TPS40090: QCW Laser diode driver, 18V 180A.

Hi Robert,

DROOP pin voltage is used as internal reference voltage.
100k between Vref and Vdroop pins is too large. Can you use 0ohm resistor and try?

Thanks
Qian

Hi Qian,

thank you for your help.

I would prefer to explain to you how we are using TPS40090 to drive our laser diode.

Our first purpose to realize our laser diode driver was to give the reference voltage signal to DROOP pin to get the same output current waveform. Our intentions were to vary the amplitude of the reference voltage signal  on DROOP pin to get an output current value varying from 30A to 180A. 

We realized the feedback with a sense resistance of 1 mohm (200 mV at 200 A) and a differential amplifier with its output linked to te FB pin through 10 Kohm and a gain of  5. 

We verified that lowering the reference signal voltage amplitude on DROOP output current did not vary at all (fixed to 200A) untill, under a certain threshold, the controller showed its protection feature.

Because of this unwanted behavior, later, we doubled the sense resistance value and we noticed current splitted in half (100A). With this result in mind, we tought to go back to the sense resistor of 1 mohm and to use a variable gain amplifier (LMH6505) to vary the voltage signal amplitude measured on the sense resistance. We stated varying the gain from 1 to 10 we got a varying current from 20A to 200A.

The problem is sometimes and randomly the controller show its protection feature also at low current. We think feedback loop is not reactive enough to regulate output current.

How we could resolve this problem? About you is TPS40090 suitable to our purpose of driving in current-controlled pulsed mode a laser diode?

Attached you will find in blue the output current sensed with a fluke current probe (consider 1V/1A).

Thank you for your help.

P.S. We found SLUS578B document like more to an application note rather than a datasheet.

Hi Robert,

TPS40090 is a peak current mode controller, but it may work at voltage mode.
I think the better way for your application is to let the controller work at voltage mode, then use the output of external current sense amplifier to control TPS40090 COMP pin voltage directly, because COMP voltage determine the PWM width directly hence determine the current you wants.

Thanks
Qian

Hi Qian,

we gave it a try in open loop with the following controller's pin configuration:

- FB = 0,7V;

- DROOP floating,

- 0-5 V potentiometer on COMP.

The PWM modulation seems work fine, but we have some other issue. I'll let you know when the whole system will work.

Thank you a lot for your help.

Roberto

Hi Qian,

don't ask me why but with our configuration (FB= 0,7V and DROOP floating) varying the voltage on COMP the duty-cycle varies linearly and continuously.

With the configuration you suggested we saw only some periodic burst (that is the component modules for some period then shuts down, then restart).

In our configuration we observed the behavior showed in the video: in blue TPS40090 PWM output (ch1), in yellow MOSFET gate signal (after the MOSFET driver).

When we supply the TPS40090 and the power stage is off all goes well (frequency is ok, duty-cycle is ok).

When we switch on the power stage (24 V) PWM controller's output signal is heavily reduced. Why?

We made this tests in open loop.

Thank you

Roberto

Hi Roberto,

I suspect the problem that PWM signal is reduced when power stage is on is related to current sense signal.
Can you remove R62, R68, R69 and R71 in your design and check again?

Thanks
Qian

Hi Qian,

already done, yesterday before I wrote to you. We noticed that behavior in both cases.

Thanks

Roberto

Hi Roberto,

Can you check if there is noise on COMP and SS pin when power stages are on?
Also, can you short DROOP pin to REF pin (0.7V) ? FB is already connected to REF (0.7V),right?

Thanks
Qian

Hi Qian,

I don't know if is this the right place to write to you, but anyway we need some more help.

We observed a strange behavior. The controller works more or less fine (we have to improve our feedback network), but sometimes, and mostly when it drives currents over 130-140 A, PWM signals on the four channels disappear and their output stuck at 1,21 Vdc so we have to change the TPS40090 every time it happens.

We are putting our reference voltage on DROOP and the feedback is connected to FB pin. Other signals are ok: SS oscillates, Vref is ok and BP5 also.

I don't know why we have this behavior, I supposed the problem could be electromagnetic but I'm not sure. Have you any suggestion?

Thank you.

Roberto

Hi Roberto,

After the issue happens, if you power recycle the board, will the problem goes away?

Is there resistor between DROOP pin and VREF pin, or DROOP is driven by external circuit only?

Thanks

Qian

Hi Qian,

if I power recycle the board the problem doesn't go away (we replaced many controllers).

The configuration we set is the following:

- A Rsense of 1mohm connected beteween VOUT and GNDS.

- From DIFFO to FB pin with 4,7 Kohm (from COMP and FB we have our feedback network).

- No resistor between DROOP and VREF.

- DROOP is driven by external circuit (we measured 100mV of adjustment range, starting around 620 mV).

- CSx pins are all shorted to ground.

P.S. Another question: internal current generator at DROOP pin is somehow dependent from CS pins?

Thank you

Hi Qian,

the PWM outputs of the TPS40090 don't stuck at 1,2V but to 0V: parctically the modulator doesn't start. The rest is the same. I apologize for the wrong information.

Thank you 

Roberto 

HI Roberto,

Below information is captured from datasheet page 10. 

During power on, if the part detect any CSx pin is shorted to ground, it will not startup. 

i suspect the problem on your board is related to this function. 

Hi Qian,

everything is clear. But how can I solve this issue in voltage-mode? A weak pull down (100 ohm for example) could be sufficent?

Thank you.

HI Roberto, 

Can you short CSx to CSCN pin, and connect CSCN pin to ground via a resistor, such as 100ohm?

Thanks

Qian