Johnyz
0
- Joined
- Jul 19, 2010
- Messages
- 421
- Points
- 0
The Practical Laser Driver Theory
Greetings everybody, it seems I'm back after a quite long hiatus. Anyway, during my pause, I realised how much info I have gathered when I was working on my driver designs back in the day and I'd like to write it down somewhere, just because. Greetings to drlava, rog8811, benmw, rhd, foulmist, Wolfman29 and everybody who ever got into designing a new driver circuit or radically modifying an existing one. Anyway, let's get to the business.
It turns out, laser diodes, like pretty much every modern technology, run on electrictity. But no! Not just -any- electricity, these little fragile guys require to have the energy served to them delicately. Let's begin by talking on how to do that. Laser diodes are similar to LED's, and if you ever looked into powering those, you can skip this part. The way these guys differ from most other parts that they have very little internal resistance – that means even though they were designed for a set current (big or small), if you just hook them up to a battery, they would burn out almost instantly (there are exceptions – see 'Direct Drive') because they would pull all the current they could, even though they can't take it. That means we need to limit it. But how? There are many ways to do that...
Ohm's law, the basic theory
Ohm's law defines the relationship between electrical current (marked as I), power (marked as P), resistance, (marked as R) and voltage (marked as E or U). The very basic idea of it is this: I = E / R (current equals voltage divided by resistance) and keeping in mind that power is calculated like so: P = E * I (power equals voltage multiplied by current), we can, as long as we have two of those values, calculate the remaining two. Now, to make stuff a lot easier for you, just check out this on-line calculator with all the derived equations: Ohm's Law Calculators
Now, when talking about laser and light emitting diodes (and semiconductors in general), there is one thing we need to mention – they have a property called the dropout voltage. What this means? It simply means that they drop the voltage when it's coming through them by the specified amount. For red laser diodes, it's usually around 3V, for violet ones it's around 4.5V, and for the 445 blues it's between 4V and 5V (but it varies). The new direct green diodes take up to whopping 8V. Which means, unless you have a boost driver, you need at least that voltage on your battery to make it work. Let's get to the interesting stuff then, ey?
I will start from the easiest/simplest stuff (or more precisely a lack thereof) and gradually get into the more technically complex stuff...
Presenting, at number one, the infamous...
Direct Drive
Just to start, this is NOT RECCOMENDED. I am talking about this as a theoretical/historical reference only! Any laserist who has been in this hobby for a couple of years cringes when he hears the combination "KipKay Maglite Red Laser" or some paraphrase thereof. Watch this Youtube vid - Amazing Lasers! - Laser Flashlight Hack! - YouTube – once again, do not replicate. Let's talk about this... What is he doing is that he is replacing the bulb in the flashlight by a laser module with a diode he extracted from an old DVD burner. He drills out the reflector and solders two longer pins on the diode leads, to make it fit instead of the removed bulb. What is he essentially doing is connecting the diode straight to the battery. Well! Why won't the diode die then? Let's take a look at it. He hooked up a red LD, which has a dropout voltage of around 3V, to a 2x AA flashlight, which supplies 3V. And what do you know? In this scenairo, the little resistance of the diode, plus the resistance of the contacts in the flashlight, add up and supply the diode with a reasonable current. What a coincidence..! Now, let me tell you why is this a very bad idea. First things first, you have no control over the current that will be supplied to the diode, you just have to hope it's just about right. If your LD has a dropout voltage of 2.8V, it'll probably die. If it's 3.2V, it won't probably light up. The second thing is, as the batteries get discharged, their voltage starts dropping... With it being quite close to the lowest possible voltage the diode can run on, the laser starts to dim very fast. When each battery loses only 0.1V, the total supply voltage drops to 2.8V, which may be no longer enough to power the LD and the laser stops working and you throw out batteries that would still be good for some TV remotes or other small electronics. Now let's say, dissapointed by the short battery life, you pop in some high power primary AA's like the Energizer Ultimate Lithium. These guys can supply up to 1.8V when new – which adds to 3.6V. And the fragile balance of the small resistance, small (post-dropout) voltage gets disturbed, and you can wave your diode goodbye... Also, if you put in the batteries backwards, the same thing happens. This type of "driver" experienced a bit of renaisssance with the 445 diodes being powered by single Li-Ion cells with the voltage between 3.6V and 4.2V, which is enough to light up most of these diodes, and since they can take a lot of abuse, often the current was limited enough by the internal resistance in the battery to make it work but... The next thing I'll be mentioning is very rarely used (and still risky) for laser diodes, but very common for LED's.
