Post Go back to editing

LT3467 Questions

Hi all,

I'm attempting to use an LT3467 to generate a regulated +/-15V from a 9V supply. I'm using this schematic as outlined in the datasheet:

I have this circuit built on a breadboard and it is working well. It's holding a steady +/-15V @ 350mA load (roughly 175mA per rail). I'd like to be able to set the output voltage by way of a 3 position slide switch, each position connected to a different value R1 resistor, as illustrated in this schematic:

My questions are:

1. Will there be any risk of damage to the LT3467 if the selector switch was to change to a different position while powered on? I'm thinking possibly if it were to drop down from 15V output to 12V output for example.

2. Can someone explain to me what the purpose of the 1 ohm resistors are? My first thought is that they'd need to be able to dissipate a lot of power (like 15V x 175mA = 2.6W) which would make them quite large, although maybe that's not the case. Am I thinking about this correctly? 

Thanks for any help!

Parents
  • Hi birdmin,

    1. Mechanical switches are very noisy and produce a lot of switch chatter when closing or opening. Also, this is not a make before break solution and so there will be an open circuit however briefly. This solution will not be reliable and will likely produce IC failures.

    The recommended method is to alter the pull-down resistance (R2 in the first picture). One resistor would be from FB node to GND which produces the maximum desired output voltage and then have the desired number of resistors from the FB node to GND in parallel with this resistor which is switched by a small n-channel MOSFET on the low side. This method has the following benefits:

    1. There is always a resistor divider present to provide maximum desired voltage feedback.
    2. The switched-in resistors do not open the feedback path but merely reduce the effective resistance.
    3. The MOSFET will provide clean switching with no noise or chatter.

    2. The one ohm resistor (R4) in the negative output making switching edge less sharp providing better EMI and reducing the affect on the positive rail.

  • Hi Ron, 

    Thanks for the reply! 

    Ah yes, I definitely see now why paralleling resistors is the better way to do this.

    I may have this completely wrong, but would the MOSFET switching you're describing be something like this?..

    I'm probably not understanding the idea behind this well enough but, in this schematic:

    1. I'm taking the input 9V here to turn the MOSFET on. Is this correct? Otherwise, what controls the MOSFET?

    2. I'm still using a mechanical switch to control the MOSFET. Am I negating the benefit of the MOSFET?

    3. When I run the 15V simulation, I'm only getting +/-13.9V even though the total parallel resistance of R2 and R1 are 13.3k as the original schematic. Is this likely because I need to take into account the 'Rds on' of the MOSFET?

    4. Is there any reason I should be specifically paralleling R2, and not R6? The resistors in parallel with R6 would have to be pretty large, around 500k. Maybe this would be bad for noise/EMI?..

    And thank you for the explanation of the 1 ohm R3 and R4 resistors. I asked a similar question on stack exchange and someone replied there that they serve to "add thermal surge sharing with the Schottky diodes". Is this also another reason for these 1 ohm resistors?

Reply
  • Hi Ron, 

    Thanks for the reply! 

    Ah yes, I definitely see now why paralleling resistors is the better way to do this.

    I may have this completely wrong, but would the MOSFET switching you're describing be something like this?..

    I'm probably not understanding the idea behind this well enough but, in this schematic:

    1. I'm taking the input 9V here to turn the MOSFET on. Is this correct? Otherwise, what controls the MOSFET?

    2. I'm still using a mechanical switch to control the MOSFET. Am I negating the benefit of the MOSFET?

    3. When I run the 15V simulation, I'm only getting +/-13.9V even though the total parallel resistance of R2 and R1 are 13.3k as the original schematic. Is this likely because I need to take into account the 'Rds on' of the MOSFET?

    4. Is there any reason I should be specifically paralleling R2, and not R6? The resistors in parallel with R6 would have to be pretty large, around 500k. Maybe this would be bad for noise/EMI?..

    And thank you for the explanation of the 1 ohm R3 and R4 resistors. I asked a similar question on stack exchange and someone replied there that they serve to "add thermal surge sharing with the Schottky diodes". Is this also another reason for these 1 ohm resistors?

Children
No Data