diy solar

diy solar

Ammeters and diodes

You can use a analog microamp meter with a shunt

Now, that's an idea :·)
Let me see if I get it right. I use a normal 1 milliohm shunt. Current flows both ways through it.
Then I put a resistor (to be calculated) in parallel with that. A - fixed, proportional - percentage of the current flowing through the big shunt will flow through the resistor, right?
So I can use a small diode on that part, which will dissipate almost-nothing, and current there will only flow one way.
I measure micro/milliamps over the resistor/diode, find the right ratio, and call microamps amps.

So when current flows out of the battery, my panel meter shows it. When it flows in, it doesn't.
I wouldn't even have to white-out the μ, as long as I know what they are, it could be saying "apples".
All I need is to see the needle moving :·)
 
Is there some reason you only want to measure the current one way? There are analog meters that go both ways.
 
Shunt is basically low Ohm resistor, why do you want to put another resistor in parallel with the Shunt? You put the uA meter (which has its own internal resistance) in parallel with the Shunt.
 
It is correct that if the uA meter does not match a standard shunt you want to use you can add a series resistor to make it correct. Shunt does have to produce enough voltage drop for uA meter plus any additional dropping resistor in series with meter.

Many uA meters are 0.1v full scale, but not all.

You can use an op amp to amplify shunt voltage output but you have to deal with DC offset and temp drift of op amp.
 
Is there some reason you only want to measure the current one way? There are analog meters that go both ways.
As I said:
I don't want to know if the battery is charging or discharging.
I want to see what I'm drawing. Not a number, an immediate visual representation of a physical quantity. It's the way my mind works.

So I can see what I'm charging one one meter, what I'm discharging on another. Silly? Probably. Cilly? Certainly ;·)
Call it "energy aesthetics" :·)

Now, I'm finding it difficult to find a μA meter that is not both small and ugly. I might have to resort to a heart ;·) transplant, but... aesthetics, you know... :)

[EDIT] To better explain my "predicament": I have a (pretty enough) ammeter between the panels and the charge controller.
It moves with the clouds :·) Now, that works just fine, as the controller has transistors on the end, they don't let any current flow back into it.
If I put an ammeter from the battery to the loads (including inverter) as soon as the charge exceeds the load, the needle goes the wrong way.
 
Last edited:
Some of us use both analog and digital meters, the beauty is you can place an analog and a digital instrument on the same shunt.
You can use any wire as shunt, but you may keep in mind thermal coefficient of resistance, copper is not so good because thermal coefficient is quite hi.
Ih you do not want use a shunt you can use a hall transducer open or closed loop, now are available open loop transducers precise enough for most needs with one percent initial accuracy.

If you want something stable and pecise for SOC meter for example you must use a shunt with lo thermal coefficient and precision voltage reference and amplifier.
LTO for example has better than 0.995 charge efficiency, you need one per mile intial accuracy and lo drift as possible to measure state of charge correctly, in my case drift is lower than one 16bit ADC quantum.
It goes up and down on the same path tens of cycles without top or bottom reset.
 
Last edited:
This doesn't make sense to me.

You want to see at a glance if it's charging or discharging and have a visual reference for how much, not the number itself.

The blue sea needle goes left for discharge and right for charge. Needle goes more left then it's more discharge. More right means more charge.

It tells you what's going out vs what's going in without any subtraction necessary. If you're charging more than discharging then you'll see it swing right.

I'm confused as to why that doesn't meet the requirements you've put forth and you're just dead set on using only what you've already decided is the solution.


Are you somehow trying to separate your charge current from your load current to display both at once?

Edit: I see this is in fact what you're looking for.

You can't do that on one wire and will need to sense from the charger and load side separately. Any meter or shunt would show the same result of "backwards" relative to whatever way you installed it as "forwards".
 
Last edited:
I see this is in fact what you're looking for.
Right. I want to see what I'm drawing regardless of what I'm charging. And I want to see it with a needle, not numbers.

You can't do that on one wire
Apparently, I've been given a way to do it. Have a resistor in parallel with the shunt, and put a small diode on that.
Then work out as close as possible what the resistor would have to be - with diode and all - to measure the small parallel current.
Which should be constantly proportional. Then choose a micro/milli-ammeter with a scale that looks good and sensible.
I would have tried already if I weren't really busy with other things. But I will, eventually :·)
 
Right. I want to see what I'm drawing regardless of what I'm charging. And I want to see it with a needle, not numbers.


