Erdabyz's switching laser drivers

Well, being both fluorescence scanners, must contain both a laser able to emit in UV, only, i don't have idea about what type of lasers they use (models and powers, i mean)

Maybe the seller left you look at the inside, and you can see the tags on the lasers and the date of productions, for choose the better one .....

HIMNL9 said:

Well, being both fluorescence scanners, must contain both a laser able to emit in UV, only, i don't have idea about what type of lasers they use (models and powers, i mean)

Maybe the seller left you look at the inside, and you can see the tags on the lasers and the date of productions, for choose the better one .....

Click to expand...

Well, as Daguin has pointed out here http://laserpointerforums.com/f46/tampa-florida-come-down-42658.html the Perkin Elmer prism 377 does have an argon inside. He also has a thread that shows how to extract one of these. I'm not sure, and this is definatly a question for Daguin to answer, but the genomyx may have an alternate light source like LED's or something.

Hey guys!

About HIMNL9's design simulation, it gave a wierd graphic in the simulator, giving a fluctuating and not variable current close to the max.
However I think that is something related to gain because the regulator cound't get more than about 600mV in its FB pin.

I designed today another driver with the resistor trick based on a LT1930.
I'll post details later.

^ You mean the one with 1723 and 1607 ? ..... it was a principle scheme, need to be refined.

If you can do simulations, try to add two capacitors, an 1uF from the output of the 1723 and ground, and another 2,2uF in parallel to the LD, for see if it can take away the oscillation (you still have the input capacitor, right ? ).

It's strange, about the FB pin, cause the datasheet says that it's a high impedance input, so it must not influence enough the 10 Kohm trimmer (that is one of the values that set the gain of the 1607 ..... anyway, if the problem is the gain, using 220 ohm instead 150 for the Rin , must increase it almost 1,5 times.

BTW, what do you use for simulations ?

HIMNL9 said:

^ You mean the one with 1723 and 1607 ? ..... it was a principle scheme, need to be refined.

If you can do simulations, try to add two capacitors, an 1uF from the output of the 1723 and ground, and another 2,2uF in parallel to the LD, for see if it can take away the oscillation (you still have the input capacitor, right ? ).

It's strange, about the FB pin, cause the datasheet says that it's a high impedance input, so it must not influence enough the 10 Kohm trimmer (that is one of the values that set the gain of the 1607 ..... anyway, if the problem is the gain, using 220 ohm instead 150 for the Rin , must increase it almost 1,5 times.

BTW, what do you use for simulations ?

Click to expand...

Well, to be sincere, I was using LTSpice IV (the free simulation program from Linear) and I wasn't using a 1723 but a LT1930 connected in the same way than the 1723 in your schematic. It doesn't matter because what we wanted to test was the amplifier circuit. I used that program because it is very easy to use. The problem was that the FB pin couldn't reach the required voltage so the current goes up to the max and starts fluctuating. I varied the values of the trimpot (had to use two resistors because there's no trimpot available in that program) and nothing...

I'll resimulate it and post the graphs obtained.

Oh, and I think let's stop making things with the 1723 and give all hail to the Hyp.... 1930. The 1723 based drivers I did had a small issue, they require very solid connections or they won't work. For example they required a spring to push the battery against the positive contact to work. Nothing important as most hosts have negative springs that make enough force. However, with the 1930 nothing of that happens. The driver I did yesterday just WANTS to work. And as the 1930 works only above 2.5V, a simple schottky diode in the input would work as battery protector.

Bump, can't wait for these drivers. Oh, and only the Prism 377 had an argon, apparently the Genomyx used some sort of radioactive liquids or something to sequence. Acordding to the buyer, when he finally opened it up.

