Author Topic: Current and resistance  (Read 346 times)

Offline Turbidaceous

  • Participant
  • **
  • Posts: 398
  • Likes: 1
  • Living in the UK
Current and resistance
« on: February 17, 2020, 04:15:44 PM »
I understand a constant current is ideal. But if I buy a constant current power supply and use a potentiometer to reduce the ma current to what I want, won't that cause varied results because of possible voltage fluctuations?

As I understand it the volts and the ohms of the resistor dictates the current. So if I buy a constant current power supply, then cap it to 3ma with the turn of a dial on the potentiometer, won't fluctuations in voltage cause the ma to go up and down? I'm not an electrical genius, so please if you could enlighten me on this.

I found this article but I am not sure what will happen. I am going to try with a constant current 24-42 volt range constant current power supply and a 50k Potentiometer. I just don't know what the results may be, based on this info:

Quote
A constant voltage power supply will attempt to hold the target voltage no matter how much current it's asked to source.  So for simplicity's sake, assume you have a variable resistor (a potentiometer) across the output of a constant voltage power supply and you then vary the resistance which in turn varies the amount of current drawn from the power supply.  For an ideal constant voltage power supply the voltage across the variable resistor will hold constant, no matter how much current its delivering.

Of course actual devices will have design limits and at some point you'll hit a current that's more than the circuitry of the power supply can handle and at this point the voltage will start to drop and you'll hit a maximum current.  Well designed power supplies will have current limiting or short circuit protection built in to they won't blow up if someone shorts the output terminals.

An ideal constant current power supply will deliver a constant current to the load no matter how much voltage it needs to do this.  So once again for simplicity assume the same variable resistor across the output of a constant current power supply.  This time assume that you set the resistance to zero - a short circuit, or very close to a short circuit.  The power supply will now deliver the few millivolts or whatever is necessary to drive the specified current, say 1 amp for example.

Now you increase the resistance to 1 ohm.  The constant current power supply will increase the output voltage to 1 volt so it can maintain the constant 1 amp current, and so on.

Once again, real world supplies will have design limits and at some point when the load resistance gets too high the supply will no longer be able to drive the voltage high enough to deliver the asked for current and you'll hit a maximum output voltage.

Offline Gene

  • Expert
  • Participant
  • ***
  • Posts: 1160
  • Likes: 84
Re: Current and resistance
« Reply #1 on: February 17, 2020, 09:44:33 PM »
If you buy a constant current supply like that 300ma LED supply, it will try to maintain 300ma. If you try to reduce that to say 3ma using a pot, you're not going to like what you get because it will still try to maintain 300ma and as such, peg to ITS supply voltage and it becomes a really crappy constant voltage power supply at this point because it can no longer regulate current where its also not really regulating voltage - just connecting its supply voltage to its output basically.

I'm not sure what you mean "fluctuations in voltage" using a constant current source.  The whole purpose of constant current is that regardless the load (within reason) and regardless how the load resistance changes (for a Colloidal Silver cell the cell resistance drops a bit over a run as you accumulate silver ions in the water - I've seen about a volt drop give or take from start to the end of the run variation at the fixed current due to this change in cell resistance) the current source will maintain an accurate FIXED current flow. This is the whole purpose of a current limiter because without an accurate, fixed current flowing through the cell, Faraday's law of electrolysis doesn't work and you have no clue WHAT PPM you made.

The lower resistance of the cell due to accumulated silver ions, for a fixed current, as I said above would have the cell voltage drop for the lower cell resistance.

E = I * R (voltage equals current times resistance)

If R becomes less, E follows suit.

The purpose of a 10V minimum cell voltage is a bit different. It has to do with electrochemistry and the minimum potential you want for good quality production of Colloidal Silver (actually Ionic silver - the Colloidal Silver happens after reduction). You need to maintain 10V across the cell for this reason.  I usually set my cell to about 12V to account for the drop in voltage over the course of a run so I don't have to check it at all because its STILL higher than 10V at the end of a run.  I like simple!