Apparently, I've been given a way to do it. Have a resistor in parallel with the shunt, and put a small diode on that.
Then work out as close as possible what the resistor would have to be - with diode and all - to measure the small parallel current.
Which should be constantly proportional. Then choose a micro/milli-ammeter with a scale that looks good and sensible.
I would have tried already if I weren't really busy with other things. But I will, eventually :·)
I would need to see the full circuit you are thinking of, but I don't think this will work as you think. Diodes are non-linear, not anywhere near "constantly proportional."

The voltage drop across a diode is a constant, typically 0.7 for a silicon diode. But, that is only true once it is fully conducting, which will require more than 0.7V. Below that point, it's response is a curve, not a line.

You propose this in parallel with the shunt. The shunt is of a very low resistance, and will hold the voltage across the diode very low, much lower than 0.7 V. The diode will not be conducting. Also, having anything in parallel with a shunt will change the resistance across the shunt and cause any meter connected to it to read incorrectly.

If you are going to use a diode, however you use it, you will need an amplifier to get the voltage from the shunt well over 0.7V. I would think something like coming from one side of the shunt, to an amplifier, through the diode, then an attenuator. Basically a pair of resistors, one in series, the other to ground. They would be selected so the attenuation exactly equals the gain of the amplifier. One of the resistors could be a potentiometer for adjustment. Then to the ammeter. The other side of the meter would go directly to the other side of the shunt. This would cause the meter to work in one direction, but not go negative in the other direction.

I have become unclear though on exactly what you are trying to achieve however. I am very clear you want an analog gauge, and for some reason the blue seas gauge that moves both ways won't work for you. Can you please restate what you are trying to achieve? What exactly do you want this meter to show you? Are you looking for a pair of meters, one that shows current in one direction, and the other that shows current in the other direction? The easiest solution might in fact be a digital solution, that drives 2 analog meters. Perhaps an Arduino or something. The shunt could feed an analog input to the Arduino, and the Arduino could decide what to do with it and drive the two meters. Or a single meter that always moves positive, and illuminates a charging or discharging LED, or any of a bunch of things that might be fun.
 
Now, that's an idea :·)
Let me see if I get it right. I use a normal 1 milliohm shunt. Current flows both ways through it.
Then I put a resistor (to be calculated) in parallel with that. A - fixed, proportional - percentage of the current flowing through the big shunt will flow through the resistor, right?
So I can use a small diode on that part, which will dissipate almost-nothing, and current there will only flow one way.
I measure micro/milliamps over the resistor/diode, find the right ratio, and call microamps amps.

So when current flows out of the battery, my panel meter shows it. When it flows in, it doesn't.
I wouldn't even have to white-out the μ, as long as I know what they are, it could be saying "apples".
All I need is to see the needle moving :·)
Nope, doesn't work that way. You have something like 0.05 volts over the 1 milliohm shunt and normal silicon diode doesn't conduct before 0.7 volts or so.

There is a simple trick to make this work with diode bridge but you'd take huge penalty on voltage loss, 1,4 volts lost.
 
Right. I want to see what I'm drawing regardless of what I'm charging. And I want to see it with a needle, not numbers.


Apparently, I've been given a way to do it. Have a resistor in parallel with the shunt, and put a small diode on that.
Then work out as close as possible what the resistor would have to be - with diode and all - to measure the small parallel current.
Which should be constantly proportional. Then choose a micro/milli-ammeter with a scale that looks good and sensible.
I would have tried already if I weren't really busy with other things. But I will, eventually :·)
Nope, it does not work that way.
 
He may use an operational amplifier circuit, an ideal rectifier, diode is placed in feedback loop and, non linearity is compensated, but is too complicate, it needs auxiliary power.
There are two ways in fact, wire charge and load apart as most users suggested, or place a meter with center needle.
 
I would need to see the full circuit you are thinking of, but I don't think this will work as you think. Diodes are non-linear, not anywhere near "constantly proportional."

The voltage drop across a diode is a constant, typically 0.7 for a silicon diode. But, that is only true once it is fully conducting, which will require more than 0.7V. Below that point, it's response is a curve, not a line.

You propose this in parallel with the shunt. The shunt is of a very low resistance, and will hold the voltage across the diode very low, much lower than 0.7 V. The diode will not be conducting. Also, having anything in parallel with a shunt will change the resistance across the shunt and cause any meter connected to it to read incorrectly.

