including free shipping was scarcely over $20.

Steve

 

Ron,
Your circuit was first in line (by two minutes) so that's the one I'm trying first.
I think it will work.

Basically, the zinc will supply the current for the thruhulls even if they are at battery negative.

I ordered all the stuff I needed from a guy on Ebay.
I had to buy 5 of each, ten of each etc, but the whole thing including free shipping was scarcely over $20.

Steve

I have also been thinking about your RFI concerns. I have realized that, if the circuit works, the MOSFET will be switching rapidly. I think a capacitor from the +input of the comparator to ground will slow the RATE (but not rise and fall times). I would start with 100nF.