2.2V 60A current regulator

[arhi]

Hi colleagues, long time no c .. I was busy with reprap.org project so was bit more into software and mechanics then into electronics for quite some time now ...

At the moment, I'm trying to make a driver for 40W laser diode (max voltage 2.2V, min current 11A, max current 60A) and I need to make a current regulator to drive it.

I need to
- have max voltage limited at 2.2V
- have fine tuned current regulator between 10 and 60A

I have
- 10 of BUV48 transistors
- 2 MJ802 transistors
- 100 0R47 5W resistors
- 10 47R 50W resistors
- 5 2R2 100W resistors
- 9VAC 60A source
- 5VDC 60A source
- 3v3DC 50A source
- 3VAC 50A source
- whole bunch of other stuff that is not "high current"

For the "fixed current source" I'd go with simple opamp driving paralleled transistors with leads over shunt resistor.

VIN +------------\ />----[ LOAD ]---+---[ SHUNT ]---GND
    |            ---                |
 [ RES ]          |                 |
    |          [OP-out]             |
    +-----[ OP+ ]    [ OP- ]--------+
    |
  zener
    |
   GND

and then the current trough LOAD would be V(zener)/R(shunt) but since I want to go with 60A I cannot make Rshunt variable and I want to have fine regulation of current between 10 and 60A, and I don't know a pot that will pass trough 60A nor one that is in the ballpark of 0.1R or less...

What I figured so far is that putting few (let say 6) BUV48 in parallel and driving them with another buv48 I will have the "driving" power. Using 20x5W 0R47 will get me shunt with low resistance and high power. What I don't know how to do is - how to drive that initial BUV48 to get me the current I want. Mine initial idea was just to use 10K pot as a voltage divider (ends on + and gnd and center to base of transistor) to act as rudimentary voltage regulator, but the problem here is that small voltage changes can cause big jumps in current and that's not what I can afford (as I'm driving expensive laser with it).

Additional problem I didn't even scratch is whether the 2V will be enough to even open transistors ...

Anyone have any ideas?

 

[Grossel]

How long period of time will the laser stay on?

Here's two additional ways you may solve the problem:

1. Several independent constant current regulators. Pros: You can turn them on/off independently and then regulate current in several steps + distribute power loss over several transistors. Cons: Waste lot of energy by heating up the transistors.
2. Build a Buck converter. Then you can take out fairli high current and low voltage - as oposite to input, higher voltage and lower current.

 

[arhi]

how long - hard to answer .. initial testing will be done with few seconds at a time, but finally it should work for hours ...

not sure what to think about several independent constant current regs - it's not that I can stack them like I can stack voltage regulators to increase voltage ?! What am I missing?

on the buck idea, this is something I 100% have no idea how to do... the whole chopping thing is way out of my reach (I'm proficient in software, digital electronics, mcu's .. the analog, high current, hight power stuff is soooooooo out of my area of expertise ) and iirc the whole buck thing even if you have a working schematic need to be "fine tuned" and for that you need level of exp that eludes me attm...

The direction attm is using bunch of BUV48 in parallel opened by op-amp fetching voltage from the shunt resistor... but I'm open for better ideas (and pointers to explanations as, to mention again, not too proficient in this )

 

[bychon]

Thinking out loud...I could see this as a 2.2 volt voltage regulator with huge current capabilities, but you want the regulation on the current aspect. WT? Doesn't your load have a positive resistance?

I have a current regulated source (attached) that would form the basis nicely. Add an op-amp to throttle it down in an analog way when the output gets to 2.2 volts. How to control the amps is a variation on this circuit. Do you want infinite control or chunks of amps that are switchable?

Keep talking. I don't have enough to work with.

 

[ronsimpson]

https://www.electro-tech-online.com/custompdfs/2010/05/3731Hfb.pdf
Data sheet and evaluation board. 5 to 7 volts in 2 volts out. Very little power loss.

 

[bychon]

Yep. That oughta' work.

