• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 872 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Strange output fluctuation when components heated/touched

Craig Easdale19
Expert 1425 points
Other Parts Discussed in Thread: LM3429

I noticed while modifying something else with a heat gun nearby, the output from my LM3429 was dipping/increasing unexpectedly. After messing around a bit I noticed that waving the heat gun over the  dimming part of the PCB from about 5 inches away caused this behaviour to occur. I also noticed that this would happen if I applied pressure to the top of Q2 with a pair of tweezers. I noticed waving a soldering iron around Q2 that having heat at one side will brighten the output, and the other side will dim it.

I don't understand why this would be happening, especially since it was the part recommended in the datasheet and it's for LED driving which will generate heat.

Schematic and layout are below. 

Q2: BC856BDW1T1G
U5: LM4120IM5-5.0

  • 0
    TI__Mastermind 28655 points
    Hello,

    There could be a few things going on here. How what is the dim level when this is happening? CSH is a noise sensitive pin due to being low current.
    Q2 being in the same package should keep the junction temperatures near the same (what you want for a current mirror). If they vary between each other the dim level will change but shouldn't be that drastic unless it's dimmed very low. Look at the collectors of both Q2s when this is happening to see what they are doing. Q3 will vary the dim level with regards to temperature due to Vbe variation as well but with the ratio between R40 and R38 as it is it should be small.
    My first guess would be a noise coupling issue. When probing the different points of the circuit watch for changes in light output/current.
    I looked at one datasheet and couldn't find the information on the BC856 regarding leakage current. If I recall the BC856 is a rather large die part so this could be a factor. The full range for CSH is 0-100 uA which is pretty low current and requires care in layout and parts choice.
  • 0 in reply to Irwin Nederbragt
    Expert 1425 points

    At the point I tested this, and saw the most noticeable variation it was at about the 20% output level. Admittedly it doesn't look so bad at higher levels, but it would be an issue if the level set at "20%" by the user varied wildly.

    I think you might be on to something when you talk about the junction temperatures varying between each other. Holding a low-temperature soldering iron close to one side dims the output, and held at the other side brightens it.

    The temperature around this part should be fairly consistent, but is there a better part I might be able to use in place (with similar footprint/size) that maybe wouldn't respond so badly to fluctuation in temperature?

  • 0 in reply to Craig Easdale19
    TI__Mastermind 28655 points

    Hello,

    When dimming down to 20%, 80% of the CSH current is supplied by the current mirror.  If there is a small percentage change on the current mirror it will really show up in the light output, example; a 10% change on the current mirror would be a 40% change of the light level.  You could try a different PNP pair that doesn't have the capability of the BC856.  Something like digikey part IMT4T108CT-ND.  I don't know if this will make any difference but the transistors are running pretty far down in current compared to their rating.  Another thing I would consider is adding a resistor to the right side transistor to try and keep the Vce of the two transistors closer together.  hfe will vary with Vce voltage.  This shouldn't change the issue you see with the light change though.

    Another route would be to use an op-amp in place of the current mirror.  I would still check to make sure no noise is coupling into the circuit, CSH is a noise sensitive pin.  I would use an oscilloscope to see what is actually changing on the circuit to see what is actually happening.