Where things go haywire is when the voltage of your cell plus the headroom requirement of the current source (figure a volt for a 2 transistor limit and 3V for an LM317 style limiter) exceeds the applied voltage from your power supply because now the limiter can't function and its just pegged at its maximum output voltage and you no longer have a constant current.

If you use a fixed voltage power supply and a limiting resistor, as the resistance of the cell changes due to accumulation of silver ions in the water over the run, the total resistance (cell plus limiting resistor) will drop and consequently the current will go UP.  The voltage of the cell will change also where the cell voltage is a direct result of the current times cell resistance.

Offline Turbidaceous

  • Participant
  • **
  • Posts: 398
  • Likes: 1
  • Living in the UK
Re: Current and resistance
« Reply #2 on: February 18, 2020, 01:54:05 PM »
I understand potentiometers and getting the ma I need. This transistor business totally and complete looses me. I have not found anywhere enough information for me to make sense of it and put it into practise. I ordered that constant current power supply hoping that it would take the headache of the "constant" current part of it, so I can then just limit the ma. I am happy with my current setup but I really want this constant current people keep talking about. I have no problem stripping wired, bending them around contacts and electrical taping them securely. I just am clueless with regards to this transistor business, 1 of them? 2 of them? How do you use them? voltage headroom? When Amazon searching I found nothing but confusing things there too. I have not found a single shred of info that lets me know exactly what I need to do.

Right now I have a wall-wart 30v 500ma power supply which I have no idea if anything about it is constant, both the voltage and the ma fluctuate by a tiny amount every second but only a small amount 0.1 of a volt and maybe 0.02 of a ma. Then I have put a potentiometer inline to tune it down to the ma I want. What would I need in order to make this constant current?

Offline axkman

  • Participant
  • **
  • Posts: 36
  • Likes: 0
Re: Current and resistance
« Reply #3 on: August 16, 2020, 01:44:07 PM »
I use a LM317 as a current regulator which is set to 2ma. It always tries to set the current to 2ma, no matter the load or voltage.

LM317 voltage regulators can be bought very cheap on aliexpress and then slightly modify it to turn it from voltage regulator to current regulator.

Offline Gene

  • Expert
  • Participant
  • ***
  • Posts: 1160
  • Likes: 84
Re: Current and resistance
« Reply #4 on: August 17, 2020, 01:35:48 AM »
The LM317 has been discussed many times.

The TO220 packaged device is only spec'd to regulate down to 10ma.
The LM317LZ (TO92 leaded) packaged device is only spec'd to regulate down to 2.5ma.

While you can probably push it 20-30% lower than this, I wouldn't bank on it due to process spread (variations from lot to lot for when the wafers are processed). The only thing you can rely on is what the manufacturer guarantees.

If you want a fixed regulator, just pick the right resistor.

For a variable one, put a 50Ohm resistor in series with a 1KOhm potentiometer wired as a variable resistor.  Your top end current will be about 25ma. The low end, if you're using an LM317LZ regulator would be 2.5ma or a bit lower but I doubt seriously it'd go as low as 1.5ma.

The current regulator topology is simply the regulator device with a resistor (fixed or variable) in series with its output with the other side of this going to your load AND to the regulator ADJ pin.

The maximum voltage your load could ever see would be Vin- (3V + 1.25V) but choose a higher Vin to give yourself maybe 5V of headroom so you never go out of regulation.  The LM317 drops about 3V across it when its operating - nothing you can do about this - it just is.

The circuit functions by keeping the voltage across the current setting resistor to always be 1.25V. If the voltage across a resistor is constant, the current through it must be also.  The load voltage will change but the current will not unless your load voltage exceeds the highest operating voltage as stated above.

Offline axkman

  • Participant
  • **
  • Posts: 36
  • Likes: 0
Re: Current and resistance
« Reply #5 on: August 17, 2020, 09:38:11 AM »
Yep its an awsome method.

