Hi Peter - This looks like fun. Thanks for caling my attention to it.

John

Hello John,

Although I don't have your circuit (you could post it) I suspect that your constant current source won't work at low supply voltages because you are using  a 741 op amp. This device was first designed and sold by Fairchild in 1968 so the design is 46 years old. Things have moved on.

Looking at the TI data sheet (SNOSC25C –MAY 1998–REVISED MARCH 2013):

The part is only specified for operation with +/- 15V supplies - it'll work with less but they don't guarantee anything.

The input voltage range is worst case +/- 12V - this means that if either input gets within 3V of the supply rails then the amplifier won't work properly, so the least supply you could ever use is 6V.

The output voltage swing is worst case +/- 12V so once again it can't drive near the supply rails.

For your current source application pick a rail to rail input and output op amp (often described as RRIO in data sheets) so that you can work with inputs and outputs near the supply rails. Pick one which is specified to work with a 3V or less supply, (but OK with maximum).

You could consider the Microchip MCP6281 but watch out for the maximum 5.5V supply.

It is getting hard to find the spec you need in dual in line packages - there is much more choice in surface mount.

Make sure the MOSFET in the current source will turn on enough with the gate voltage you have available.

Image is for illustrative purposes only.
Please refer to product description

 

Manufacturer:

MICROCHIP

Order Code:

1439456

Manufacturer Part No:

MCP6281-E/PMCP6281-E/P

MK

Hi Michael -Thanks for your insight. You are surely correct. I took the circuit esign from an example in "Practical Electronics for Inventors" and I am sure it wasn't intended for a 3.7 volt source. Your comment is exactly what I like about this site. I learn new things and am starting to progress from the 46 year old technology that has been the basis of my electronics for too long. I lament the loss of the through the hole components as they were easy to work with but I am finding ways to adapt to the surface mount. Hopefully I can learn to use the surface mount before they come out with the next big paradigm shift in design. I will research the MCP 6821.

John

Hi Peter,

I took your advice and got some of the LM3410 LED drivers to play around with. I set up the circuit and it worked perfectly. What a neat little IC. I always appreciate your wisdom. I also got some of the MCP6281 that Michael mentioned in his post. I will play with those next.

Thanks John

I am only to happy to help and happy it is working out for you, they are certainly easy to use and have a minimal BOM to get up and running

show us what you build , could be of interest for others

Hi Peter,

Most of my builds are very simple and temporary. I find that I can read the data sheet and study the part in question but until I actually build a test circuit I am not sure I really understand what is going on. I will try to  add some more pictures and diagrams to my posts in the future.

John

Hi Michael,

It has been two weeks since you told me about the MCP6281 Rail to Rail Op Amp. I just want to tell you that I got some from Newark and have wired one into the constant current circuit that I was trying to make work with the 741. It worked beautifully. I have no present application but I learned a lot from the exercise of trying to help Dave Bailey with his LED question and from the follow up that you and Peter gave me. Here are two pictures of my test setup. One shows the MosFET gate voltage and drain current with one LED and the second picture is of the same setup with a second LED in parallel with the first. The drain current remained rock solid at my preset 4.7 mA.

Thanks Again for your expert insights.

John

I'm happy that it's working for you. Do you have an oscilloscope and  pulse generator. If so it can be very interesting to look at the current (by looking across the current sense resistor)  while applying step changes to the current demand. Sometimes you get a lot of ringing and even oscillation - then you can experiment by adding a capacitor in the right place to control it.

MK

I thought id better come and apologize for not posting in a while. I recently lost my job, so i  had to put things on the back burner for a while, which included the internet and lots of other non essentials. including my spare time lol, temp work is still work.

So the wife and myself had a talk and she came up with an idea  "Go to Uni and study Electrical/Electronic Engineering"  (SHHHH don't tell her but i think she is a GENIUS) so i applied and a day later they had me sat in a lecture hall.

thanks for all the help guys it is much appreciated

Dave

Well Now you have even more reason to come back for help… Assignments, Homework, Labs etc etc