Number two, the still very simple...
Single Resistor Drive
Not exactly sure if it's the correct name, I'll be brief about this one. As I said, very rarely used for LD's, but common for LED's, this method uses a single resistor to limit the current. Let's imagine you have a 3x AAA host which provides 4.5V, and a red LD which has a dropout voltage of 3V. This leaves us with 1.5V left over, and let's say we want to give it 250mA of current. The ohm's law tells us that resistance equals voltage divided by current: 1.5V / 0.25A = 6 ohm. That means we need a 6 ohm resistor – if we want to be extra perfect, we can calculate the power as well, so we know how what power rating we need to get: power, as I already said equals voltage times current: that gives us we need atleast a 0,375W resistor, for good measure, it's advisable to get atleast a 0,5W one. But keep in mind, this has the same issues as the above method: since it's set up only for a fully charged battery it'll dim very quickly, and is likely to burn out if you somehow give it more voltage and has no reverse polarity protection. Once again, not reccomended... The easiest way to account for the dropping battery voltage is with a simple IC – which lets us move into the next thing.
Which is number three, the (probably) most polular driver group...
Linear Drivers
The most widely used thing among hobbyists – reliable, easy to build, and used probably by every single laserist at some moment during their time with the hobby. It's commonly called the DDL circuit, after the LPF member Daedal... Take a look here: Laser driver - It can be done We use a LM317 linear voltage regulator, which can be hooked up as a current regulator as well. If we pop in a protective capacitor to account for the voltage spikes, and a reverse diode to protect the diode from getting blown up if you put in the battery backwards (it'll conduct the current instead of the diode). Also popping a trimmer in parallel with the resistor, which is now only used to limit the maximum current now, we can set the current by turning the little screw on it! How cool! The downside is that the regulator IC needs another extra 3V for it to function, but the advantage is that now we can pop in a bit of extra voltage and it adjusts itself, when the batteries discharge it adjusts as well and the laser only dims when the batteries are fully discharged and actually can't supply the current. Another, more or less practical disadvantage is that, especially when running off a voltage significantly higher than the LD's dropout + the 3V required to make the regulator work it tends to warm up quite a bit (think putting 10440's in a 3x AAA holder). Never the less, I think every laserist needs to build one at some point. Just to add, the resistance is calculated with the IC's internal reference of 1.25V.
Low DropOut Linear Drivers (LDO's)
The thing is, especially when the high power 445's appeared, it turned out, that at currents exceeding 1A, even the 3V required by the LM317 were quite a limiting factor, especially at 1A (or more) it produces 3W (or more) of heat, which is quite a lot in a handheld. Fourtunately, the design of the LM317 can be improved to only require around 1.25V – the same as the internal reference. This gives us an obvious advantage, requiring less extra voltage to work. Also it makes it heat up less at high currents... One example of these is this: http://laserpointerforums.com/f64/fs-ghostdrive2-linear-drivers-6-99-a-71279.html, though these guys keep popping up from various members of LPF and dissapearing again. The IC frequently used for this is LM1117, but there are more, like the range from LM1083 to LM1086 – some of these being capable of handling up to 7.5A, but that's unrelistic for LD's unless you're talking multidiode setups. Further improvements in this design are possible but not without significant changes in the design. A noticable example of a non-LM317-related linear driver is the Blitz-Linear by Wolfman29: http://laserpointerforums.com/f67/n...op-out-1a-adjustable-linear-driver-79920.html As you can see, it's a very different IC with a very different design. However, these guys still have one disadvantage. No matter what, they have to burn off the extra (also reference) voltage... This can't be avoided with linear drivers, so let's move on to some switched mode ones, then! Starting with...
Number four, the underrated...
