Linear Current Modulators
for high-power LEDs

About this project:

After constructing the Pulse Width Modulator for High Power LEDs I needed to build another LED modulator for another optical transceiver.  For this project I decided to take a different approach and use solely linear techniques for the audio modulation.  Like the PWM circuit, this circuit also uses the "precision current sink" to regulate set the LED current but in addition to the linear modulator, a tone generator - based on the same source code as in the PWM circuit - was added to facilitate testing and aiming.

In order to linearly modulate the intensity of an LED it is necessary to vary the amount of current flowing through the device rather than the voltage across it.  It is fortunate that the luminous output of an LED is linearly proportional to the current flowing through it:  At higher currents, the "current versus light output" curve "flattens" a little bit, but this results on only a very small amount (a few percent at most - see the sidebar below) of distortion and is unnoticeable in voice communications.
Figure 1:
Schematic of the high-compliance current sink, used as a high-power LED modulator.
Click on the image for a larger version.
Simplified schematic of current sink to
                    modulate a high-powered LED.


First, a bit of explanation as to how the circuit works.

The "Precision Current Sink":

Several means have been devised to modulate high-current LEDs, variously using bipolar and power MOSFET devices, several examples of which may be found on the page "The 'Luxeon':  New Light of Hope for Optical Comms".  I decided to try a different approach - the Precision Current Sink - which is very similar to some types of current sources.

The basic current sink circuit may be seen in Figure 1.  This circuit uses a single section of an operational amplifier "wrapped around" a transistor and using a current sense resistor and it works thusly:
One of the most important things to note about this circuit is that LED and supply voltage is irrelevant - provided, of course, that the supply voltage (V+) is sufficient to overcome the voltage drop across the resistor, LED and the resistance of the FET.  In other words, the voltage drop across the FET could be zero volts or 5 volts and as long as the supply voltage high is enough to overcome the voltage drops across the circuit's components (7 volts would probably be adequate) it doesn't matter!  If you use much more voltage for your power supply (V+) than the LED and drop across Rsense require, you'll be excess power in Q1 as heat:  As long as Q1 is properly rated and heat-sinked, it will work fine, but such operation may be wasteful - particularly if you are using battery power!

To make this circuit practical for use as a linear modulator for speech, several refinements need to be made:
Comment:  While an N-Channel power MOSFET is shown in the circuit diagrams, a NPN transistor could also be used.  The MOSFET was used because its drive requirements are negligible at audio frequencies (e.g. only capacitance) and because they are rather ubiquitous (even available at Radio Shack) and they are inexpensive.  If an NPN transistor were used for a high-power modulator, a Darlington arrangement would likely be required as the current-sourcing capability of a typical op amp is likely inadequate for reliable operation.  It should also be noted that if an bipolar transistors are used that a few other changes in the stabilizing resistors/capacitors would also be required to maintain circuit stability.

Figure 2:
Schematic of the "simplified" high-power LED Linear Modulator.  The circuit depicted is for demonstration purposes and it is recommended that if you wish to build a modulator, construct that depicted in Figure 3 (below) instead.
Click on the image for a larger version.
Small version of LED Linear Modulator
Circuit description - "Simple" version:

The circuit shown in Figure 2 addresses the added requirements.  The "heart" of this circuit is the "precision current sink" consisting of U1B, Q1, and R9.  I don't recommend this circuit for serious use, but rather as one to illustrate the concepts involved for the reasons mentioned below:  I would recommend the circuit depicted in Figure 3, instead.

In this circuit, the current through the current-sense resistor, R9, produces a voltage that is proportional to the current through the LED.  U1B is wired such that, because of the closed-loop feedback, it will attempt to make the voltage on pin 6, the inverting input, the same as that on pin 5, the signal input:  If the voltage on pin 6 is below that of pin 5, the output voltage of the op amp will increase, causing Q1, a power FET, to conduct more, until the voltages on pins 5 and 6 match each other.  Because the luminous output of LEDs is proportional to their current, the result is an extremely linear modulation curve with excellent audio fidelity.

In order to properly modulate an LED one needs to establish a "resting" current:  A "100% modulated LED" will experience a current excursion from zero to twice the resting current, with the resting current being half of the maximum current and to do this, R7 sets a DC bias.  With the 5 volt reference supply shown this resting current could be set from zero all the way to 2.5 amps, but with a 3-watt Red Luxeon device, this would be set to 1.1 amps as measured by observing a drop of 1.1 volts across R9.