Also the limitations of LM317 needing atleast +/- 10ma of minimal current to operate correctly can be bypassed to put another resistor on LM317 Vout to ground. That way you can set your "base current" load independend of what the current towards the output to your circuit has to be (towards led, motor, silver electrodes etc).
So you can set a base current of for example 10ma so that whenever you not connect anything to it, you already have 10ma going so the LM317 specification of minimal 10ma current is met.
Then with the resistor between Vout and Vadj you can set the current that you want for your application (silver electrodes etc) to be whatever low current you want. So in that way you can use the LM317 current source for even 0,2 ma if you desired 0,2 ma current thru your silver electrodes. Very easy and cheap way for having very low current source!

Offline Gene

  • Expert
  • Participant
  • ***
  • Posts: 1160
  • Likes: 84
Re: Current and resistance
« Reply #6 on: August 17, 2020, 08:47:05 PM »
Kind of...

If your ballast resistor is connected to the VOut of the LM317 to ground, depending on what voltage the regulator is putting out, the current through it will vary.

The limiter circuit works by maintaining 1.25V across the current setting resistor. That means that VOut could range from VIn-3V if the circuit is powered up and not yet hooked to your cell, to as low as 11.25V if you set your cell to exactly 10V though I'd recommend more like 12V if you can get there - a little more voltage helps rather than hurts.

The issue is, given the range of voltage the regulator can vary over, you need to compute your ballast resistor at the lowest possible voltage.

For safe area, lets say 10V in case your cell coasts a little below 10V during a run.

So, R = 10/0.01 (10ma) = 1000Ohms

OK, fine and dandy BUT when you haven't yet connected the cell and have powered up the limiter, its putting out say 21V if you're powering it with 24V as an example. So, at 21V, a 1k resistor would pass 21milliamps.  Sure, no big deal, or is it?

What about power dissipation in the ballast resistor?  It will be, in the 21V case, dissipating 0.021^2 * 1000 = 0.441 WATTS!  You had better consider using at least a 1 watt resistor unless you want a branding iron instead. A 1/2watt WILL get VERY HOT with this much power dissipation.

Sure, when the cell is operating with 10V across it the current through the ballast resistor would be 11.25v/1000 = 11.25ma where the power would be about 0.127watts but you can't design to this - only to worst case because over time it WILL happen.

HAHAHA, this is sick but it'd work well.  Make another current limiter and set it to say 12-15ma just to be safe using a fixed value current setting resistor and connect Vin of this circuit to the output of the real current limiter regulator (LM317 VOut) and the output of this ballast limiter to ground. Now, the ballast resistor only ever has 1.25V across it and even a 1/8 watt would be fine and even in a situation where the real cell current limiter regulator is putting out 11.25V, you still have plenty of safe operating area for this "ballast limiter" to continue functioning - 11.25  - 3 - 1.25 = 7V.

If for the ballast current regulator current setting resistor you used an 82Ohm resistor, that'd be 15.24ma. The power dissipation of the current setting resistor for this would now be 0.019watts (19milliwatts). Even an 1/8W resistor would do wonders and would stay cool as a cucumber.

There is a simpler solution though.

Go buy an LM317LZ TO92 packaged device and crack out of the box you'll easily be able to go as low as 2.5ma but probably even a bit below 2ma though the manufacturer doesn't guaranteed anything but 2.5ma.  I bought 20 of these at the last aliexpress sale, with shipping, for less than a buck total and yeah, I said for 20 of them.  They're dirt cheap but you do have to shop around a bit because sometimes, onesies will cost you as much as 20-50 (I've seen it - thats why I bought 20).

Offline axkman

  • Participant
  • **
  • Posts: 36
  • Likes: 0
Re: Current and resistance
« Reply #7 on: August 20, 2020, 07:29:08 PM »
Thanks for your reactions man!

