LMG1020: Help with eval board

Hi John,

What LDO is installed on your EVM? It is the component labeled U1, and is located straight up from the Texas Instruments text.

Measuring the current in the loop is tricky. The issue is, if you break the loop to insert a current-measuring element (resistor or loop for current probe), you will degrade the performance due to the increase in inductance in the loop. We usually recommend customers measure the output of the laser diode to correlate back to the current in the loop.

Hi John,

Im an apps eng with this device and work with Don.
Thanks for reaching out to us on this.
What frequency, pulse width, bus voltage, are you using the LDO?
do you have a dummy load or laser or both current on the EVM?
whats your application?

thanks,

Hi Jeff

 I have tried 1Khz to 1 Mhz, with short pulses and square wave. Without the laser diode attached it does not draw excessive current when I apply the high voltage. Soldering the laser diode in and applying the high voltage my current meter quickly goes to very high, so I turn it off.

I have not tried a short across the pad yet ( I do this with other drivers to see if I can get a 2 ns pulse).

I may have damaged the eval board to cause this excessive current. If so I will order another.

I am building a laser source for a small tactical ground robot. It will be the pulsed source for the 3D Lidar.

I would like to try your device because it says it goes up to 100 amps. I have laser diodes that can go that high.

If I cannot estimate the current in the pulse I will watch the optical power.

If you have some suggested settings let me know, I will try. If they do not work I will get another eval board and try again. If I can get it to work close to 2ns I will design a driver and use it in the production have some prototype testing.

John Hanlin

303 204 8676

4D Tech Solutions Inc.

Hi John,

thanks for explaining in detail,
what do you mean you have not tried a short across the pad yet? you mean you pulse the driver with a 2ns fast pulse?
why do you think you damaged the EVM? are you still seeing switching at Vg and K probe points?
what laser diode are you using? what are you doing for eye safety?
LMG1020 is only the driver for the GAN which drives the laser. There are 2 loops, gate loop with 1020 and power loop with bus caps, GAN and laser. The 100A pulses can get you 300m laser light distances however you need some very tight layout about the size of a stick of gum.
also, you need to remove the 4 parallel resistors (R5-R8) off the EVM if you test with the laser diode. These are for the dummy load for testing LMG1020 right out of the box.

thanks,

Jeff

Thanks for some more details. It was not clear to me that I needed to remove the resistors. I will get some help and do that. So far looking at the Vg and K points I have monitored them and I see the switching is occurring. I can see the voltage increase at the K point. I am interested in two exercises, one to to see what the 2ns pulse through a high energy pulsed laser diode, and if that is sufficient I want to design a driver  for my project. Your device seems to be better than the DEI devices I have used in the past. I really need to do a comparison.

I will let you know once I find someone to take the resistors off the board.

BTW, does TI sell this board with the resistors already off the board? This is not an eye safe project as close as I am to the outputs on my bench.

John

hi john,

sounds good,
what does DEI stand for?
sorry, we dont remove the dummy load but we will make a caution for the user not to run both simultaneously in the EVM users guide.
Hope this helps let me know if you have any more suggestions or questions.

Thanks,

Jeff you have been very helpful. I appreciate the time you are taking for explanation. DEI is Directed Energy Inc. they are pout of Colorado and use to be the only company that made high current switching components for laser drivers.

I am of the opinion that your component GAN gate maybe be better. 

I still have not been able to see a 2ns pulse from the eval board, can you point me to the location where I can measure the voltage on the resistive load.

Thanks again, I am looking for one of my EE friends who has the equipment to remove the resistors.

John

Hi Jeff, sorry to be so dense, but I do not understand the pulse shortener technique. If I input 100Khz with a sort duty cycle I can see about 1.5us pulse width at the K ( cathode ).
If I add up the caps and the resistors I get around 0.4uf and 1 ohm the RC is about 1.75us.

Am I supposed to keep increasing the frequency from the signal generator until I get the 2ns or less pulse width?

John

Hi John,

thanks for the update,
The pulse that is sent from the function generator J3 should be about 100ns at first so you can see the pulse shortener working.
the pulse shortener can be seen working by probing the input header pin on the input J3, while also probing IN+ or TP8 to see the shorter pulse occur. From my testing Ive see that 78ns J3 input gives a 1-2ns pulse on TP8. The frequency is independent of the pulse shortener.
For the 1.5us GAN pulse on the K waveform, are you still testing with both dummy load and laser installed? you have to look at TP8 to see the exact pulse width you are applying to the LMG1020 and therefore the GAN/laser combo.

thanks,

I am getting the four resistors disconnected from the eval board. I should have them back this week and will install my laser for testing. I will send you some results. I am in hopes that I can build a driver using your components for this project.
John

thanks for the update John, looking forward to your results.
Thanks,

Jeff, the 4 resistors are removed now. I put a laser diode on the pads, Following the direction in the description I started the frequency generator at 100Khz. As soon as I apply the high voltage, a few volts, I get very large current drain. I started to increase the frequency and I noticed the current went down. so I tried to balance the two. Since I really do not understand how the circuit actually functions I am worried about damage to the laser diode and the circuit board. With all the inputs off and disconnected and I measure across the pads with the laser I get less than 0.1 ohms, I wonder about that.
I applied the inputs again and ran the frequency high to Khz I started to see a sine wave shape at the K contact. As I ran the frequency higher the valley of the wave started to show some higher frequency oscillations. I was trying to keep the current from getting too high so I stopped.

Is this some of the functions I am supposed to see? The high currents worry me because other drivers do not have this effect when the lasers are attached.
Is there an explanation of how this circuit is suppose to function with a laser diode available to read.?

Thanks
John

Hi John,

thanks for your update!
The operation of the LMG1020 EVM circuit is similar to a lowside FET switching power GND to the laser anode to pulse the laser using the charged up bus caps. The equation that governs the circuit is Ldi/dt = v which tells us that the faster the current through the laser happens the larger the voltage drop across the source inductance L, this drop can affect the gate driver with unwanted noise and ground bounce. This is why the loop inductance needs to be on the order of magnitude of the change in current.

You can check out the TIDA users guide as well as he EVM users guide using the links below:
www.ti.com/.../tidue52.pdf
www.ti.com/.../snou150.pdf

let me know if you have any more questions,
thanks,