The audio is applied and modulated atop this resting current, coupled via C3, a 0.1uF capacitor.  The value of 0.1uF was chosen purposely as it will limit the low-frequency response of the modulation to about 150 Hz or so at the -6dB point - and this small value will also allow the circuit to stabilize more quickly when it is powered up and to minimize effects of wind or breath on the user's microphone.  The circuit consisting of U1A is a simple audio amplifier, the gain being adjustable using R4, with the audio coming from an inexpensive electret microphone.  If the gain of this circuit is not adequate for the microphone being used, it can be increased by decreasing the value of R3 (to as low as 1k) or by increasing the value of potentiometer R4 to as much as 1 Meg.

In order to maintain stability of the circuit with varying power supply voltages, U2, a 5 volt regulator is used because without a stable voltage reference the LED's resting current would vary:  A regulator is chosen over a simple Zener-based system as it removes from the reference supply any traces of modulation that might be imposed on the power supply, improving stability and preventing "motorboating."  Another critical component is C5, a 2200uF capacitor and this provides a low-impedance path, preventing current swings from excessively modulating the power supply:  Without it, the entire circuit could be unstable, and modulated audio might ingress any other devices operated from the same supply.  R11/C4 are also used to improve stability of the circuit at higher frequencies.  It should go without saying that Q1, the N-channel power FET, must be heat-sinked!

When building any circuit that mixes both low-level (microphone) signals with high-current ones (from the LED) it is important that single-point grounding techniques be employed.  What this means is that all ground connections having to do with the LED, its power transistor (through R9) and bypass capacitor (C5) should be connected to one point.  The other ground connections on the rest of the circuit should be connected to that "high-current" grounding point with just a single wire.  The idea here is to prevent any voltage drops across any wires that carry current from being communicated to the low-level portions of the circuit.  It's easy to forget that a few centimeters of small-gauge hookup wire could easily have a few hundredths of an ohm of resistance:  If a couple of amps is flowing through that wire, the crop across it could be 10's of millivolts - a level not too different from that coming out of the microphone!

Current limiting:

Another component worth mentioning is the current-limiting resistor in series with the LED (RLimit.)  If the LED module is not constructed with a means of Overcurrent Protection then the builder may feel more comfortable with the addition of this component.  For a 12.0 volt supply (representing a partially-depleted lead-acid battery) using a Luxeon III, one would wish to limit the absolute peak current to 3 amps or less.  Assuming a nominal peak current of 2.2 amps one would incur 2.2 volts across R9, about 1 volt across an inexpensive MOSFET, and 4 volts across the LED - plus another half volt or so due to IR drop in the connecting wiring for a total voltage drop of 7.7 volts or so - the "remainder" being about 4.3 volts.  In this particular case, using a 1.5 ohm resistor for RLimit with a dissipation of 5 watts would be sufficient, as the average current would be 1.1 amps.

Another way to limit the maximum current to the LED would be through the use of a larger ohmic value of resistor for R9 - but remember that doing so will not only require a higher-power resistor, but still-higher peak-to-peak audio levels from U1A, requiring one to take into account that sufficient audio levels would be available.

"How 'linear' is this modulator?"

To determine the relativel linearity of the modulator, the amount of distortion was measured using a Hamamatsu S1223-01 photodiode in parallel with a 10k resistor, the output of which was connected to a computer sound card.  In each case, a red Luxeon III high power LED was used as the light source, using the circuit in Figure 4 as the modulator.

Before any other measurements were to be done, the amount of distortion of the current waveform applied to the LED was measured to provide a reference point as to how "clean" the signal being applied to the LED might be.

First, the Spectran program was used to measure the distortion produced by the 1 kHz tone generator itself:  The 2nd and 3rd harmonics were found to be at least 59dB below the fundamental.

Next, the distortion of the modulator itself was measured by observing the voltage waveform across the current-sense resistor (
Rsense in Figure 1.)  With the modulator set at 100% and a small amount of audio compression occurring, the measured distortion was:
  • 2nd harmonic:  -50.7 dB (0.29%)
  • 3rd harmonic:  -54.3 dB (0.19%)
(Distortion of higher-level harmonics was noted to be somewhat lower, but not included in this measurement.)

The total distortion of the two harmonics was about 0.5% and did not change more than 1 dB between 1.1 amps and 100 milliamps.  It should be noted that the above test measures the linearity of the modulator only
:  Remember that the curve relating and LED's luminous output versus current is not perfectly linear, so we need to run a few more tests.

Next, the distortion was measured using a plain photodiode, illuminated by the LED and configured as noted above:  Care was taken to avoid saturating the diode.  To determine the effect of nonlinearity of current-versus-luminous output, two sets of measurements were taken as noted below:  

Measurement 1:
  1.1 amps of average current, peak current of 2.2 amps
  • 2nd harmonic:  -42.1 dB (0.79%)
  • 3rd harmonic:  -49.2 dB (0.35%)
  • (Other harmonics were noted to be at least 10 dB lower)
Measurement 2:  100 milliamps of average current, peak current of 200 milliamps.
  • 2nd harmonic:  -32.2 dB (2.5%)
  • 3rd harmonic:  -41.5 dB (0.84%)
  • (Other harmonics were noted to be at least 10dB lower)
  • Total = 3.4% (approx.)
As can be seen, the overall linearity is quite good, although I was surprised to note that the low-current linearity was somewhat worse than the high-current linearity.

