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.

TPS65145: DLP4500 voltages from 3.3V rail

Part Number: TPS65145
Other Parts Discussed in Thread: DLP4500

Hello,

I am trying to design a circuit that provides the Voffset, Vbias and Vreset voltages of 8.50, 16.0 and -10.0 for the DLP4500 DMD using the TPS65145. I am a bit confused about the negative charge pump. In the datasheet I see that the most negative voltage that can be generated is

VOUTmin = –(VO1 – 2 VF – IO (2 × Rds(on)Q8 + 2 × Rds(on)Q9 + Xcfly))

Note that there is no reference to the Vin voltage.

This formula suggests that the voltage magnitude of the negative charge pump will always be smaller than that of the boost converter. This implies that you cannot generate 8.5V from the boost converter and -10V from the negative charge pump.

However, looking at the schematic of the Lightcrafter 4500, (tidr157a) page 24, it is shown that the TPS65145 is in fact used to generate these voltages. However, there is a connection from the SW nodes to the charge pump part of the schematic. It seems some clever engineering is used to generate a larger (more negative) voltage with the charge pump). I don't understand how the circuit from the reference design works. I would like to use the same technique, however in my board the Vin voltage is 3.3 instead of 5V. Therefore, I am not sure if I can use the same circuit from a 3.3V rail. I would like to avoid adding a 5V rail to my circuit as all other components use the 3.3 rail.

So, my questions are:

- What is the name for the circuit that is used in the Lightcrafter 4500 schematic?

- Is is possible to use this circuit when Vin in 3.3V instead of 5V?

Looking forward to your reply,

Jim

  • Hi Jim,

    I have notified our expert regarding this topic. Please expect a response by 5/26/20. 

    Thanks,

    Costin

  • Great, thank you for the update.

  • Hello Jim,

    “It seems some clever engineering is used to generate a larger (more negative) voltage with the charge pump). I don't understand how the circuit from the reference design works.”

    Ahh... I see you’ve discovered the confounding, double-secret charge pump circuit... It is designed to intimidate and confound!  (Insert evil laughter here.)

    At least, I thought it was intimidating.  lol….

    I had to look into a very similar circuit quite recently for a customer issue and found it a bit confusing.  I am not a power supply expert; I am a DLP® / DMD guy!  This charge pump made my head hurt! 

    I worked through it by doing some research, talking to one of the TPS experts, and measuring each node of our circuit with an oscilloscope to see what was really going on.  Oh, and coffee... coffee definitely helped.

    All joking aside....  

    What we have implemented to create Vreset at -10V is a two-stage charge pump with an unregulated stage (voltage inverter) creating approximately -8.5V from Voffset’s switching node, and a ~-1.5V stage created by the TPS65145’s negative charge pump. The final output is regulated by the TPS's negative charge pump feedback circuit.

    NOTE: I am ignoring the diode drops, etc.; the voltages are approximate ("~")!  You can see example calculations in the E2E blog mentioned below.

    The portion of the circuit that is AC coupled from the Voffset boost converter switching node to the Vreset output is an unregulated inverting charge pump driven by the Voffset boost regulator switching node (C220 “Flying Capacitor” + DN2 Rectifier Diodes). 

    At power on, even before the TPS charge pump is enabled, it will create approximately negative Voffset (~ -8.5V) at the Vrst (“Vreset”) node (C230) (ignoring diode drops, etc.). 

    An example of a boost converter driving an inverting charge pump from "Pump it up with charge pumps – part 3" :


    Figure 4: Boost converter driving an inverting charge pump

    The other stage is created by the TPS65145 negative charge pump through Flying capacitor C227 and diodes DN1 (and the TPS internal negative charge pump).  When the TPS negative charge pump is enabled, it will drive drive this stage to approximately -1.5V DC @ C228 (again, ignoring diode drops, etc.).  

    ~-1.5V added to the ~-8.5V creates -10V at Vrst and is regulated by the negative charge pump’s feedback circuit. 

    Figure 2 on this TI E2E Blog shows an example of a regulated and unregulated charge pump combined to create a higher output voltage:

    Pump it up with charge pumps – Part 4

    What we are doing is the same as what is shown here, only with a negative charge pump -- so the four diodes are reversed.

    The whole “Pump it Up” series might be useful: https://e2e.ti.com/tags/Pump%2bit%2bup%2bseries

    Dave over at EEV Blog has a pretty good tutorial on a Voltage Inverter that might be helpful as well: https://www.youtube.com/watch?v=I4ED_8cuVTU.


    Regarding the question of whether you could do this with a 3.3V Vin instead of 5.0V.... Yes, you can.  We have done this on other designs.  The DLP470TE EVM is an example.  That design uses the same TPS65145 circuit design.  In that case the input is 3.3V; the feedback resistors set for different DMD voltages; and there is a modification for power sequencing (not related to Vin = 3.3V).


    I encourage you to checkout any hardware you build to be sure there is no overshoot, etc. on the supplies.  The power supplies can be pretty sensitive to vias in the feedback circuits, different inductors and capacitors, etc..  Also, please be sure to review the TPS65145 layout guidelines as you design your board.

    Hopefully that addresses your questions.

    Good luck! 

    Best Regards,
    Gary

    PS:  Thanks for giving me the opportunity to describe this circuit.  It has been cathartic.  

    All information in this correspondence and in any related correspondence is provided “AS IS” and “with all faults”, and is subject to TI’s Important Notice (http://www.ti.com/corp/docs/legal/important-notice.shtml).

  • Hi Gary,

    Thank you so much for a clear and very interesting answer! It seems I have to stock up on coffee and do some studying ;).

    Kind regards,

    Jim