Hey! just to say that I just found an IC that is just what we've been asking for all this time. IT HANDLES UP TO 2.8AMPS and UP TO 26 VOLTS! and what's even better, it has just 190mV as Vref, meaning that we can use really small value resistors (in the range of 0.1-1 ohms). And what's even better, it cuts off at 2.7 volts Vin, which is the cutoff voltage of a lithium battery, meaning that it would perfectly work with unprotected cells because it wouldn't underdischarge them!!!
It will drive 8x's at around 85% efficiency, and up to 1amp

The final driver will give up to 1amp at 10 or 16 volts (but upgrading it would be a matter of changing a schottky diode and a capacitor)

And it has also another feature, PWM current mode dimming!!! Yeah, the driver will support dimming

I am currently designing how the adjusting potentiometer would work, as I can't find less than 10ohms potentiometer, and I'd need a 1ohm one. This time i think i'm gonna "mass produce" and sell this drivers.
So stay tuned! it won't take very long until the first prototype arrives!
It will be blu-ray only however, but I guess that's not a big issue as there are many drivers available for reds...

erdabyz said:

Hey! just to say that I just found an IC that is just what we've been asking for all this time. IT HANDLES UP TO 2.8AMPS and UP TO 26 VOLTS! and what's even better, it has just 190mV as Vref, meaning that we can use really small value resistors (in the range of 0.1-1 ohms). And what's even better, it cuts off at 2.7 volts Vin, which is the cutoff voltage of a lithium battery, meaning that it would perfectly work with unprotected cells because it wouldn't underdischarge them!!!
It will drive 8x's at around 85% efficiency, and up to 1amp

The final driver will give up to 1amp at 10 or 16 volts (but upgrading it would be a matter of changing a schottky diode and a capacitor)

And it has also another feature, PWM current mode dimming!!! Yeah, the driver will support dimming

I am currently designing how the adjusting potentiometer would work, as I can't find less than 10ohms potentiometer, and I'd need a 1ohm one. This time i think i'm gonna "mass produce" and sell this drivers.
So stay tuned! it won't take very long until the first prototype arrives!
It will be blu-ray only however, but I guess that's not a big issue as there are many drivers available for reds...

Click to expand...

Is it similar to this IC http://www.national.com/ds/LM/LM3410.pdf ? Just curious

Ramsey_innovations said:

Is it similar to this IC http://www.national.com/ds/LM/LM3410.pdf ? Just curious

Click to expand...

It wasn't that hard right?


A piece of win like that requires a very well designed PCB however to give all its powah. Let's see what it can do, i'll receive some samples in 1-2 days and first prototype will be done like this weekend.
I'm still figuring out how to make an adjustable driver out of that IC, as it will be kinda hard to adjust such small resistances required... I have some methods but all of them require "high" value resistors. Maybe the best would be just making different models with different ranges.... or just make them fixed current.

wow, good news!
that sounds like the perfect driver of the future!
if its small enough, that is.. for unprotected "feature", the notoriously unprotected 10440 cells come to my mind. if the driver isnt too big..? perhaps long and slim, which would be ok for the aixiz too?
then a larger one with integrated pwm? yay!

will definitely watch this thread!

manuel

Uhm, i'm very sorry to have to brake a bit some enthusiasm, but the 2,8A from the datasheet is referred to the maximum current that THE INTERNAL MOSFET can switch, not to the maximum current that you can draw from the component .....

I also tried it (but still not finished my tests), and can't get it working more than 600 mA, at least til now (but can always be due to my errors, cause i can work on these things only occasionally) ..... i also made the same mistake, the first time i started to experiment with boost drivers ..... i had found an IC that was looking perfect for me, constant current loop, 2,7V input, til 12V output, 1A switching rate ..... my enthusiasm fall under the shoes, when after some tests i realized that 1A was the mosfet current, and that it can give just 140mA max to the load

I cannot believe I have missed this thread. Great information about some new driver designs. So I thought, I'd share two of mine.

My first successful attempt at a driver was with the TPS61200 regulator from TI. The 61200 is a step-up voltage regulator with "down-mode". Starts up from as little as 0.5V Vin, but the Vout is limited to 5.5V. Using an idea from an old thread here at LPF I used a Zetex 1009 high-side current sense amplifier to make the TPS common ground and be able to use smaller sense resistors.
The PCB is rather small and fits into an Aixiz module easily. Drives all of my low power PHR builds. Efficiency isn't great (70-80%), probably due to my poor PCB design Substitue a TPS63000 for the 62100 and you get full buck-boost capability.