If you are going to use a diode, however you use it, you will need an amplifier to get the voltage from the shunt well over 0.7V. I would think something like coming from one side of the shunt, to an amplifier, through the diode, then an attenuator. Basically a pair of resistors, one in series, the other to ground. They would be selected so the attenuation exactly equals the gain of the amplifier. One of the resistors could be a potentiometer for adjustment. Then to the ammeter. The other side of the meter would go directly to the other side of the shunt. This would cause the meter to work in one direction, but not go negative in the other direction.

I have become unclear though on exactly what you are trying to achieve however. I am very clear you want an analog gauge, and for some reason the blue seas gauge that moves both ways won't work for you. Can you please restate what you are trying to achieve? What exactly do you want this meter to show you? Are you looking for a pair of meters, one that shows current in one direction, and the other that shows current in the other direction? The easiest solution might in fact be a digital solution, that drives 2 analog meters. Perhaps an Arduino or something. The shunt could feed an analog input to the Arduino, and the Arduino could decide what to do with it and drive the two meters. Or a single meter that always moves positive, and illuminates a charging or discharging LED, or any of a bunch of things that might be fun.
He wants to see how much charge he's getting from pv as well as how much load. Separately and simultaneously.

I'm not sure why this is so hard.

One meter on the charge cable from the charge controller.
One meter on the cable out to the load.

If you try to put a meter on a single cable that's covering both charge and discharge you'll run into all the issues being discussed.
 
I already have the ammeter on the panels. Before the controller, actually. Works like a charm.
I also have one on the load output of the controller, but that shows only a small percentage of what I draw.
I also have a kill-a-watt thing on the inverter output.

What I want to see, though, is how much current is going out of the battery. Regardless of what I'm charging.
Including what the inverter draws, all of it.
And I really wonder how anyone could possibly not want to know that.

Now, I want to see it analogically, see a needle go up and down, others may prefer to have bright numbers whizzing on a display.
And I obviously don't want to go subtracting charge from load to work it out :·)

So, putting a big diode and an analog ammeter on the battery cable would do that.
The diode would dissipate heat under load. My load is usually well under 100W, so not a big deal, but it's occasionally more.
Now, if I can do it with a parallel shunt, and measure proportional current on that, it would work, for me.
Otherwise I'll just put a big schottky diode, or mosfet, or whatever dissipates the least, there and be happy :·)
When I have a little time.
 
I do not know why you keep talking about putting in the diode.
You should provide the wiring as to what you are trying to accomplish with the diode and the meter that you think it will work the way you want it to work.. That will really clear out the confusion, just show us your wiring of what you have right now with the existing meters, and such and how you are going to wire in the shunt, resistor, and diode into the system.
Do you really have separate wires one for charging the battery and another one for discharging the battery?

"What I want to see, though, is how much current is going out of the battery. Regardless of what I'm charging.
Including what the inverter draws, all of it." How are you going to see the current draw from battery and the current being discharge at the same time? You will be either drawing from the battery (when charger cannot supply enough current for the load that is when the battery will discharge the current to the load to supplement the load requirement, and if the charger is not on at all then the battery will supply the load alone) or charge the battery when battery is too low to supply the current to the load, at that point the charger has to be able to supply the current to the battery and the load at the same time if the charger can handle it.

"
Now, I want to see it analogically, see a needle go up and down, others may prefer to have bright numbers whizzing on a display.
And I obviously don't want to go subtracting charge from load to work it out :·)" What subtracting are you talking about?


Regarding to Center meter type and your reply in Your post #3:
"I did think of that (and actually, the Blue Sea one is pretty, and I did consider it).
But, simply, it won't tell me what's going out without subtracting what's going in. Defeats the "visually pleasant" concept."

I and other also do not see why you cannot just put Center meter needle display on just one wire to the battery to show the direction of the current flowing or flowing out of the battery just like the one used in the car.

Bottom Line: Your wiring diagram will surely clear the confusion as to why it can be or cannot be done.
BTW, I do have DC Amp meter that show if the battery is being charged or if it is discharging, simple.
 
Last edited:
What I want to see, though, is how much current is going out of the battery. Regardless of what I'm charging.

Do I really have to draw you a picture?
arrabbiato1.gif
 
So put the Amp meter on the battery cable, simple as that. Are you EXPECTING to see the charging current to the battery and the discharging current from the battery at the same time? You are ether pulling the current from the battery or pumping the current into the battery, but it will not happen at the same time.
Where do you think the current to the load come from when the battery is fully charged and the charger is running?
 
Back
Top