Additional testing was done using the
KA7OEI Version 3 optical receiver with the lowpass filter bypassed and light attenuated to avoid saturating the receiver:

LED Current of 1.1 amps:
  • 2nd harmonic:  -44.2 dB (0.62%)
  • 3rd harmonic:  -51.3 dB (0.27%)
  • Total = 0.9% (approx.)
LED Current of 100 milliamps:
  • 2nd harmonic:  -32.3 dB (2.4%)
  • 3rd harmonic:  -43.6 dB (0.66%)
  • Total = 3.1% (approx.)
As you can see, the linearity, in all cases, is quite good, especially at higher LED currents.

The use of more than one Luxeon in series:

This circuit, as shown in figure 2 (assuming that RLimit is omitted and is zero ohms) is capable of driving two Luxeon III LEDs in series while operating from a 12.0 volt supply.  If one is driving a pair of LEDs in series, more careful attention must be paid on the value of RLimit, if one chooses to use this resistor in the first place.  If voltage drop is to be minimized, the selection of a FET for Q1 with a low ON resistance is also advised.

It is also worth pointing out that reducing the value of R9 further-reduces the amount of voltage drop across the modulator, allowing one to modulate a pair of Luxeon III LEDs in series with a power supply voltage below 10 volts.  If, for example, R9 was changed to 0.25 ohms then only 0.55 volts of peak-to-peak audio would be required at C3 to modulate one or more Luxeon III LED's (in series) to 2.2 amps.  If the value of R9 is decreased, it is advisable to increase the value of R5 (to about 33k or so in the case of R9 being 0.25 ohms) to put a "safe" upper limit on the amount of the LED's resting current as adjusted by R6.


It is recommended that one uses a current-limited supply (maximum of 2.5 amps or so) or replace the LED with, say, a 1 ohm resistor, when checking out the circuit and setting things up:  It would be easy, due to a wiring error or misadjustment, to destroy the LED.  The first step would be, with the gain set all of the way down (R4 at minimum resistance) to set the idling current at 1.1 amps with R6:  These recommended currents are those suggested if one is using a 3-watt Red Luxeon:  Different devices may require different current.

The next step would be to adjust the audio gain.  For this, one starts out with minimum gain and simply increases it until 100% modulation is achieved.  Even without any test equipment this is easily done as one can tell by ear when full-modulation is achieved by noting that audible clipping starts to occur.

It should be noted that if the resting current is adjusted using R6, the audio gain should be readjusted at the same time to assure 100% modulation.

Notes on operation and construction:
Additional comment:
Why I don't recommend this circuit:

Minor circuit refinements - improving current limiting and adjustment:

If you are going to build a simple modulator, I would recommend the circuit in Figure 3 over that in Figure 2 as it is much easier (and safer, from the LED's viewpoint) to use.

The LM324 is cheapest, most commonly-available op amp available that can operate with an input and output down to zero volts, and with it we have four available op amp sections.  By using just one more section than the circuit shown in Figure 2 and adding two resistors we can take advantage of the op amp's supply-limited voltage swing to put an absolute limit on the maximum amount of current that could be applied to the LED and provide a very easy means of adjust LED current while maintaining full modulation:  Figure 3 shows such an adaptation.

In this circuit, U1C is used as a microphone amplifier, capable of a gain of up to about 20dB, followed by U1B which is wired as another op amp section with a fixed gain of about 13dB - or a voltage gain of 5:  Note that the value of R7 could be increased (to several hundred k-Ohms) if it turns out that U1A/R4 do not yield enough gain from your microphone to adequately drive the LED to 100% modulation.  It is important to notice that U1C is AC coupled to U1B to prevent cumulative DC offsets from being amplified from U1C.  Wired in the way shown, U1B's output voltage is centered around 5 volts as referenced from U2, the 5 volt regulator, with modulation riding atop it with 100% modulation being 10 volts peak-peak.

