Fast modulable Laser diode driver with high side current control

Printable View

05-30-2009, 09:30
andythemechanic

4 Attachment(s)

Fast modulable Laser diode driver with high side current control

Hi everybody,

some time I asked about a modulable driver with high side current monitor ans switching here: http://www.photonlexicon.com/forums/...ead.php?t=7146
From the discussion on this thread a nice driver evolved, which I just finished prototyping and testing. Results look really promising, and I like to share them.
So Voila! Here are the schematic the layout from eagle and a pic of my first test assembly with a bluray laser diode for testing.
Have fun!
In the next post I will show some test results
05-30-2009, 09:35
andythemechanic

2 Attachment(s)

Here are some test results with a 47Ohm resistor as load. Current is 100mA. I have not testet higher currents so far but simulations point out that with some modifications (change of Resistor R1 and R10) currents up to 10Amps shouldn't be a problem (with proper heatsinking)
05-30-2009, 09:40
andythemechanic

4 Attachment(s)

Finally some results with a bluray diode, recorded with an ultrafast silicon photodiode (100ps risetime).
05-30-2009, 09:42
andythemechanic

As can be seen the driver behaves quite nicely at this low currents. It has the big advantage that the diode is grounded. It has softstart and current limiting.
This was actually my first layout in Eagle and I had quite some fun learning and designing this drive.
A big thanks to Steve (aka mixedgas) for the help with this drive!!!
Hope you enjoy!

Andreas
09-02-2009, 15:54
chad

I can't believe I didn't catch this a couple of months ago. Looks good.

How do the drivers work? I saw the pictures of the little projector you finished.

Could you take a couple of pics that might show the modulation of the lasers with your drivers? Maybe a couple of the test patterns?

chad
09-03-2009, 08:27
andythemechanic

2 Attachment(s)

Quote:

Originally Posted by chad

Could you take a couple of pics that might show the modulation of the lasers with your drivers? Maybe a couple of the test patterns?

chad

Hi Chad,

The pics in the third post here is the actual modulation of a bluray diode recorded with a fast photodiode. I tried to make a simple test to visualize the modulation: I scanned a circle with 400Hz and modulated the diode with a higher harmonic (rectangular). Here are the images at 10kHz and 36kHz Modulation frequency. In real it looks better than with my camera. Also I think the modulation in the scanner is more limited by the noisy AudioDac than the drivers, the above was recorded with a prper function generator.
If you have another testframe (in ILDA format) you would like to see you can PM it to me.
The drivers work really good in my machine. The current limiting feature is really nice when adjusting the drivers, you never have to worry to kill your diodes :) . They are not so good for pump diodes because these are case positive so you have to isolate them.

Greetings
Andy
09-03-2009, 08:57
ElektroFreak

Looks very nice. I'll be trying one of these out for sure!
09-03-2009, 09:54
andythemechanic

Quote:

Originally Posted by ElektroFreak

Looks very nice. I'll be trying one of these out for sure!

If you like to test it, I could send you one for parts cost. I made 8 proto boards and ordered all the parts for 8 drivers. Send me a PM if you're interested!
09-03-2009, 10:27
ElektroFreak

PM sent to you!
09-04-2009, 01:58
flyowynd

You're my savior, I've got so many issues isolating my red diodes, I will run to my local electronic dealer to buy the components.

High side current drive is very hard to find, and my previous tests with home made one were not very successful.
09-04-2009, 05:58
chad

Cool! Looks pretty good.

Nice Job!

chad
09-04-2009, 06:17
flyowynd

Hi again

I was looking at your design and maybe I got something wrong but I don't see any AOP on the board you design.
It seems to me that the board and the schematic are different.

Another question, do you think it's possible to do it with normal component instead of CMS. I will have four of these to build for my diodes so it could be easier for me.

Thx
09-04-2009, 14:21
andythemechanic

Quote:

Originally Posted by flyowynd

Hi again

I was looking at your design and maybe I got something wrong but I don't see any AOP on the board you design.
It seems to me that the board and the schematic are different.

Another question, do you think it's possible to do it with normal component instead of CMS. I will have four of these to build for my diodes so it could be easier for me.

Thx

Hi,

what did you mean with AOP, maybe the Opamp ? It's hardly visible on the circuit because it is a very tiny SOT23 package. There is one difference between the schematic and the board, on the board is a 10uF capacitor between opamp V+ and V- which is not in the schematic.
When you want to do it with normal components, you have to find an opamp in a DIL case with similar specs which could be difficult. All other components are available. Hope that helps!
09-13-2009, 06:17
andythemechanic

1 Attachment(s)

Problems

Hi Everbody,

Here an update to this circuit: I changed the current sense and limiting resistors to 0.1Ohms to be able to use the circuit for much higher currents. Now I have a damped oscillation on the current each time I switch it (see attached image). The oscillation is always the same regardless of current, frequency etc.
So here the $64 question to all the electrical engineers out there: Any Idea what this could be?
09-13-2009, 09:20
krazer

By reducing the current sense resistor, you increased the gain and it is now under damped, probably due to delays in the circuit. I am not really familiar with the components you are using, but it could be due to the gate capacitance of the mosfet, or possibly the opamp is just to slow for the slew rate you are asking of it. Also, if you are measuring the voltage at the output of the driver its very possible that the issue is the inductance of the leads (or if you have one, the capacitance of the esd protection capacitor), and the driver is in fact giving a constant current output but that requires varying the voltage a bit.

