Maker Pro
Maker Pro

Simple current circuit regulator issues

Allenvik

Jan 25, 2016
9
Joined
Jan 25, 2016
Messages
9
Hi everyone,

First of all, thank you very much for this platform ! I am interested in electronics, and I am trying to learn some or let's say relearn it, as I had some knowledge about it 20 years ago, but long forgotten.

Anyway, I am using this circuit for a diy hobby application for some time, and I have attempted to improve it without success.

This simple current regulator circuit is supposed to provide steady and a smooth current up to 3.5mA, for an applications' resistance which is around 5kΩ, but that can go up to 20kΩ. The circuit has two regulations, one small trimpot for fine regulation which I use a 100kΩ and an 50kΩ linear potentiometer to change the output current. Recently I downgraded the pot from 100kΩ which seemed better to operate.

Currently, the circuit is very stable once the current is set, due to the LM317, and is reasonably smooth with the 47μf capacitor. But the regulation and the fine adjustment is not good.

Since the final current was steady, I didn't bother to fix it, but finally I am trying to get it right once it for all.

I will appreciate any ideas to improve this.

Many thanks for your time and kindness!

P.S: I made a diagram picture of the circuit, please check it below (if shows up on you). It's basic, very simple stuff... but so far 10 to 1 for this bugger on me :)

Simple_current_Circuit_regulator_issues_01.png
 

Bluejets

Oct 5, 2014
6,901
Joined
Oct 5, 2014
Messages
6,901
To begin with , it is much easier to read a circuit diagram if it is drawn in a conventional manner such as below.
With regard to your circuit, I think you will find with the addition of the large value resistors, you are operating the LM317 beyond it's ability to regulate.
Think you'll find a minimum of around 10mA might be the standard.
 

Attachments

  • 317.jpg
    317.jpg
    49.7 KB · Views: 132

AnalogKid

Jun 10, 2015
2,884
Joined
Jun 10, 2015
Messages
2,884
There are a few issues with your circuit, and all of them are covered in the datasheet. A 317 will not regulate down to zero anything, zero volts or zero milliamps. This is because the little bit of current that the 317 needs to function flows through the load. Also, somewhere on the datasheet is the circuit for turning the 317 into a current regulator, including the equation for calculating the output resistor that sets the current. If you run the numbers you will see that 100K is way too big. Finally, if the current regulator is workng correctly the 50K output resistor is not needed and will not affect the output current.

ak
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
On top: 3.5 mA through a 20 kΩ load requires 3.5 mA*20 kΩ = 70 V therefore the 18 V source is totally insufficient.
The LM317 is suitable for an input-output voltage differential of max. 40 V.
At the minimum load of 5 kΩ the voltage drop across the load will be 14 V. This allows for a max. input voltage of 14 V +40 V to the LM17. There is no way this circuit can drive 3.5 mA through 20 kΩ!
 

Allenvik

Jan 25, 2016
9
Joined
Jan 25, 2016
Messages
9
To begin with , it is much easier to read a circuit diagram if it is drawn in a conventional manner such as below.
With regard to your circuit, I think you will find with the addition of the large value resistors, you are operating the LM317 beyond it's ability to regulate.
Think you'll find a minimum of around 10mA might be the standard.


Thank you very much for taking the time to share some knowledge ! I understand...It never occurred to me that the problem to begin with was that I was using the wrong current regulator perhaps...:(

I have being using this setup for a while, and like I said it works but is not precise which is what I want do improve now...

So the 317 is not appropriated for the intended purpose I guess, I had a look in the data sheet as wisely pointed out and I found that the 'min load current for operation is 3.5mA' (if I am interpreting it right), and this level is the ceiling of my required output current, so I am amazed that is working the way it is.

I am now looking for a substitute for the LM317 that could work with lower current loads, any suggestions? I found the LM2936, could that be an better option? but I am confused with the specs... i couldn't find the info for a min current output or current load...

Just to recap, my circuit was supposed to be um adjustable current supply from 0 to 3.5mA, for loads around 5k to 20kohms with a supply input between 9-18 volts. I also though about a charge pump? but I couldn't find anything under my requirements so far...

LM317 data
http://www.ti.com/lit/ds/symlink/lm317.pdf
LM2936 data
(http://www.ti.com/lit/ds/symlink/lm2936.pdf)

LM317_specs.png
 

Allenvik

Jan 25, 2016
9
Joined
Jan 25, 2016
Messages
9
On top: 3.5 mA through a 20 kΩ load requires 3.5 mA*20 kΩ = 70 V therefore the 18 V source is totally insufficient.
The LM317 is suitable for an input-output voltage differential of max. 40 V.
At the minimum load of 5 kΩ the voltage drop across the load will be 14 V. This allows for a max. input voltage of 14 V +40 V to the LM17. There is no way this circuit can drive 3.5 mA through 20 kΩ!