When operating from a 12 volt supply, the output of U1B is limited in its swing from zero volts, representing zero current (or "0%" modulation) to just short of 11 volts - this upper voltage being limited by the supply voltage and the ability of the LM324's output to swing only within about a volt or so of the positive supply rail.  Because of this intrinsic, positive upper limit of the audio voltage swing, the maximum LED current is limited to a reasonable positive value - about 110% when operated from a 12.0 volt supply, or about 125% when operated from 13.8 volt source - no matter how much audio drive is applied.  Of course, if one operates U1 from a higher voltage supply (say 24 volts) this intrinsic "positive modulation" current limiting does not necessarily apply.
Figure 3:
Schematic of the "simplified" high power LED Linear Modulator with improved current limiting.
Click on the image for a larger version.
Small version of
                    LED Linear Modulator schematic

The actual LED currents are set using R8 and R9 to scale down the voltage output from U1B:  R8 is used to set the maximum current obtainable when R9 is set to full current and this arrangement has the advantage that if R9 is front-panel mounted, one may use it to continuously adjust the LED current from "full" current (1.1 amps of idle current, for example in the case of a red Luxeon III) all the way down to just a few milliamps - still have 100% modulation of the LED at any selected current.  If one does not require the need to adjust the LED's idle current from the front panel, the one can eliminate R9 completely, using R8 to set the idle current once and forgetting it.

Because of a quirk in the way the LM324 works, R5 and R8 are used to load the outputs of U1C and U1B, respectively, to prevent crossover distortion.  In the case of R8, the resistance value is not critical - but it should be less than 5k.  Practically any value of potentiometer may be used for R8 from 1k up to 100k, but if more than 5k is used, simply parallel a lower-value resistor (anything from 1k to 4.7k is fine) across it to maintain loading at pin 7.

As noted above, the beauty of this scheme is that by adjusting R9, the LED's current can be adjusted from "full" output all the way down to zero while maintaining a constant level of modulation throughout the range!

A few comments applicable to all circuit variations on this page:

In order to maintain stability all of the completed circuits have capacitors at the output.  Taking the schematic in Figure 3 as an example:
The importance of maintaining 100% modulation:

Up to this point "100% modulation" has been mentioned - but not explained.  For the purposes of this discussion, 100% modulation refers to the fact that the LED's current varies about a "resting" current from zero to twice the resting current.  As it turns out, this range represents the maximum that the amplitude-modulated LED can be driven and avoid distortion of the original signal.

If you were to attempt to exceed 100% modulation, the most obvious side-effect would be that some portion of your waveforms would try to go below zero current - something that is clearly impossible - and that portion of the audio would be hard-clipped, causing distortion.  On the "positive" side of the modulation, it may be possible that the current could go above twice the resting current - at least until the op-amp, transistor and/or power supply ran out of voltage/current "swing" at which point it would "hard clip."  As mentioned above, the circuits described, when operated from 12-14 volts - a range that encompasses portable battery operation - can produce around 125% positive modulation.

Practically speaking, however, one can overdrive the audio circuit and cause a significant amount of clipping (perhaps 10-20%) and actually improve intelligibility somewhat under noisy/weak-signal conditions.  While the fidelity of the speech is obviously reduced, the impact on its intelligibility is generally negligible and the fact that the "peak-to-average" ratio is reduced (that is, the "quiet" portions of speech are louder by comparison than the loudest portions) and the overall "speech power" is actually increased and can be better heard when conditions are poor.
"Why the LM324?"

You may have noticed that in the parts lists, it says not to substitute the LM324 for other op amps.  Why is that?

The main reason for this is that the LM324, unlike most op amps, has both an input
and output voltage range that includes ground and this allows a simple modulator circuit to be built without the need for a negative supply voltage!  If you try this with most other op amps, it may not work very well (e.g. modulation may not go all the way to zero) or, worse, some op amps do "odd" things (such as work "backwards") if you exceed their input voltage range.  If the op amp's output can't go all the way to zero, it may not be possible for the LED's current to be reduced completely to "zero" during 100% modulation - factors that can come into play in both U1B and U1C in Figure 3.

There are other (better!) op amps that can work to the negative supply rail, but the LM324 is arguably the cheapest and most common of the bunch.  As shown, the modulators depicted in figures 2, 3 and 4 pretty much run out of "steam" by 10 kHz because of the bandwidth limitations of the LM324, but if it's only speech audio that you want, this is sufficient.  Even so, it's not the current sink itself that (e.g. U1C in Figure 3) but rather the high-level driver (U1B in Figure 3) that first suffers at high frequencies - something that can be remedied by running this stage as a unity-gain follower and amplifying the signal to 10 volts peak-to-peak using a different op amp.  The bandwidth of the LM324-based current sink itself is closer to 20 kHz at 100% modulation.

If you want more bandwidth a higher-bandwidth op-amp must be used - but it, too, must be capable of operating down to the negative rail unless a negative voltage supply (2-3 volts will suffice) is provided.  One suitable op-amp is the National LMC660 - also a quad op-amp - but if this is simply dropped in place of the LM324 you can expect oscillation owing to the radically different amplitude/gain/phase response of that amplifier:  In other words, it's the insufficiencies of the LM324 that keep the circuit in Figures 3 and 4 stable under all operating conditions:  Different op amps would require different techniques.