Good luck!
09-13-2009, 10:50
andythemechanic

Quote:

Originally Posted by krazer

By reducing the current sense resistor, you increased the gain and it is now under damped, probably due to delays in the circuit. I am not really familiar with the components you are using, but it could be due to the gate capacitance of the mosfet, or possibly the opamp is just to slow for the slew rate you are asking of it. Also, if you are measuring the voltage at the output of the driver its very possible that the issue is the inductance of the leads (or if you have one, the capacitance of the esd protection capacitor), and the driver is in fact giving a constant current output but that requires varying the voltage a bit.

Good luck!

Hm Thanks, but I don't think it's that easy. Slew rate of the Opamp is really high (130V/us) and the mosfet was the one with lowest gate capacitance for the wanted power rating.
Also it should be possible to compensate this effects, by changing C4,R9 and R6 and R4. Doing this had no effect at all on this oscillation it always stayed there and with the same frequency.
My nex thought was, that it could be parasitic capacitances on the inputs of the opamp. To test this I removed the copper of the ground plane under the opamp and the leads toward it, no effect.
Any other ideas what it could be?
09-14-2009, 02:29
flyowynd

Hi

I may be completly wrong but when I look at the trace, it seems that you have an instability like in a simple proportional driver. I've seen the same when driving motors and you want a precise control of it. That's why we use PID drivers.
But normally using a single proportional driver will cause the first oscillation to be above the targetted point.

I pretty sure it won't help but who knows.
09-14-2009, 04:01
andythemechanic

Quote:

Originally Posted by flyowynd

But normally using a single proportional driver will cause the first oscillation to be above the targetted point.

When you look at the rising edge of the pulse closely you see that the oscillation starts with going down a little bit.
Also it is not a pure proportional driver, the capcitor c4 integrates a bit. If it is something like this, shouldn't I be able to influence the oscillation with changing c4 ? Maybe not removing it but changing it's frequency? Nothing like this happened. When I make it much smaller then the circuit starts to oscillate at high frequency what is expected. When I make it bigger the rise and fall times get slower but the oscillation always remains like it is. I don't understand this. :(
09-14-2009, 06:19
flyowynd

That's really sound strange.
Just a question like this, in most of the current controlled source I've seen, they are using 2 Opamps one for current sensing and the other as a simple follower to drive the transistor.
Could the problem come from this? adding a second Opamps after the first one will isolate the two parts of the circuit and maybe get rid of the oscillation.
(I'm really not sure of this)
09-14-2009, 10:40
krazer

I still suspect that delays in your circuit may be contributing to the issue.

Looking at your mosfet (I can't quite read the part number, but I am guessing it is a IRLU9343) has a gate charge of about 5e-9 C Your opamp is supposed to be good for .1A, which can charge the capacitor in 5e-8s, or 50ns. The fet also has about 30ns of delay times on the turn on (10ns delay time, 20ns rise time), which all adds up to somewhere between .05-.1us of delay, which really isn't negligible in the circuit.

Although thinking about it more, the fact that the current takes a dive before it hits the set point is more worry some, and I have a hunch is being caused by stray inductance in the circuit somewhere. You did use a low inductance .1r resistor, right?

Also, what are you using for your dummy load? And what current are you running at for that matter.
09-14-2009, 11:04
andythemechanic

Quote:

Originally Posted by krazer

I still suspect that delays in your circuit may be contributing to the issue.

Looking at your mosfet (I can't quite read the part number, but I am guessing it is a IRLU9343) has a gate charge of about 5e-9 C Your opamp is supposed to be good for .1A, which can charge the capacitor in 5e-8s, or 50ns. The fet also has about 30ns of delay times on the turn on (10ns delay time, 20ns rise time), which all adds up to somewhere between .05-.1us of delay, which really isn't negligible in the circuit.

Although thinking about it more, the fact that the current takes a dive before it hits the set point is more worry some, and I have a hunch is being caused by stray inductance in the circuit somewhere. You did use a low inductance .1r resistor, right?

Also, what are you using for your dummy load? And what current are you running at for that matter.

Yes the sense resistor is a low inductance thick film resistor. Test load was in this case just a 0.5 Ohm resistor. Wires were short arround 3cm or so. Peak current is 2A but the effect is also at 0.5A there.
I also think it is some parasitic effect. Maybe the only way to find out is to make a new layout. I think I will wire the circuit without board just with wires and see if the effect is still there. Just to exclude parasitic effects from the layout.