Step Down "Buck" Switch-mode Drivers
Now. Hold right there. First, you should take a look at this: http://fayazkadir.com/blog/?page_id=460 to explain how Switch-mode (also known as DC-DC) converters work in general. Is your mind blown yet? Calm down, I am not forcing you to build one
But as always, you're welcome to try. So what exactly do they do and what are their (dis)advantages? The biggest advantage is that they can indeed convert the extra voltage into current and not waste it. I'll explain. For example, when driving a red LD (3V dropout voltage, remember?) at 500mA from a fully charged Li-Ion (4.2V) we have an extra of 1.2V... We need 1.5W of power for this and assuming 90% efficiency, which is not uncommon for these (high efficiency is one of the advantages), it's abou 1.66W. Now, let's take the mighty Ohm's law into practice, we have 4.2V and 1.66W... I = P / E (current equals power divided by voltage) and that gives us a little bit under 400mA. What just happened? The step down circuit converted the extra voltage into current – in result, we are supplying 500mA to the LD but only drawing 400mA from the battery, with very little heating up, since only 10% of the power (0.166W) is now being wasted. Sweet, isn't it? But what about when the diode needs more voltage than your single-cell host can provide? Time to talk about the popular stuff, then!
Presenting number five, the popular...
Step Up "Boost" Switch-mode Drivers
Simply put, for some reason most laserists want to use blue and violet (and now green) diodes, which have a big dropout voltage, in hosts that only have one Li-Ion cell. (Actually, almost all actual electrical engineers I've discussed this with find this quite stupid, no offense though.) We can do just the opposite of what a "Buck" converter does – we can draw more current from the battery than the laser diode needs and convert the extra into voltage. The downside is that when you're driving a diode that already pulls a lot of current and you also need the extra to generate more voltage... Well, you're off to pull ludicrous (for handheld means) amounts of current. But, with modern Li-Ion technology and beefy cells - 18650 and larger, this is very doable. Just make sure to buy verified quality ones, because cheap cells from Chinese suppliers are notorious for not being capable to reliably handle this. (However, if you need unusual shaped ones, you often have no other choice – in that case, look for reviews and feedback from other people.)
And lastly, number six, the most convinient...
Step Up/Step Down "Buck-Boost" Switch-mode Drivers
Now, to be completely honest, even I don't exactly understand how these work, internally. For a regular user though, these are precisely the combination of the last two categories – they can both convert extra voltage into current, or extra current into voltage, depending on what they need at the moment. This is utilised either when the battery voltage tends to swing a lot, or when the battery voltage is very close to the required diode dropout voltage, or simply for convinience. The first situation is simple – when you have the new green diode that takes 8V and two Li-Ions to power it then they go form 8.4V when fully charged to 7.2V when discharged. As the battery voltage approaches the required diode voltage (plus some losses), the driver switches itself to boost to still satisfy the diode and keep the laser running. (NOTE: This is purely theoretical! I have not yet seen a finished buck-boost driver that can work as high as 8V. Don't even think about hooking a FlexDrive and two Li-Ion cells to your precious diode!) An other, more realistic example would be what (I believe) the FlexDrive was originally intended to be used in – take a 637nm red diode with a dropout voltage of around 3.3V and three NiMH rechargeable AA batteries – these can have a total voltage of 3.6V or more when charged and discharge to 3V or less. The idea is the same – when the supply voltage drops close towards the requirement, we switch to boost and enable ourselves to more time spent firing our lasers and less time spent charging. OR, since these situations are becoming progressively rarer and rarer, and we are becoming lazier and lazier, we can use these guys as universal laser drivers – got a 405nm project? Pop a buck-boost driver in and it'll spend the rest of its life working in the boost mode. It's a bit of a wasted potential, but hey, if you're building a red laser next, you can pop in the very same driver! Again, making it spend all the time as buck, but it's very convinient at the same time.
And that concludes this post. If there is any other brand new and exotic way of driving laser diodes invented, I'll try to add it somewhere around here. If you're still reading, thank you for your time, you are awesome and I hope that you find the information contained useful. If there are any errors in it, both factual and gramatical, or you just have something to add, let me know! I'll be more than happy to improve my work and I will ofcourse credit you. Now just for a little disclaimer...
DISCLAIMER: The author of this post is NOT RESPONSIBLE for any direct or indirect damage caused by the usage of information contained in the above text, including but not limited to, on health or equipment!