In a prototype
(to be documented elsewhere) the LMC660-based circuit provided "flat" frequency response at 100% modulation to over 40 kHz and with minor tweaks, usable response to nearly 200 kHz making it a good candidate for use with VLF "subcarrier" schemes.

Eyebrows may be raised by having circuits that do not lend themselves well to substitutions, but real-world op-amps in closed-loop feedback circuits always require that a bit of care be taken to assure phase stability.  Since different types of op amps can have radically-different properties, it is not unexpected that in some cases simple substitution may result in instability!

Exactly how much "clipping" is "too much"?  That's mostly a matter of taste.  Clearly, if one runs the audio too "hot" then excessive clipping will result in enough audio distortion to reduce intelligibility.  Again, it should be noted that the circuits in Figure 3 and Figure 4, when operated from a 12-14 volt supply, can be run into clipping without much fear of damaging the LED as they intrinsically limit the maximum amount of LED current to a reasonably safe value.  The same cannot be said of the circuit in Figure 2, however!

What about lower than 100% modulation?  If one doesn't fully-modulation the LED, the result is that the amount of audio being conveyed by the lightbeam is reduced and it will sound "quieter."  Where this becomes an issue is where the signal is already weak or competing with noise:  A badly-undermodulated (or "quiet") signal is at a significant disadvantage and is clearly not being used to its full potential.

In other words, in terms of overall effectiveness it's better to run the audio a bit "hot" and put up with a little bit of distortion than run it too low!

Why I recommend the circuit in Figure 3 over those in Figures 1 or 2:

As noted above, when first setting up the circuit with the LED, R8 is set to a setting correlating with the maximum "idle" current for the LED and if this control is not disturbed (e.g. not accessible from the front panel) one cannot easily damage the LED by either overdriving it or through misadjustment, making the circuit nearly foolproof!

As is also mentioned, when properly configured, one can also adjust the LED's drive current from the pre-set maximum down to zero, but not have to adjust the audio level to keep the LED fully-modulated at any setting.  With this capability, you can do things such as:

Adding a few features:

The "simple" modulator shown in Figure 3 work nicely, but in order to improve intelligibility and facilitate alignment of the optical gear a few extra features would nice to have:

Circuit description - "Fancy" version:

Figure 4:
Schematic of the high-power LED Linear Modulator.
Click on the image for a larger version.
Small version of fancy LED Linear Modulator

Signal input stage:

The audio input can be one of three sources:  A built-in electret microphone for those situations where you forgot to bring one, an external microphone via J1, or an external line input via J2.  S1, an SPDT switch, selects which source is to be used, and if a "center off" switch is used, the center position can serve as a "mute" setting.  Note that J1 is a "disconnecting" type of jack and is wired to disable the internal microphone when an external one is plugged in.  In experimentation, it has been noted that computer-type microphones are wired in one of two ways:  While the audio is always on the "tip", some connectors apply bias to the tip and leave the ring disconnected while others apply the bias voltage only to the ring.  The circuit shown accommodates both wiring schemes.

Note also that J2 is wired such that the two resistors will sum (and attenuate) a line-level stereo input (from a computer or an audio player) to a monaural signal.  The resistors (R3 and R4) are necessary in many audio amplifiers because it has been observed that with many stereo-output audio devices, simply shorting the left and right channels together often results in distortion as the two amplifiers will "fight" each other.


C1 is used to limit the low frequency response to about 30 Hz and U1C, a non-inverting amplifier, contains a gain cell, OC1 that consists of a Cadmium-Sulfide (CdS) photocell and LED, optically coupled in a light-proof package.  The CdS cell (which is in parallel with R9, the "Max Gain" control) will decrease in resistance when the LED, driven by U1D, is illuminated at high audio levels, thus causing the resistance to go down and reduce the circuit gain.  In this way, an "AGC" or Automatic Gain Control circuit is formed, keeping "loud" signals from overdriving the modulator and bringing up quiet audio to a higher level, keeping the level fairly constant.  Under poor conditions, it is advantageous to keep the audio level as "loud" as possible - but not so loud that overdriving and distortion occurs.

U1B is used as a summing/gain amplifier (a voltage gain of 5 is set) to provide additional amplification.  The output of this amplifier is fed to U1D which has a variable gain setting that effectively sets the maximum amplitude of signal that can be present at the output of U1B.  The output of U1D goes to a full-wave bridge rectifier (D1-D4) that allows the LED to be illuminated on either positive or negative excursions of the audio waveform. C3/R12 prevent "clipping noise" from being put onto the +5 volt supply line as well as preventing excess positive excursions of the +5 volt rail by averaging out the conduction current through the LED.