For driving an 8x I needed to overcome the 5.5V limit of the 61200. So I looked at LED drivers and found the TPS61165: Very small (2x2mm package), 18V maximum Vout and being a LED driver it is current regulated from the start and requires just a few external components. On the downside you loose common ground (not a problem for BluRay LDs as of now) and it needs at least 3V to run.
The PCB is even smaller than the driver described above and my 8x runs happily on a single 14500 Li-Ion cell at 290ma putting out ~400mW.

HIMNL9 said:

Uhm, i'm very sorry to have to brake a bit some enthusiasm, but the 2,8A from the datasheet is referred to the maximum current that THE INTERNAL MOSFET can switch, not to the maximum current that you can draw from the component .....

I also tried it (but still not finished my tests), and can't get it working more than 600 mA, at least til now (but can always be due to my errors, cause i can work on these things only occasionally) ..... i also made the same mistake, the first time i started to experiment with boost drivers ..... i had found an IC that was looking perfect for me, constant current loop, 2,7V input, til 12V output, 1A switching rate ..... my enthusiasm fall under the shoes, when after some tests i realized that 1A was the mosfet current, and that it can give just 140mA max to the load

Click to expand...

erdabyz?

I have a small number of LM3410s, I was going to make a pair of drivers with it, but HIML9 was correct, the MOSFET's current limit is 2.8A, meaning the max output is around 700mA or so. You can create an example design on their online simulator tool, however, which I found very useful. If anyone was interested, I could probably do soldering of a small batch (with hot air gun), but I can't currently use Eagle, so I couldn't make the PCBs for mass or batch production. If anyone's interested in seeing precisely what these chips can do, I could probably invest a few hours learning Eagle and ordering some inductors/capacitors. For SMT pots, LPC-815 sleds have some good ones I could probably re-use on the new design...

Well, yeah, 2.8 Amps is the maximum it "could" give but it also doesn't mean it couldn't give 1 amp or so. I knew it wouldn't give 2.8 amps, it's just "impossible"
Getting high currents, however, it's just a matter of getting the proper inductor with a saturation current like 25% bigger than the current you want it to give and a big enough output capacitor. 1 amps sounds perfectly doable for me with the proper componets. It's also a matter of how much current a cell can deliever. Think that if it has to give 1 amp at 6 volts, we have 6 watts of power. Considering a 80% efficiency, it would drain 7.5W from the cells, that at 4 volts mean that it would drain 1.8amps from the cells.

Yeah, some testing has to be done and a good design is just crucial to get the best out of this IC's.
Also, with "mass production" I meant about 25drivers/month. Nothing more


The problem with this IC will be making it adjustable, as we need resistance values in the range of 0.19-3 ohms.


quantile, did you include some sort of capacitor in paralell with the amplifier output? All my tests with high side amplifieres turned out to be a fail, because the driver was faster than the amplifier so the current fluctuated a lot. A capacitor in paralel with the output of the amplifier made the vref stable but also increased a lot the stabilization time, meaning that the driver was giving max current (around 700mA) for about 0.5-1 millisecond, enough to kill or damage a diode.


HIMNL9: I doubt a lot that your driver could only give 140mA to the load having a 1A switch. You did something wrong.
I did extensive testing with the LT1930 and i can assure you that this IC, having a 1A switch, can perfectly give around 650mA at 5.5V from a 4V source. Even more with LT1930A. It's REALLY important to make a well designed PCB for this drivers and to carefully pick the inductor and output capacitors. A fail in that will result in a fail in performance.

erdabyz said:

quantile, did you include some sort of capacitor in paralell with the amplifier output? All my tests with high side amplifieres turned out to be a fail, because the driver was faster than the amplifier so the current fluctuated a lot. A capacitor in paralel with the output of the amplifier made the vref stable but also increased a lot the stabilization time, meaning that the driver was giving max current (around 700mA) for about 0.5-1 millisecond, enough to kill or damage a diode.

Click to expand...

No, I did not, it just worked in my naive "just plugging it together" approach. But, the only regulators I was able to get working with the high side monitor were the TPS61200 and TPS63000. I tried some others and failed, because the Vref didn't stabilize. I just had another look at the datasheet for the TPS61200, it states that it uses average current mode control. Could it be that that makes them slower to respond to the FB (thus giving the amplifier enough time to stabilize)?
Thank you for the suggestion with the capacitor in parallel to the amplifier, I think I'll have another go at this in the future.