Fun things those current regulators!
I've seen someone put up a different solution than putting a resistor load across Vout and Gnd. I'm trying to find that circuit back, he made it as a school project and needed 1 or 2 ma from a normal LM317.
Hope I'll find it.

What do you think about putting an extra load on the output (like output from Vadj and Gnd) like the connections that go to the electrodes. Probably the same problem as you mentioned above when putting a resistor load on Vout-Gnd.

Offline Gene

  • Expert
  • Participant
  • ***
  • Posts: 1160
  • Likes: 84
Re: Current and resistance
« Reply #8 on: August 20, 2020, 08:03:26 PM »
Well, the resistor between VOut and ground will do what you want but depending on the current you're running and the top end output voltage, that resistor might need to be a 1-2 watt one as I'd said, so it doesn't become a toaster. This will NOT affect the final output current because its not in series with the current setting resistor.

If you put the resistor across the cell (limiter out to ground), you will not have ANY idea how much current is flowing through the cell.  Though the resistor is fixed, the cell resistance isn't and as it changes due to accumulation of silver oxide or Colloidal Silver in the water, where the cell voltage will change some, the "split" in current between the fixed resistor and that which is going through the cell will change.  Think about it. It makes sense. As the cell voltage rises, the current through the fixed resistor will too and  now you're getting less through the cell. Vice versa, same thing just more current through the cell. This is absolutely NOT what you want.

The current regulator maintains 1.25V ACROSS the current setting resistor, regardless of load voltage so long as the load voltage isn't so high that the regulator craps out (Vin - 3v - 1.25v). I you don't exceed this voltage (and yeah, shoot to stay several volts below this for safe area), the current regulator will do its thing properly.

The ADJ pin is NOT a ground pin on this device and I absolutely would not suggest grounding it nor putting a load resistor on it. This device is a little different than a normal fixed regulator.  Its the voltage set pin.  To use it as a normal regulator, you use a resistive divider on VOut to cause the ADJ pin to be 1.25V when Vout is what you want.  Without the resistive divider, the ADJ pin is still looking for a voltage of 1.25V and it will do what it needs to VOut to get it.  In the current limiter, its looking for 1.25V and it will adjust Vout to cause its input (ADJ) to be 1.25V lower than Vout. VERY simple.

Offline axkman

  • Participant
  • **
  • Posts: 36
  • Likes: 0
Re: Current and resistance
« Reply #9 on: August 20, 2020, 08:27:42 PM »
Yeah, sorry if it was written confusing, I didnt mean Vadj was ground. Its just voltage setting for voltage regulation or "output" when used in current source configuration.

I'm still searching for that other circuit someone made as school project. He didn't put a resistor across Vout and GND but he did something else so his LM317 was able to precisely manage 1-2ma regulation. When i found it i'll post it here.

Offline axkman

  • Participant
  • **
  • Posts: 36
  • Likes: 0
Re: Current and resistance
« Reply #10 on: August 23, 2020, 11:49:28 AM »
Luckily I changed my resistor between Vadj and Vout of LM317 to a blue bournes potmeter set at about 620 ohm. My output is a steady 2ma now! Without having to fiddle with  setting a base load resistor between Vadj and Gnd. Those LM317 voltage regulator boards from Aliexpress are very decent :)

Then next came the question:
I was wondering yesterday if i should modifify my switching psu adapter from 18v to 50v. (by putting a 50v zenerdiode in it and higher voltage capacitors). Then order a LM317HV as it can withstand 60v  between Vin-Vout. But. I doubt giving my current regulator a higher input voltage is any beneficial. As the current is set at 2ma anyway and because of the electrolyte in the cell the the current increases and the voltage get lowered because of this. The only way to get higher electrode voltage is raising the cathode electrode. So inputting a 50v into my current regulator looks like a non-beneficial effort.

Do you have any thoughts on this Gene?
Like, would it be beneficial for me to modify my powersource to output 50v instead of 18v it does now? I am for 2 ma, so because of the electrolyte the current regulator will lower the voltage anyway to fix the current to 2 ma.