There is another input to U1B:  The signals from an audio tone generator (see below) are input via R21 and C8.  Note that the audio levels from the tone generator are not affected by the AGC amplifier, but the AGC does detect the presence of those tones being generated and will decrease the gain of U1C accordingly to prevent overmodulation of the composite signal.

    Comment about the gain control device:
It is possible to use other devices for gain control:  The CdS cell and LED method mentioned above has the advantage of consisting of components that are probably already in your parts box as described below!  Other suitable devices include JFETs, MOSFETs, gain-control devices like the NE/SA570 (or the '571) and those could have been used as well.

Switch SW3 provides high-frequency boost to the transmitted audio - mostly to accommodate conditions in which the signal-to-noise ratio at the receiving end is quite low.  While the frequency response of the transmitter and accompanying receive system are fairly flat, the human voice has relatively little energy at higher frequencies (above 1 kHz or so) but this is the same range in which un-voiced consonants have their energy.  With weaker signals it was found that these un-voiced speech components - those that allow one to tell an "f" from an "s", for example - were among the first to be lost, making deciphering speech a bit more challenging.  This capability, along with the liberal use of phonetics, can help improve intelligibility under such adverse conditions.  If digital modes were used, pre-emphasis would probably be disabled.

Peak Indicator:

The Q2/Q3 circuit forms a "peak" indicator to let the user know that the audio level is sufficient to fully modulate the transmitter.  In this circuit, R27 sets a threshold voltage:  When audio from the output of U1d drops below this voltage, it causes Q2 - and, in turn, Q3 - to conduct, turning on the LED.  Even though this indicator works only with the downward modulation, it serves the purpose of letting the user know that something is happening.

The purpose of this circuit is simply to let the user know that an "adequate" amount of audio is present:  It is expected that it will flash frequently under normal conditions, with the AGC amplifier preventing gross overmodulation.  What is most important is that the user note if the LED is NOT flashing occasionally - a conditions that means that either there's no audio at all (which could happen for a number of reasons - such as forgetting to move S1 to the proper position, having a microphone disabled, etc.) or too little audio to properly modulate the system - which could occur if R9 is set too low or if the external audio source were running with an inadequate output level.

About OC1:

OC1 is simply an LED that is optically coupled to a Cadmium Sulfide photocell.  While these are commercially available (many are made by Vactec) they can also be homebrewed rather easily.  All that is required is an LED (a high-brightness red one will work fine) and a small CdS cell.  This CdS cell should exhibit a resistance of several megohms after a few moments in total darkness and under 10k in bright room lighting.

To construct a homebrew gain cell:
It is also possible to use a green LED if a high brightness red LED is unavailable.  In some ways a green LED is a better match for the CdS cell as it is more sensitive to the green light than red.  A word of warning, though:  If you use a high-brightness LED, verify that it is of the "low voltage" variety - that is, it illuminates at 2.1 volts or lower.  Many of the "super bright" green LEDs need 3-4 volts to light up, and this voltage is too high for the circuit to work properly.

Finally, some CdS Cell optical couplers contain a pair of LEDs connected back-to-back so that they respond to either positive or negative voltages.  If you are constructing your own optical coupler, you can do this and avoid the need for D1-D4 completely - just make sure that you try to position the two LEDs so that they illuminate the CdS cell more or less equally:  It is possible that the 2-leaded "bi-color" LEDs (e.g. one that lights with red with for one polarity and green with the other) will work nicely as well.

Comment:  With this linear modulator it is more important that the downward modulation not exceed 100% too often - that is, the LED current cannot go below zero - as this will cause distortion.  On the other hand, occasional excursions above 100% (that is, above twice the average, resting current) may be permitted.  If the full-wave rectifier (D1-D4) is omitted and a single LED is used, the LED could be connected so that it responded only to downward ("negative") peaks of the audio waveform.  Note that both U1B and U1D invert the audio, so the LED would, in fact, be wired to illuminate as the output voltage of U1D went downwards.

Output driver and monitoring:

U1A is wired as a "Precision current sink" and with R20, the 1 ohm "sense" resistor, one volt of input to pin 3 will result of one amp of current flow.  Using U1A in this way guarantees that LED current will be proportional to the voltage present across R18.

Being that the output from U1B is AGC-limited to an amplitude pre-set during the adjustment of R16, this signal is always representative of one with 100% modulation.  If one wanted a resting current of 1 amp, this output, through R17, a trimmer potentiometer,  is set to provide 2 volts peak (1 volt "resting") voltage at its wiper.  In this way, R18 may be used to provide a 100% modulated signal over a continuous range from full current to just a few milliamps.

Two monitoring points are provided:  The junction of C11/R22 may be used to measure the average LED current, where 1 volt = 1 amp, while J3 is used to monitor the modulated audio with headphones, the level being adjusted by R23.  S2 is in series with the LED, allowing it to be shut off without having to power down the entire circuit and PB1 allows simple on/off keying of the LED:  If the LED is being modulated with an audio tone, MCW keying can be done using PB1.

