DX 30mW Green - V(in) vs I(LD) & Board Pic

[chimo]

I opened up my DX Green 30mW again to get some better pics.  I also ran some tests for various input voltages and measured the voltage through the 1 ohm sense resistor.  Here are my results:

Vin = 2.4V -> Isense = 239mA
Vin = 3.0V -> Isense = 287mA
Vin = 3.5V -> Isense = 324mA
Vin = 4.0V -> Isense = 360mA
Vin = 4.2V -> Isense = 374mA
Vin = 4.5V -> Isense = 394mA

With a couple of partially depleted Lithium L92s, the draw from the batteries was 277mA.

Heres a pic of both sides of the board. I removed the switch so I could trace the vias.

Schematic from my post below.


Paul

 

[Benm]

Interesting.. i have a dx20 i'm unable to open, bar really brute force.

Were you able to change output power using the pot, and if so, how did that affect the current readings?

Also, is the text on the IC legible (hard to see in the pic)?

 

[chimo]

Interesting.. i have a dx20 i'm unable to open, bar really brute force.

Were you able to change output power using the pot, and if so, how did that affect the current readings?

Also, is the text on the IC legible (hard to see in the pic)?

It's LA6358NT - a dual op amp.

To open, use the dual wooded block/vise method and some heat.

I did not mess too much with the pot.

 

[Benm]

Was there any sort of compound between the head and the rest of the laser that required melting? Or is it only a matter of expanding the head bit just a tad?

 

[chimo]

Was there any sort of compound between the head and the rest of the laser that required melting? Or is it only a matter of expanding the head bit just a tad?

There was some adhesive that was weakened with the heat.

 

[yew.sir-name]

Is the pot on the underside of the board?

 

[chimo]

Is the pot on the underside of the board?

On mine - Yes. However, I understand that there may be different versions of this pointer with different driver boards.

I am still working on figuring the schematic - lots of hidden vias under things.

Paul

 

[Things]

the pot was on the top of my board, so they must be different versions?

 

[chimo]

I have had a crack at creating a schematic for my DX 30.  Here it is.  It's not surprising that the pot does nothing!  Note that there is no feedback for the photodiode!!!

If the part marked 431 is a LM431 equivalent, Vref should max out at 2.5V.  

The voltage divider for the op amp fed from the sense resistor should swing at 515mV (which would equate to a 515mA current through the LD - perhaps a more suitable current for the 200mW model).  



Paul

 

[Hemlock_Mike]

The pot appears to set some bias point for the driver transistor but it is swamped out by the feedback circuit driven by the limiting resistor. I'd work in that feedback area. Always monitor LD current.

Mike

 

[chimo]

Mike, that's the existing circuit schematic from the board on my DX30. It certainly explains why the pot does nothing. I have seen a design that's fairly close to that one on Sam's FAQ, but that one used one of the op amps for PD feedback. I'm not overly impressed with the DX circuit - but.... I guess it's worth what was paid for it.

Paul

The pot appears to set some bias point for the driver transistor but it is swamped out by the feedback circuit driven by the limiting resistor.  I'd work in that feedback area.  Always monitor LD current.

Mike

 

[Daedal]

Hmm... that seems pretty simplistic and rather effective if you ask me...

Does that say that the PD is not connected anywhere? :-/ I thought they used PD feedback to control the current :-?

--DDL

 

[chimo]

I have to say I was a little baffled by the configuration. With two transistors, two op amps and 11 resistors, you would think they could have worked in PD feedback. It explains the rising LD current with the increased Vin.

Paul

Hmm... that seems pretty simplistic and rather effective if you ask me...

Does that say that the PD is not connected anywhere?  :-/  I thought they used PD feedback to control the current  :-?

--DDL

 

[Benm]

Are you sure this circuit is correct?

The reference seems to be ill designed really, running of 3 volts it wouldn't even come close to raising the voltage across the 413 to 2.5 volts. I guess that does explain the bad Vin/Ild regulation observed.

It would be possible to improve this circuit quite easily though - switch out the 431, and 15k/3k9 divider for a proper reference, and it should be quite steady at a current of choice.

The part around opamp 1 doesnt do anything at all, it just pulls the output of that opamp to Vcc - connecting the 100 ohms resistor to Vcc would have been easier

 

[chimo]

Are you sure this circuit is correct?

The reference seems to be ill designed really, running of 3 volts it wouldn't even come close to raising the voltage across the 413 to 2.5 volts. I guess that does explain the bad Vin/Ild regulation observed.

It would be possible to improve this circuit quite easily though - switch out the 431, and 15k/3k9 divider for a proper reference, and it should be quite steady at a current of choice.

The part around opamp 1 doesnt do anything at all, it just pulls the output of that opamp to Vcc - connecting the 100 ohms resistor to Vcc would have been easier

I believe it is correct. It took a longer than expected to trace everything because the hidden traces did not all go where I expected. It's certainly a bit of a wonky design. The pot does nothing. Considering the very low cost of these lasers (<$30) - they could have been from a "defective" run that should had been scrapped or they have modified the board's "design".

Paul

 

[Benm]

I guess so... perhaps it was intended to use the photodiode, but they switched to a LD that doesnt have one inside and had to hack it for that reason.

Still, the reference is a complete misdesign for anything intended to run of 2 AA cells. To me, it looks like it's designed to run at 5 volts, and someone put it in a battery operated device. No doubt some component values have been hacked to it still lights up, but any form of regulation is gone.

So i guess you got the circuit drawing right, but the designer did not