NOTE: Short citations of this text are allowed as long as you credit the author and include a link to this page in your work. Please contact the author for further permissions.
Greetings everybody, it seems I'm back after a quite long hiatus. Anyway, during my pause, I realised how much info I have gathered when I was working on my driver designs back in the day and I'd like to write it down somewhere, just because. Greetings to drlava, rog8811, benmw, rhd, foulmist, Wolfman29 and everybody who ever got into designing a new driver circuit or radically modifying an existing one. Anyway, let's get to the business.
It turns out, laser diodes, like pretty much every modern technology, run on electrictity. But no! Not just -any- electricity, these little fragile guys require to have the energy served to them delicately. Let's begin by talking on how to do that. Laser diodes are similar to LED's, and if you ever looked into powering those, you can skip this part. The way these guys differ from most other parts that they have very little internal resistance – that means even though they were designed for a set current (big or small), if you just hook them up to a battery, they would burn out almost instantly (there are exceptions – see 'Direct Drive') because they would pull all the current they could, even though they can't take it. That means we need to limit it. But how? There are many ways to do that...
Ohm's law, the basic theory
Ohm's law defines the relationship between electrical current (marked as I), power (marked as P), resistance, (marked as R) and voltage (marked as E or U). The very basic idea of it is this: I = E / R (current equals voltage divided by resistance) and keeping in mind that power is calculated like so: P = E * I (power equals voltage multiplied by current), we can, as long as we have two of those values, calculate the remaining two. Now, to make stuff a lot easier for you, just check out this on-line calculator with all the derived equations: Ohm's Law Calculators
Now, when talking about laser and light emitting diodes (and semiconductors in general), there is one thing we need to mention – they have a property called the dropout voltage. What this means? It simply means that they drop the voltage when it's coming through them by the specified amount. For red laser diodes, it's usually around 3V, for violet ones it's around 4.5V, and for the 445 blues it's between 4V and 5V (but it varies). The new direct green diodes take up to whopping 8V. Which means, unless you have a boost driver, you need at least that voltage on your battery to make it work. Let's get to the interesting stuff then, ey?
I will start from the easiest/simplest stuff (or more precisely a lack thereof) and gradually get into the more technically complex stuff...
Presenting, at number one, the infamous...
Direct Drive
Just to start, this is NOT RECCOMENDED. I am talking about this as a theoretical/historical reference only! Any laserist who has been in this hobby for a couple of years cringes when he hears the combination "KipKay Maglite Red Laser" or some paraphrase thereof. Watch this Youtube vid - Amazing Lasers! - Laser Flashlight Hack! - YouTube – once again, do not replicate. Let's talk about this... What is he doing is that he is replacing the bulb in the flashlight by a laser module with a diode he extracted from an old DVD burner. He drills out the reflector and solders two longer pins on the diode leads, to make it fit instead of the removed bulb. What is he essentially doing is connecting the diode straight to the battery. Well! Why won't the diode die then? Let's take a look at it. He hooked up a red LD, which has a dropout voltage of around 3V, to a 2x AA flashlight, which supplies 3V. And what do you know? In this scenairo, the little resistance of the diode, plus the resistance of the contacts in the flashlight, add up and supply the diode with a reasonable current. What a coincidence..! Now, let me tell you why is this a very bad idea. First things first, you have no control over the current that will be supplied to the diode, you just have to hope it's just about right. If your LD has a dropout voltage of 2.8V, it'll probably die. If it's 3.2V, it won't probably light up. The second thing is, as the batteries get discharged, their voltage starts dropping... With it being quite close to the lowest possible voltage the diode can run on, the laser starts to dim very fast. When each battery loses only 0.1V, the total supply voltage drops to 2.8V, which may be no longer enough to power the LD and the laser stops working and you throw out batteries that would still be good for some TV remotes or other small electronics. Now let's say, dissapointed by the short battery life, you pop in some high power primary AA's like the Energizer Ultimate Lithium. These guys can supply up to 1.8V when new – which adds to 3.6V. And the fragile balance of the small resistance, small (post-dropout) voltage gets disturbed, and you can wave your diode goodbye... Also, if you put in the batteries backwards, the same thing happens. This type of "driver" experienced a bit of renaisssance with the 445 diodes being powered by single Li-Ion cells with the voltage between 3.6V and 4.2V, which is enough to light up most of these diodes, and since they can take a lot of abuse, often the current was limited enough by the internal resistance in the battery to make it work but... The next thing I'll be mentioning is very rarely used (and still risky) for laser diodes, but very common for LED's.
Number two, the still very simple...
Single Resistor Drive
Not exactly sure if it's the correct name, I'll be brief about this one. As I said, very rarely used for LD's, but common for LED's, this method uses a single resistor to limit the current. Let's imagine you have a 3x AAA host which provides 4.5V, and a red LD which has a dropout voltage of 3V. This leaves us with 1.5V left over, and let's say we want to give it 250mA of current. The ohm's law tells us that resistance equals voltage divided by current: 1.5V / 0.25A = 6 ohm. That means we need a 6 ohm resistor – if we want to be extra perfect, we can calculate the power as well, so we know how what power rating we need to get: power, as I already said equals voltage times current: that gives us we need atleast a 0,375W resistor, for good measure, it's advisable to get atleast a 0,5W one. But keep in mind, this has the same issues as the above method: since it's set up only for a fully charged battery it'll dim very quickly, and is likely to burn out if you somehow give it more voltage and has no reverse polarity protection. Once again, not reccomended... The easiest way to account for the dropping battery voltage is with a simple IC – which lets us move into the next thing.
Which is number three, the (probably) most polular driver group...
Linear Drivers
The most widely used thing among hobbyists – reliable, easy to build, and used probably by every single laserist at some moment during their time with the hobby. It's commonly called the DDL circuit, after the LPF member Daedal... Take a look here: Laser driver - It can be done We use a LM317 linear voltage regulator, which can be hooked up as a current regulator as well. If we pop in a protective capacitor to account for the voltage spikes, and a reverse diode to protect the diode from getting blown up if you put in the battery backwards (it'll conduct the current instead of the diode). Also popping a trimmer in parallel with the resistor, which is now only used to limit the maximum current now, we can set the current by turning the little screw on it! How cool! The downside is that the regulator IC needs another extra 3V for it to function, but the advantage is that now we can pop in a bit of extra voltage and it adjusts itself, when the batteries discharge it adjusts as well and the laser only dims when the batteries are fully discharged and actually can't supply the current. Another, more or less practical disadvantage is that, especially when running off a voltage significantly higher than the LD's dropout + the 3V required to make the regulator work it tends to warm up quite a bit (think putting 10440's in a 3x AAA holder). Never the less, I think every laserist needs to build one at some point. Just to add, the resistance is calculated with the IC's internal reference of 1.25V.
Low DropOut Linear Drivers (LDO's)
The thing is, especially when the high power 445's appeared, it turned out, that at currents exceeding 1A, even the 3V required by the LM317 were quite a limiting factor, especially at 1A (or more) it produces 3W (or more) of heat, which is quite a lot in a handheld. Fourtunately, the design of the LM317 can be improved to only require around 1.25V – the same as the internal reference. This gives us an obvious advantage, requiring less extra voltage to work. Also it makes it heat up less at high currents... One example of these is this: http://laserpointerforums.com/f64/fs-ghostdrive2-linear-drivers-6-99-a-71279.html, though these guys keep popping up from various members of LPF and dissapearing again. The IC frequently used for this is LM1117, but there are more, like the range from LM1083 to LM1086 – some of these being capable of handling up to 7.5A, but that's unrelistic for LD's unless you're talking multidiode setups. Further improvements in this design are possible but not without significant changes in the design. A noticable example of a non-LM317-related linear driver is the Blitz-Linear by Wolfman29: http://laserpointerforums.com/f67/n...op-out-1a-adjustable-linear-driver-79920.html As you can see, it's a very different IC with a very different design. However, these guys still have one disadvantage. No matter what, they have to burn off the extra (also reference) voltage... This can't be avoided with linear drivers, so let's move on to some switched mode ones, then! Starting with...
Number four, the underrated...