Modifications to minimize voltage drop:

The following comments about minimizing voltage drop are generally applicable to both the "simple" and "fancy" circuits although specific component designations refer to those in Figure 4.

In testing, it has been noted that the LM324 used will properly operate down below even 10's of millivolts input and because of this, it is possible to reduce the value of R20 down to at least 0.1 ohms.  If a low on-resistance MOSFET is used for Q1, one can construct a circuit that will fully-modulate the LED with less than 0.5 volts of additional voltage drop.  What this means is that with these lower resistances it is possible to run a single Luxeon III from a 6 volt supply (with appropriate circuit modifications) or up to three Luxeon III's in series from a 12 volt battery supply!

With the circuit shown and a value of 1 ohm selected for R20, there is adequate headroom to modulate two red Luxeon III LEDs wired in series to 2.2 amps peak while using a single 12 volt "gel cell" as a power source.

With a 1 ohm resistor used as R20, it is possible to set the LED's resting current to 5 milliamps and still achieve nearly 100% modulation!

Comments about the circuit:
Adjustment of the modulator:


S1 is used to select between the line input or microphone input while R9 is used to adjust the maximum gain that the AGC will allow.  If a high gain setting is used on a fairly high audio level, overmodulation will not occur, but there will be considerable compression - depending on the audio level at the input:  High compression may be desired under conditions of low signal-noise in order to maximize intelligibility.

Note that in all positions of S2, the audio input is still active.  This means that even when generating a tone the microphone will pick up and transmit audio, although the level will be reduced by the AGC amplifier because of its detection of the tone.  If S1 is a "center-off" type of switch, the middle position can be used to mute the audio inputs or, if no line input audio source is present, muting can be accomplished by selecting that position.

The LED current may be set with R18, providing 100% modulation at any current from the full setting all the way down to a few milliamps.

Comment:  Again, 100% modulation is defined as modulation that goes all the way from zero up to twice the average (unmodulated) current as set by R18.

Tone generator:
Figure 5:
Schematic of the PIC-based tone generator that is integrated with the linear modulator.
Click on the image for a larger version.
PIC-based audio tone generator

On the same board, I also built a tone generator using an 8-pin PIC, the 12F683, which has an onboard A/D converter as well as a hardware-based PWM generator that can be used as a D/A converter.  Much of the source code was "borrowed" from my PWM-based modulator, hence the similarities in operation.

Clocked at 20 MHz, there is a 19.53125 kHz interrupt that is used to generate sine waves using DDS techniques and a 10-bit sine lookup table, using the PIC's built-in PWM hardware as a D/A converter to produce 78.125 kHz square waves of varying duty cycle.  The lowpass filter for the PWM-generated audio consists of R103, R104, C105 and C106, effectively removing most of the 78 kHz PWM energy while causing little attenuation of the highest-frequency audio tones.

Modes are selected by applying different voltages to GP1:  These voltage are created using a resistive divider on the rotary switch and the digitized voltage is used to select the appropriate mode.  Also present is R102, a pot that produces a variable voltage read by the PIC's onboard A/D converter and is used to select the various tone modes and frequencies.

On the output of the PWM network are two pots, R105 and R106, that are used to set the output tone levels to correspond with 100% modulation in the case of the test tones, and 25% modulation in the case of the pilot tone:  The second half of the rotary switch (S1B) is used to select the tone level in the case of the test tones, the attenuated tone in the case of the pilot tone, or no tone (in position "E") should the pilot tone not be used at all.  Note that C108 isn't needed if you use this circuit with Figure 4 as there's already a blocking capacitor, C8, in that drawing.

Pin 4 (GP3) is used to select whether or not the 1 kHz tone should really be 1 kHz.  This arose when this circuit was to be used in areas with 50 Hz mains (e.g. most areas of the world other than the Americas) where a mains harmonic would coincide with the 1 kHz tone.  By tying pin 4 high, the 1 kHz tone selected in "Tone Mode A" and the one available in "Tone Mode C" (see below) is, in fact, 1000 Hz.  If pin 4 is grounded, a 1020 Hz tone is generated at those positions, instead.  Note that even though the 4 kHz "pilot" tone is also harmonically related to 50 Hz mains, its frequency is not altered as it serves simply as an amplitude reference.

The nominal voltage used to select the tone mode is noted below and assume that the PIC is operating from a 5.0 volt supply.  The actual voltage thresholds for selecting the modes are midway between the voltages specified.

