LED Driver with Current Limiter

 LED Driver with Current Limiter

 

Recently my AKAI GX-M10 cassette deck's tape progress lighting failed. I noted that the filament had burnt out and so I contemplated with the notion of making my own, but this time the lighting was to be driven by a simple LED display.

Having arranged and soldered two 1.8mm LEDs in parallel, the next task was to design a circuit to run the LEDs at the recommended diode current of 20mA, at 3V.

Later, after observing that this was probably excessively bright, I then experimented by driving the LEDs with voltages varying from zero to 3v, and noting the brightness. It was then I realised that an ideal solution would be to have a current controller fitted to a LED circuit – thus this transistor based, LED driver circuit with adjustable current control was about to be analysed and tested.

Of course a much simpler LED arrangement in series with a suitable resistor would be sufficient, but the challenge of an alternative 'on line' controller for the LED's brightness was a plus-factor.

A standard current limiter circuit is shown below, together with my labelling of various (conventional) current equations, passive, and active components. 

 

Brief Circuit Analysis

In a working circuit after switch on, the circuit stabilises with both T1 and T2 conducting, and importantly - the higher power transistor T1 controlling the current passing through both diodes entering the collector Ic1.

In a steady state condition - a potential difference is developed across Re, which according to ohm's law is virtually Ie*Re. This voltage Vbe2 (Ie*Re), if increased (via Re → Re+𝜟Re) also increases transistor T2 into further conduction - resulting in the collector-emitter Vce2 voltage of T2 to drop momentarily, thus increasing the current through Rb (a voltage drop), and thereby dropping the voltage at Vbe1. This transient Vbe1 drop then lowers the current Ic1 (and Id), and Ie1, and thus lowers Vbe2. This new Vbe2 voltage is then fed back into the loop again. The new decreased Vbe2 now increases Vce2, thus reducing the current through Rb, and increases Vbe1 again. 

This is a feedback cycle that loops instantaneously until equilibrium current Ie1, Ic1, or Id is established.

There is another factor that also determines the driving voltage across the base-emitter of T1 - Vbe1. And that is the voltage developed across Re (approximately Re*Ie) which effectively adds Voltage Series negative feedback to the loop. 

Hence, there are two influential factors in driving the Vbe1 voltage: the feedback loop containing T2, and the in-series voltage across Re, ie Re*Ie. 

So which then will be the more dominant? - this will depend on the input current to T2 (Ib2), and the current gain (hFE2) of T2.

In summary then, with Re static, we observe that Ic1 is static, and then find that varying Re, ie Re → Re±𝜟Re allows us to control the feedback loop, and with this control Ic1, Id, and thus LED brightness.  

Varying Re is the key to controlling LED brightness. 

 

Detailed Circuit Analysis

 

Basic Circuit Equations - although not strictly neccessary, the circuit could be worked with just a basic understanding of how it operates, and a careful trial and error approach - nevertheless, I propose a mathemetical approach. 

 

It is also of academic interest to me to see how close the solutions to the basic equations are to actual real world measurements.

First consider the circuit without the presence of transistor T2 - this is the equivalent to setting Re=0Ω.

 

Kirchoff's Voltage Law is applied to determine the value of biasing resistor RB. The supply voltage from the AKAI GX-M10 deck is approximately 16.7v, the recommended rated current of each LED is 0.020A, and the transistor T1 (BC142 NPN) is labelled in the picture to have a current gain hFE1=230 (at 0.040A), here labelled as 𝛃1. 

 

 

Inclusion of Transistor T2 (2N2222A)

 

With the addition of T2, the circuit become complex - here I label all relevant currents and voltages.

 

 

Applying Kirchoff's voltage law is the key to success assessment, note again that hFE1 is re-labelled as 𝛃1 in the analysis as it's quicker to write ...

 

 

 

The term Rb represents the input dc resistance looking into the current limiter and controlling transistor T2, that is Rb≡Vbe2/Ib2.

 

Measured, Rb was high and averaged around 98KΩ for Vbe changes of 0.58v .. 0.61v.

 

 

 

In summary - I arrive at a simplfied expression for the collector current Ic1, remembering that each LED is handling Ic1/2.

The accuracy of the expression is going to be dependant on the variables Rb, and 𝜷1 and 𝜷2 which vary depending on collector currents. T2's measured 𝜷2 was relatively static at around 𝜷2~10, while 𝜷1 was more dominant and variable depending on Ic1. 

 

Measured current gains for T1 ranged from 193 to 232. Variations in Vbe1 will be small, somewhere around 0.6 .. 0.7v?, this won't affect the overall result since Vcc is dominant in the expression.

It was decided that Re was not going to vary much, between 0Ω .. 100Ω was sufficient enough to control Vbe2.

So then, setting Vcc=16.9v, Vbe1∼0.65v, RB=98Ω, Rb∼98KΩ (measured average), and setting 𝜷1=230, I present the plot below where - 

 

GREEN is the theoretical or formula Ic1 vs Re,  and  

RED is the actual measured Ic1 vs Re data.

 

Current Limiter

 

As the title suggests, this is primarily a current limiter circuit which earned its title when I shorted the path of the LEDs - almost no change in current was noted. The current is limited by the setting of Re.

 

Power Dissipation in BC142

 

Maximun power dissipation in the BC142: Pmax ~ Vce1*Ice1 ~ 14*0.04 = 560mW.

 

Although the BC142 can handle ~500mW, setting Re=0 permanently is not recommended due to the heating up of the BC142. And with no effective voltage-feedback (Ie*Re) to limit increasing Ie, thermal runaway will follow. Recommended that Re is set to at least 25Ω.

 

Power dissipation in Re: Ie^2*Re

 

At 5mA, Re=100Ω: 0.005^2*100 = 2.5mW   

At 26mA, Re=20Ω: 0.026^2*20 = 13.5mW.

 

No problem there.

 

Other Supply DC Voltages

 

The above circuit will work well anywhere from about 9v to 20v dc, and still be able to maintain the same LED current: ±2 mA at worst, but usually ±1.5 mA or less. If you wish to drive your own circuit from a higher voltage, say 30v, you may need T1 to be capable of handling 1 watt constantly as a precaution, and look for a higher Vceo rating.

Current Limiter Fitted to the AKAI GX-M10

The circuit has been tested and fitted to the AKAI GX-M10 - 100% success!

The 'earth' (0v) side has been secured - the initial problem was finding an easy accessible 16.7v point which I eventually did.

 

This LED Driver with Current Limiter blog/article is periodically updated and corrected mainly due to mistakes, typo errors, blogspot.com editor bugs, and will be updated without notice.

13/1/2021, 14/1/2021, 16/01/2021. 17/01/2021, 18/01/2021, 20/01/2021, 21/01/2021, 22/01/2021.

cassettedeckman@gmail.com