 

[arhi]

@bychon, it's not that simple as simple positive resistance, it's a laser so it has variable resistance depending on the output efficiency factor at different power levels, heat etc.. so with a simple voltage regulator a small increase in voltage might make a big increase in current, the ripples will be too big. It's not a simple heater I'm powering - so it has to be a current controller, not voltage one.

@ronsimpson, this looks like a solution, the only problem is - I never worked with buc's before so for starters after reading datasheet twice I still have no idea how it works but that should not be a problem, I will read it few more times that usually helps. I need to figure out how to incorporate feedback loop (from the shunt/sense resistor back to buck so I do not have voltage but current regulation). I'll update the thread if I manage to do something

 

[colin55]

The simplest way is to use a single cell and a small-value resistor. Just put many rechargeable cells in parallel.

 

[arhi]

The simplest way is to use a single cell and a small-value resistor. Just put many rechargeable cells in parallel.

single cell of what? I need regulated current between 10 and 60A ?!

https://www.electro-tech-online.com/custompdfs/2010/05/3731Hfb-1.pdf
Data sheet and evaluation board. 5 to 7 volts in 2 volts out. Very little power loss.

I went trough data sheet multiple times. I understand what it does and the general idea of how it does that, but the problem is - there is no way I know how to make regulated output with it. There are examples how to make stable voltage output with it, and to change voltage output (from 1.2V to 5V for e.g.) a whole bunch of things are different (different L's, different C's ...) .. and I need to have regulated current .. so either I'm missing something obvious or ..

 

[colin55]

A single "battery" is called a "cell." Use AA cells or larger and you will be able to meet the 60A current demand. You can also be charging the cells at C/10 or C/15 and not overdrive the laser diode. It is far too dangerous, with your limited knowledge of electronics, to be making a 60A supply. This is a very technical specification.

 

[arhi]

I really hoped you are not thinking about batteries. In order to get 60A using AA's I need 30 cells and 2A discharge current will pretty fast kill em .. so to use more moderate discharge current I'd need 60 cells .. that;s kinda lot

I'm not "non-technical", just that psu design is really not something I ever did before ... I'm more proficient with low power and ultra low power electronics, mcu's etc .. this high power stuff is really not my thing . Anyhow, I don't intend to kill myself with it anyhow so no worries if I don't fully understand the inner workings, I ain't gonna play with it... for now, the simple current multiplier with bunch of BUV48 will do the trick for testing and later on I'll probably get some commercial led driver.

btw, more I think about it, more I like your idea with batteries 30x2700mAh would be able to give me 60A for 15-20 minutes (iirc when they discharge at high rate as 2A they loose charge exponentially), that is more then enough for test I want to perform and I don't have to worry about cooling 10 transistors

 

[wkrug]

I think the way is a switched Current souce.
Most of the popular chips are designed for voltage regulation.
But you can change that into current regulation, when the Fold Back Pin will get the right voltage driven by current.
If th Fold Back Voltage is specified with 1,76V, so you'll put a 0,16Ohm resistor into ground line.
The resistor will heated with 20W.
That is'nt very effective, so you'll put a little operational amplifier between the shunt and the Fold Back pin.
So you can lower power loss at the shunt resistor.

This solution have 2 little problems.
1. If the line to the laser ist open, the Voltage increases to very high values.
I solved this problem by setting a second voltage divider with an decouple diode at the Fold Back pin.
2. By amplyfy the Fold Back Voltage ( Current ) it is possible the voltage regulator start oszillation.
That can only be tested. Some capacitators at critical points shoud be help to avoid this problem.

 

[arhi]

I think the way is a switched Current souce.
Most of the popular chips are designed for voltage regulation.
But you can change that into current regulation, when the Fold Back Pin will get the right voltage driven by current.

The problem here is that I don't understand how this buc works. From what you wrote, and from some idea I have after looking at the data sheet, the output voltage is held at a desired level with combination of (no idea how exactly)
- voltage on UVADJ
- voltage feedback from vout on EAIN

the foldback is the EAIN part... now if I understand how the bloody thing exactly change behavior of the device I might understand how to change this buc to be current regulator and not voltage regulator - not to mention, I need to be able to attach a bloody POTIO to be able to manually change the desired current output - and, well, I don't see how to do it still .. (yeah I know, colin said already, I'm stuped).

I'm thinking that maybe using rsense bit differently can help me .. but .. I'll find a buck and play with it .. things will get much more clear

 

[Damo666]

Forget AA cells. What about using D size cells? They come in capacities around 10Ah or more nowadays.

 

[arhi]

rechargeable D size cell's cost arm and a leg over here, it will be cheaper to get someone to design and make the regulator for me ..

one other thing .. the rechargeable batteries are 1.2V (NiMh), I need max 2.2 (2.0 is more likely), lasers are very sensitive here, 2.4 could kill it and 1.2 might not be enough to push 11+ amps (11A is minimum threshold under which laser is off)

 

[colin55]

The cheapest and best thing is to use a single cell from a 12v car battery. Simply tap into the bus with a drill and screw a terminal to the end cell. Add a small resistance and connect to the laser. The resistance must be included, even though the cell is 2v2 and the laser is 2v2.

 

[wkrug]

IMHO the akku is not exact enough for driving this laser circuit.
An fully charged NiMH Akku has an voltge of 1,35V or more.
When it will be discharged the voltage get down until 1...1,1V.
So at the laser is never a stabile voltage availible.

I try to explain how a simple switching power supply works:
There are some different methodes how to create an step down converter.
At least the simple ones work with an pulse width signal on the output.
This PWM signal drives a switch ( e.g an P-Channel MOS-Fet )
When the output voltage goes down, the ON time of the PWM Generator will get a little bit longer.
The Fold Back pin get the output voltage information by dividing the output voltage with an resistor network and regulates the pulsewidth of the PWM generator.
The inductor at the output of the switching power supply will hold the current while the switcher circuit is not active. To make the circuit complete there is a diode needed that take the current back to inductor in OFF times of the power switcher.
If the power switcher is on this diode is closed becuse the voltage is reversed at the diode.

If you'll send back this fold back voltage driven by current, you get an constant current regulator.
The problem - When no load is connected at the output there is no current present - thats clear.
So the voltage will exeed until some parts will be damaged.

You'll avoid them by building a second way to the fold back pin of th switching network IC that will stop the voltage overshot at an fixed maximum level.

I think by the high current you need, there are only switching IC with an external power switch rational.

 

[colin55]

I don't have the expertise to make a 60 amp switching regulator circuit, and that's after 40 years in the profession. It's absolute madness to even think you are going to be able to make a circuit as complex as this. And to be goaded by others into thinking this is a possibility; is senility.

 

[Diver300]

It might be possible to use the buck regulators on a PC motherboard. The current and voltage needed for a CPU core are about what you want.

This application note is a bit old https://www.microsemi.com/micnotes/1301.pdf but the principles still apply.

 

[arhi]

I don't have the expertise to make a 60 amp switching regulator circuit, and that's after 40 years in the profession. It's absolute madness to even think you are going to be able to make a circuit as complex as this.

colin, don't get me wrong I'm not trying to make a "final solution" - I'm just trying to make a 60A 0-2V regulator that will work "few seconds" - nothing more... as the whole idea is to just test what I can do with this laser in few seconds.... if the results are ok, then I will pay someone to make me a regulator or purchase commercial solution. I do not intend to make a final solution myself as I kinda know my limits

Thing is - in theory a simple current injection into base of transistor and multiplying that current using power transistors should get me quick & dirty solution that will (on a big alu fan cooled cooler) provide regulated current source that will work for few seconds... Anything more then that is not expected nor required at this moment. The whole point is I do not want to spit few hundred bucks for a regulator if I don't test if laser strength is enough for what I want to use it, but I don't mind wasting 40-50 euros worth of transistors ... The solution with transistors will require "lot of them" and they will produce "a lot" of heat. The laser diode is also producing a lot of heat too (~120W) and I have not even begun to work on the cooling system ... but .. as I said .. for 2-3 seconds a simple "big aluminium cooler" will get the job done...