Tone modes:
A - (0 volts)  1 kHz tone, fixed.  The 1 kHz tone is a "standard" tone used for peaking the receiver using the audible S-meter system, for MCW keying, for measurement of scintillation, or just as a source of audio.  Even though it is possible to set Mode C to generate a 1 kHz tone, this mode allows the convenience of simply turning the switch all the way to one end of rotation - something that's easy to do in the dark!  In this mode, R102, the "Tone" pot has absolutely no effect.  Note that if pin 4 is grounded, a 1020 Hz tone will be produced instead.
B - (1.0 volt)  The tone pitch is variable from 20Hz to about 2.5 kHz using R102.  Note that inside the PIC this adjustment is de-linearized so that when a linear pot is used, the rotation is "stretched" at the low end and compressed at the high end, making frequency adjustment seem more "natural."
C - (2.0 volts)  Eight fixed tones are selectable - see below.
D - (3.0 volts)  Tone sequences.  Below the midpoint of rotation, a descending tone sequence is generated, while above the midpoint an ascending tone sequence is generated.  The rate of repetition is faster the farther the pot is turned toward either end of rotation.
E - (4.0 volts)  No tones generated:  Used for "normal" audio mode.
F - (5.0 volts)  A 4 kHz tone is generated as a "pilot" tone.  Using R106, the level of this tone is typically reduced by 12dB to 25% of the maximum level (referenced to "100% modulation") and this tone is mixed with normal microphone/line audio.  As shown in Figure 4, the modulator's AGC is connected such that the amplitude of this pilot carrier is taken into account and prevents overmodulation of the LED with the combined audio sources.  At the receive end, this pilot tone can be filtered out and is available for analysis and/or compensation of scintillation.
The 8 fixed audio tones available in Mode C are:
1 - Musical note B0 (actual freq. = 30.9944 Hz)  (Lowest voltage on pin 7)
- Musical note E1 (actual freq. = 41.1295 Hz Hz)
- Musical note C4 - middle C (actual freq. = 261.6674 Hz)
- Musical note F4-sharp (actual freq. = 369.8468 Hz)
5 - Musical note A5-sharp (actual freq. = 932.26912 Hz)
- Musical note - E6 (actual freq. = 1318.52896 Hz)
- 440 Hz - Musical note A4 (actual freq. = 439.907 Hz)
8 - 1kHz tone (actual freq. = 999.9242 Hz) or, if pin 4 is grounded this will be 1020 Hz (actual frequency = 1020.13 Hz) (Highest voltage on pin 7)

Figure 6:  Pictures of the as-constructed linear modulator.
Top:  Front panel of modulator.  Middle:  Circuit board of modulator.  Bottom:  Wiring of front panel and bottom of circuit board.
Click on any image for a larger version.
                    version of image showing front panel of linear
                    version image showing circuit board of linear
                    version of image showing wiring and backside of
                    front panel of linear modulator

Operation in the various modes

Using DDS techniques, low-distortion sine waves can be generated at practically any audio frequency below the Nyquist limit with a resolution of 0.298 Hz..  Having this capability allows several tone generation modes:
Note about Mode D - the tone sequence generator:

The tone sequence mode (Mode D) can generate either an ascending tone sequence consisting of tone #'s3, 4, 5 and 6 (in that order) or a descending tone sequence using the same tones in reverse order.  The tone mode (and sequencing rate) is adjusted via R102:

Adjustment of the tone generator:
Important note:  It is strongly recommended that you never operate any modulator or LED without having current limiting on the LED.  This may take the form of a resistor, or a current limit circuit such as one using an LM317.


S2 positions A, B, C and D are intended to be used to generate audio tones at 100% modulation.  Position F generates a low-level 4 kHz pilot tone for reference/analysis purposes and position E is for generating audio with no pilot tone at all.

When in modes A, B, C and D, R102 is used for setting the tone frequency, selecting one of eight fixed tones or selecting the tone sequence mode and rate, respectively.

Note that in all positions of S2, the audio input is still active.  This means that even when generating a tone, the microphone will pick up and transmit audio, although the level will be reduced by the AGC amplifier.  Again, if S1 is a "center-off" type of switch, the middle position can be used to mute the audio inputs or, if no line input audio source is present, muting can be accomplished by selecting that position.

Note that the (partial) component list below applies only to Figures 4 and 5.



Return to the KA7OEI Optical communications Index page.

If you have questions or comments concerning the contents of this page, or are interested in this circuit, feel free to contact me using the information at this URL.
Keywords:  Lightbeam communications, light beam, lightbeam, laser beam, modulated light, optical communications, through-the-air optical communications, FSO communications, Free-Space Optical communications, LED communications, laser communications, LED, laser, light-emitting diode, lens, fresnel, fresnel lens, photodiode, photomultiplier, PMT, phototransistor, laser tube, laser diode, high power LED, luxeon, cree, phlatlight, lumileds, modulator, detector
This page and contents copyright 2007-2013.  Last update:  20130719
Page count since 201005: