negative voltage lt3453 according to Design Note 359

Hi

I'm designing a powerboard which needs multiple positive and negative output levels. In such I need a +-40V output.

Therefore I used a setup which first generates a stable 12V input (by using http://www.ti.com/lit/ds/symlink/lm1085.pdf, verified 12V stable level) going into a lt3463, which is designed as the second example of Design Note 359. 

First to be sure: I'm not entirely sure if this topology is the inverting charge pump regulator, yet if I read the specs voltage in this design example outreaches the max voltage at pin SW2 (42V) so inverting charge pump design is suggested. As I don't see any inductor in the output stage I'm guessing this should be the inverting charge pump topology, correct?

My design would otherwise put 52V at SW2 pin, which would kill it ofcourse.

Now I get the positive side of the output (+40V) yet my output stays around 0V (some mVs, so guessing multimeter error). All connections have been verified and I tried swapping even all components which influence the negative side of the LT3463.

When checking the internal diode (D2) I get a reading of approx. 0.7V so seems correct functioning.

SW1 as well as SW2 measure approx 50Ohms to ground. 

shdn1 & sdhn2 are tied high (at the 12V input).

All resistors, input inductors  and schottky diode are chosen equally as in the design paper. Output caps I did choose larger, 0.47uF instead of 0.22uF but shouldn't have influence.

Is there any way to verify where it's going wrong if not for the design?

In attach my LTSPICE file and a picture of eagle layout of this part.

 spice_lt3463.zip

Parents
  • 0
    •  Analog Employees 
    on Feb 4, 2019 4:35 PM

    Hi, Wouter,

    The negative converter is a negative charge pump. as you have correctly assumed, so during operation, the SW voltage should be similar to the Boost switch. 

    I cannot comment on your layout because i cannot open it, but your schematic appears to mimic the simulation and the schematic in DN359, except your input voltage is higher, 12V. This higher input voltage might not be the reason for causing the converter to malfunction, but it is a reason for concern going forward. I would recommend you power the LT3463 from a 5V input source, or lower,  and use the capacitor values used in DN359. The reason for that is explained on page 6 of the datasheet, "Inrush Current" paragraph. 

    It is not clear to me whether the boost is working or not.

    First you say,

    "Now I get the positive side of the output (+40V) yet my output stays around 0V (some mVs, so guessing multimeter error)."

    When you say "you get it", do you mean you understand it?

    My assumption is both outputs are dead, since later on you say both switches have low impedance to ground. 

    If that is so, remove the inductors and see if the impedance clears. If you still read low impedance after isolating the switches, the LT3463 is dead. 

    If the switches read high impedance, find out where the short is and clear it. 

    After you fix the low switch impedance issue, try powering the converter with a lower input voltage, 5V or less. 

    If your required load current is more than about 5mA, this might not be the converter for you.

Reply
  • 0
    •  Analog Employees 
    on Feb 4, 2019 4:35 PM

    Hi, Wouter,

    The negative converter is a negative charge pump. as you have correctly assumed, so during operation, the SW voltage should be similar to the Boost switch. 

    I cannot comment on your layout because i cannot open it, but your schematic appears to mimic the simulation and the schematic in DN359, except your input voltage is higher, 12V. This higher input voltage might not be the reason for causing the converter to malfunction, but it is a reason for concern going forward. I would recommend you power the LT3463 from a 5V input source, or lower,  and use the capacitor values used in DN359. The reason for that is explained on page 6 of the datasheet, "Inrush Current" paragraph. 

    It is not clear to me whether the boost is working or not.

    First you say,

    "Now I get the positive side of the output (+40V) yet my output stays around 0V (some mVs, so guessing multimeter error)."

    When you say "you get it", do you mean you understand it?

    My assumption is both outputs are dead, since later on you say both switches have low impedance to ground. 

    If that is so, remove the inductors and see if the impedance clears. If you still read low impedance after isolating the switches, the LT3463 is dead. 

    If the switches read high impedance, find out where the short is and clear it. 

    After you fix the low switch impedance issue, try powering the converter with a lower input voltage, 5V or less. 

    If your required load current is more than about 5mA, this might not be the converter for you.

Children
  • Hi

    as a start, I need more then 5mA of output so I guess the IC isn't the one I need. How should I have read the datasheet to find this out? 

    I've been looking further in that case to split my outputs to get higher currents (I need max 600mA each output), would it be a better combo to use an LTC3786 to generate the +40V and an LTC3896 to generate the negative -40V?

    "Now I get the positive side of the output (+40V) yet my output stays around 0V (some mVs, so guessing multimeter error)."

    When you say "you get it", do you mean you understand it?

    Sorry badly written: when I'm probing I get +40V at one output, yet the other output (where I expect -40V) got approx 0V. Therefore one output is definitely not dead, which is weird as they both measure the same path resistance. I'll have to take a look into the path without the inductors, yet they are inline (according to schema in datasheet) so perhaps it is indeed the inrush currents which are my problem.

    Thanks already for the assistance

  • +1
    •  Analog Employees 
    on Feb 8, 2019 3:58 PM in reply to wouter.devriese

    Hi, 

    Most datasheets have a typical application on page one, and several at the end. 

    It is usually a good idea to browse through those applications and see if any of those fits your needs for both, output voltage and load current.

    Your application requires 24 watts of output power, 40V*600mA. This is a lot of  power, and you would not have seen anything in the lt3463 close to that.

    The two parts you have mentioned seem to be more than capable of providing what you need. If anything, they can do a lot more than what you need.

    Another option to consider could be to use LT8364 to generate the +40V with the typical, page 1 application, modified for 40V. And a;so, the -24V application seen on page  25 can be modified for -40V. 

    The negative converter might be close, but i think it can handle 600mA with a 12V input.

    Side note:

    When both converter are operating at full load, the input source will have to source about 5A.

    Regards.

  • Hi

    thanks a lot for the answer, I'll start over again with these changes in mind and with the advice!

    Sincerely