Step Down "Buck" Switch-mode Drivers
Now. Hold right there. First, you should take a look at this: http://fayazkadir.com/blog/?page_id=460 to explain how Switch-mode (also known as DC-DC) converters work in general. Is your mind blown yet? Calm down, I am not forcing you to build one
But as always, you're welcome to try. So what exactly do they do and what are their (dis)advantages? The biggest advantage is that they can indeed convert the extra voltage into current and not waste it. I'll explain. For example, when driving a red LD (3V dropout voltage, remember?) at 500mA from a fully charged Li-Ion (4.2V) we have an extra of 1.2V... We need 1.5W of power for this and assuming 90% efficiency, which is not uncommon for these (high efficiency is one of the advantages), it's abou 1.66W. Now, let's take the mighty Ohm's law into practice, we have 4.2V and 1.66W... I = P / E (current equals power divided by voltage) and that gives us a little bit under 400mA. What just happened? The step down circuit converted the extra voltage into current – in result, we are supplying 500mA to the LD but only drawing 400mA from the battery, with very little heating up, since only 10% of the power (0.166W) is now being wasted. Sweet, isn't it? But what about when the diode needs more voltage than your single-cell host can provide? Time to talk about the popular stuff, then!Presenting number five, the popular...
Step Up "Boost" Switch-mode Drivers
Simply put, for some reason most laserists want to use blue and violet (and now green) diodes, which have a big dropout voltage, in hosts that only have one Li-Ion cell. (Actually, almost all actual electrical engineers I've discussed this with find this quite stupid, no offense though.) We can do just the opposite of what a "Buck" converter does – we can draw more current from the battery than the laser diode needs and convert the extra into voltage. The downside is that when you're driving a diode that already pulls a lot of current and you also need the extra to generate more voltage... Well, you're off to pull ludicrous (for handheld means) amounts of current. But, with modern Li-Ion technology and beefy cells - 18650 and larger, this is very doable. Just make sure to buy verified quality ones, because cheap cells from Chinese suppliers are notorious for not being capable to reliably handle this. (However, if you need unusual shaped ones, you often have no other choice – in that case, look for reviews and feedback from other people.)
And lastly, number six, the most convinient...
Step Up/Step Down "Buck-Boost" Switch-mode Drivers
Now, to be completely honest, even I don't exactly understand how these work, internally. For a regular user though, these are precisely the combination of the last two categories – they can both convert extra voltage into current, or extra current into voltage, depending on what they need at the moment. This is utilised either when the battery voltage tends to swing a lot, or when the battery voltage is very close to the required diode dropout voltage, or simply for convinience. The first situation is simple – when you have the new green diode that takes 8V and two Li-Ions to power it then they go form 8.4V when fully charged to 7.2V when discharged. As the battery voltage approaches the required diode voltage (plus some losses), the driver switches itself to boost to still satisfy the diode and keep the laser running. (NOTE: This is purely theoretical! I have not yet seen a finished buck-boost driver that can work as high as 8V. Don't even think about hooking a FlexDrive and two Li-Ion cells to your precious diode!) An other, more realistic example would be what (I believe) the FlexDrive was originally intended to be used in – take a 637nm red diode with a dropout voltage of around 3.3V and three NiMH rechargeable AA batteries – these can have a total voltage of 3.6V or more when charged and discharge to 3V or less. The idea is the same – when the supply voltage drops close towards the requirement, we switch to boost and enable ourselves to more time spent firing our lasers and less time spent charging. OR, since these situations are becoming progressively rarer and rarer, and we are becoming lazier and lazier, we can use these guys as universal laser drivers – got a 405nm project? Pop a buck-boost driver in and it'll spend the rest of its life working in the boost mode. It's a bit of a wasted potential, but hey, if you're building a red laser next, you can pop in the very same driver! Again, making it spend all the time as buck, but it's very convinient at the same time.
And that concludes this post. If there is any other brand new and exotic way of driving laser diodes invented, I'll try to add it somewhere around here. If you're still reading, thank you for your time, you are awesome
and I hope that you find the information contained useful. If there are any errors in it, both factual and gramatical, or you just have something to add, let me know! I'll be more than happy to improve my work and I will ofcourse credit you. Now just for a little disclaimer...DISCLAIMER: The author of this post is NOT RESPONSIBLE for any direct or indirect damage caused by the usage of information contained in the above text, including but not limited to, on health or equipment!
NOTE: Short citations of this text are allowed as long as you credit the author and include a link to this page in your work. Please contact the author for further permissions.
Last edited:
