How set the current on a DIY power supply

[stu808]

Hiya all,

I'm in the planning stage of building a new DIY power supply and one of the features I wish to incorporate is current adjustment/control; what I'm trying to work out is how to set the current.

Obviously when you "dial in" the potentiometer controlling the voltage the panel meter reading the voltage will display the set voltage as the voltage is a constant (yes/no?). However when you turn the potentiometer for the current adjustment the ammeter won't move until you put a load on the power supply; as the current is not constant (yes/no?).

What I'm looking for is a way to temperately put a load on the output. So far all I can think of is putting a momentary DPDT switch on the output that when pressed moves the output to the "load" and lets me dial in Amps before using the power supply. But what do I use as a load? This power supply is only going to have a maximum output of 75 watts (24V-3A), but this is a test bed for bigger unit later in the year... 300 watts.

The guy at the local electronic parts store said to wire in a halogen bulb or similar, but that seems a bit rough and ready and I don't think a halogen will take too kindly to a 30V-10A load for the next power supply. Is there another way?

Many thanks,

Stu
P.S. This isn't a homework question, I'm a 36 year old who likes scratch building his own guitar and RC stuff...

 

[alec_t]

A common way to control/limit current is to pass the load current through a low-value (e.g. 0.1Ω) sense resistor, compare the resulting voltage across the resistor against a reference voltage, then switch off or 'turn down' the supply according to the comparator output. Varying the reference voltage enables the current limit to be varied. Providing the resistor value and reference voltage are accurately known you don't need a dummy load to set the limit (though it's useful for test purposes).

 

[MrAl]

Hi,

Welcome to the forum first of all. I think you're in the right place.

I understand your problem pretty well. You have a power supply and you want to use it to run something, but first you need to set the current level so you dont accidentally blow out the device being tested. It's pretty hard to do with a dummy load because the dummy load ends up being as big as the power supply. So you want an alternate method.

The simplest answer is if you dont need super accurate current setting (usually you dont) you can add a calibration plate to the power supply right behind the current set knob. It would be a round circular pattern like a clock, but instead of hours and minutes, you print amps and milliamps. So when you go to set the current level you dont need a load, you just dial it to the right value as you read off the plate.

Another method if you are building your own supply, is to use a dual ganged pot for the current control. The front pot portion is used for the normal current setting, and the back portion is used for viewing the setting of the pot on a digital voltmeter built right into the power supply. The calibration is such that the volt meter reads out in amps rather than volts. So when you turn the pot the volt meter gets say 1.000v and because of the mechanical linkage between pot sections and your predetermined calibration, that means 1.000 amps. This would be easy to implement provided you can get a dual gang pot.

 

[crutschow]

..........................

Obviously when you "dial in" the potentiometer controlling the voltage the panel meter reading the voltage will display the set voltage as the voltage is a constant (yes/no?). However when you turn the potentiometer for the current adjustment the ammeter won't move until you put a load on the power supply; as the current is not constant (yes/no?).

What I'm looking for is a way to temperately put a load on the output. So far all I can think of is putting a momentary DPDT switch on the output that when pressed moves the output to the "load" and lets me dial in Amps before using the power supply. But what do I use as a load? This power supply is only going to have a maximum output of 75 watts (24V-3A), but this is a test bed for bigger unit later in the year... 300 watts.

.........................................

I assume you mean "current limit" when you say current adjustment. If so then you can just short the output with a push-button switch to adjust the current limit. But, of course, the power supply needs to be able to handle a short-circuit without failure for this technique (which a good power supply design will do). I have used laboratory supplies that use this technique.

 

[KeepItSimpleStupid]

You can do it with a read/set button for both voltage and current. In the set mode, the supply just reads the setpoint of the current and appropriately scales it to engineering units.

You have to decide if your making a CC/CV power supply or a CV supply with current limit. If the latter you have to decide the type Foldback or Current limit.
In a foldback supply if a 10 A supply was to deliver 10A, and the output was shorted, the voltage would nearly drop to zero and the current would be 10A.

 

[dr pepper]

I posted a schematic for a psu a while back, it uses an integrator to limit current, the op amp integrator robs current from the base drive of the o/p tranny, the design uses a classic lm723, but the display is an lcd and shows volts,current, watts and the heatsink temp.
Care in design is important if you dont want noise on the supply o/p.

 

[stu808]

Cool thanks for your replies so far... please bear with my replies. I call my electronics background to date self-taught, no... badly self-taught

Also had an idea today at lunch toady, shorting the output over two 50W resistors in parallel (because 50W resistors are $3 each and 80W and above are uber-$$$). And if my dodgy napkin math is right, I'd need a resistance of 8.2Ω, but as I'm running two in parallel I'd need two 18Ωs or one 18Ω and one 15Ω (8.182Ω). Unless that logic is flawed.

A common way to control/limit current is to pass the load current through a low-value (e.g. 0.1Ω) sense resistor, compare the resulting voltage across the resistor against a reference voltage, then switch off or 'turn down' the supply according to the comparator output. Varying the reference voltage enables the current limit to be varied. Providing the resistor value and reference voltage are accurately known you don't need a dummy load to set the limit (though it's useful for test purposes).

Excuse me if I read that incorrectly, the circuit I'm using already has a current adjustment in the design. Unless I'm missing something?

Hi,

Welcome to the forum first of all. I think you're in the right place.
I understand your problem pretty well. You have a power supply and you want to use it to run something, but first you need to set the current level so you dont accidentally blow out the device being tested. It's pretty hard to do with a dummy load because the dummy load ends up being as big as the power supply. So you want an alternate method.

I know what you mean.

The simplest answer is if you dont need super accurate current setting (usually you dont) you can add a calibration plate to the power supply right behind the current set knob. It would be a round circular pattern like a clock, but instead of hours and minutes, you print amps and milliamps. So when you go to set the current level you dont need a load, you just dial it to the right value as you read off the plate.
Another method if you are building your own supply, is to use a dual ganged pot for the current control. The front pot portion is used for the normal current setting, and the back portion is used for viewing the setting of the pot on a digital voltmeter built right into the power supply. The calibration is such that the volt meter reads out in amps rather than volts. So when you turn the pot the volt meter gets say 1.000v and because of the mechanical linkage between pot sections and your predetermined calibration, that means 1.000 amps. This would be easy to implement provided you can get a dual gang pot.

I did consider that first option and it would be fine if I was using a standard 270° pot. But I'm using a multi-turn pot and that sends the scale idea out the window. Another option I'd like to have is a two pot setup, one for course adjustments and one for fine tuning; but that's another post me thinks.
The second option might be 'do-able' as I already employ a similiar setup on one of my homebrew radio control transmitters, in an attempt to show servo throw/rate.

I assume you mean "current limit" when you say current adjustment. If so then you can just short the output with a push-button switch to adjust the current limit. But, of course, the power supply needs to be able to handle a short-circuit without failure for this technique (which a good power supply design will do). I have used laboratory supplies that use this technique.

Sorry, that just gives me the hebejebes. I did consider shorting the output with piece of aluminum rod but thought there must be a cleaver way.
The circuit says "Current Adjust Range" which I thought was a current limit (based on my first PSU which came in a kit form MANY years ago). Does Current Adjust Range mean something different to Current Limit? And... now that I've looked at the circuit it has a third variable I didn't notice before called "Full Adjustment". Now I'm scratching my head, perhaps I should just built it and find out what it means in my normal fashion; the hard/hot/smokey/"Well that didn't work" way.

You have to decide if your making a CC/CV power supply or a CV supply with current limit. If the latter you have to decide the type Foldback or Current limit.In a foldback supply if a 10 A supply was to deliver 10A, and the output was shorted, the voltage would nearly drop to zero and the current would be 10A.

Bear with me again to make sure I'm reading this right; by CC you mean Constant Current? And CV = Constant Voltage?

Stu

 

[KeepItSimpleStupid]

Yep for CC and CV.
They imply regulation.

In CC the V control sets the compliance V.
In CV the I control sets the max I.
Modes usually auto-switch depending on setpoints and load.

In CL supplies (current Limit) regulaton is not implied.

There is another PS out there called a 4-quadrant supply. These operate under V+I+, V-I-, V-I+ and V+I-. They guise themselves under different names. You can get supplies that will do: constant V, Constant I, constant P and constant R.

Bench supplies are usually the first 2. A battery connected to them (charger) will kill the supply without a series diode.

 

[crutschow]

................................

You have to decide if your making a CC/CV power supply or a CV supply with current limit. If the latter you have to decide the type Foldback or Current limit.
In a foldback supply if a 10 A supply was to deliver 10A, and the output was shorted, the voltage would nearly drop to zero and the current would be 10A.

You described a standard current limit which maintains the current independent of the output voltage. A foldback limit reduces the current as the voltage drops to limit the current and dissipation in the worst-case short. In that case the short-circuit current could be only an amp or two.

 

[KeepItSimpleStupid]

Yep, discussed earlier.

 

[crutschow]

Yep, discussed earlier.

Yes it was. How is that pertinent to my comment?

 

[KeepItSimpleStupid]

Something went awry. What? Who knows?

What's more important is that we may be talking "power supplies" at the "wrong level".

We haven't got past that the setting of a "current limit" or "constant current" can be at the virtual level and not involve a real load.

Which I think means something is amis.

To the original poster: Ask questions - we don't bite!

 

[alec_t]

the circuit I'm using already has a current adjustment in the design.

Can you post the schematic? Then we'll have a better chance of solving your problem.

 

[WTP Pepper]

Google Thurlby power supply. There are schematics that show you how they do this.