Hi Harald,

Thank you so much for taking the time !

I have realized that now, I can't believe I was so naive and haven't check the basic ohms law calculations... you are completed correct! The circuit works better with lower resistance, and the current declines a lot when the resistance is increased.

Regarding the higher out put voltage, I would have to up scale it some how, I think this would take another circuit just to do that.

As I am searching for solutions, I came across 'charge pump' components which upstream voltage, do you see this is an option? also micro-controllers maybe? but I would had to learn how to program these... and I am struggling with basics so far:(

I was thinking in redesigning and increasing the voltage supply, however, the max safe working voltage for the load is around 18 volts... so ohms law will always prevail ? meaning independently of which circuit or trick I do on the other side, if I want a supply output of 3.5ma for a 5k-20k ohms load I will definitely need a 23-90v output?

Thank you
 
Last edited:

Allenvik

Jan 25, 2016
9
Joined
Jan 25, 2016
Messages
9
There are a few issues with your circuit, and all of them are covered in the datasheet. A 317 will not regulate down to zero anything, zero volts or zero milliamps. This is because the little bit of current that the 317 needs to function flows through the load. Also, somewhere on the datasheet is the circuit for turning the 317 into a current regulator, including the equation for calculating the output resistor that sets the current. If you run the numbers you will see that 100K is way too big. Finally, if the current regulator is workng correctly the 50K output resistor is not needed and will not affect the output current.

ak

Thank you for your reply !

I have realized the fundamental fault with the circuit, basically I need a new controller and so a new circuit to be precise. I have checked the LM317 specs and you are correct to point out the discrepancies regarding the intended out come for my basic circuit. I am now looking for alternative solutions. The main objective of the circuit is to regulate current between 0 and 3.5ma, considering a load resistance between 5-20 kOhms with an input voltage of 18v. Now as others pointed out, there is no way to upstream the output voltage from 18v using this controller, I would need so other ways to do that, 'charge pump' maybe? However I was considering increasing the voltage literate, however this is not a good idea I guess due to the nature of the load, which safely would take maximum 18 volts...
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Another issue is that for a constant current supply, you don't want a capacitor across the output.
 

Allenvik

Jan 25, 2016
9
Joined
Jan 25, 2016
Messages
9
Another issue is that for a constant current supply, you don't want a capacitor across the output.
Hi steve,

Thank you for your reply !

This simple circuit is working for my application for some time, imperfect but it works and is very simple, the capacitor was added after tests done in which I found that the current change was definitely more smooth with it, in fact a higher capacitor was even better, but had other unwanted effects. Eventually the small capacitor was a good thing for my application, perhaps I forgot to mention that the final load is variable.

I think the best solution to upgrade this, so far is to substitute the original circuit for a more efficient one and then joint it perhaps with another circuit to increase voltage (dc boost converter? 18-50v) so we can get the current we need on the higher load band.

So I still looking for a substitute for the 317 that can handle 0-3.5mA more efficiently...

Thank you
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
You're probably best looking at constant current circuits based on an op-amp, a voltage reference and a current sense resistor.

A circuit like this will allow linear variation of current down to zero.

Google "variable current source".

This is a simple version which may be sufficient. http://www.merghart.com/p/28/opamp-adjustable-current-source
 

Allenvik

Jan 25, 2016
9
Joined
Jan 25, 2016
Messages
9
You're probably best looking at constant current circuits based on an op-amp, a voltage reference and a current sense resistor.

A circuit like this will allow linear variation of current down to zero.

Google "variable current source".

This is a simple version which may be sufficient. http://www.merghart.com/p/28/opamp-adjustable-current-source

Hi Steve,

Thank you for sharing your knowledge.

I am looking at the info you sent, and it seems very interesting indeed. I let you know later. Many Thx
 

Allenvik

Jan 25, 2016
9
Joined
Jan 25, 2016
Messages
9
You're probably best looking at constant current circuits based on an op-amp, a voltage reference and a current sense resistor.

A circuit like this will allow linear variation of current down to zero.

Google "variable current source".

This is a simple version which may be sufficient. http://www.merghart.com/p/28/opamp-adjustable-current-source


Hi Steve,

I wanted to thank you for sending that current regulator circuit link from Ben's site. I built it in a trial board and it worked like a champ !!! It did everything I was looking for, it starts from zero and can supply very well and stable current for the loads I use.

I am in the process of tricking the resistors and introducing couple of tripots for fine regulation.

So this was in deed the solution I was looking for initially !

Thank you very much !
 
Top