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.
This Schematic Review request is in connection with suggestions of parts made in below thread
I have completed my schematic and looking for your valuable review if there are any improvements to be made to reach better noise floor.
Dear Stephen Ziel , Steve Schnier kindly look in to the attached schematic and provide your valuable review .
Thanks
Shyam D.
Hi Shyam,
I can speak to the LDOs and I'll let Steve Schnier speak to the TPS62912RPUR design.
If your load current on the TPS7A94 LDOs is constant (DC) then you may wish to increase Vin slightly. The 2.5V output has 160mV of headroom but this assumes perfect tolerance on Vin and Vout. To account for tolerance and any temperature variations in the PCB you might want some more voltage. For example, if we assume Vin and Vout have 1% tolerance, then you will want Vin to have a typical voltage of 2.75V to prevent the LDO from entering dropout on the 2.5V rail. The 5V rail will want Vin to be about 5.25V to prevent the LDO from entering dropout on the 5V rail. The input voltage can be readjusted further depending on the actual tolerance of Vin and Vout (I used 1% as an example).
If you anticipate high speed load transients on the output of these supplies you may want to increase Vin so that it is 500mV above Vout. If you don't have high speed load transients then you can ignore this suggestion.
For the 2.5V output:
The NR/SS resistor should be 16.5k resistor in series with 165 ohms to get 2.5V (so replace R126 with 2 series resistors). You probably only need 1 NR/SS capacitor, not both 4.7uF but you can leave both installed if you wish. R127 should be 12.4k and R123 should be 133k to give you a PG threshold of 95%.
For the 5V output:
The NR/SS resistor should be 33.2k resistor in series with 133 ohms to get 5V (so replace R137 with 2 series resistors). R138 should be 12.4k and R135 should be 280k to give you a PG threshold of 95%.
Negative LDOs:
I have the same feedback on Vin and Vout. If the tolerances are +/- 1% for both Vin and Vout, you may wish to increase Vin slightly so you do not risk entering dropout. The -2.5V output looks good otherwise. The -5V LDO needs updated setpoint resistors: you can use 3.32k and 1.02k or another set of your choice to get -5V.
Thanks,
Stephen
Shyam,
For TPS62912 positive outputs schematics
1, L2/L4 need follow recommendation table 8-3 in datasheet
2, A CFF could be reserved and put it parallel with upper feedback resistor, NC as default.
For inverting design
1, A CFF also could be reserved, parallel with upper feedback resistor R141/146, NC as default.
2, Ferrite bead selection follow table 8-3 in datasheet
3, Pin1/4/5/7/10 should refer to Vout before ferrite bead, please refer to TI EVM to change.
TPS62913 Inverting Buck Boost EVM User Guide
Yuchang
Dear Steve Schnier can you also please drop your comments, so that i can come back with final schematic before starting layout.
Stephen Ziel The power supplies rails are powering the high speed(500Mhz) TIA like OPA699, so your recommendation of increasing input i shall follow, may i keep 2.85V and 5.35 on both Positive and negative rails ?
1. I have considered L2/L4 from datasheet only
2. Cff reservation also incorporated
3. How to terminate PG for Inverting design ? a 10K reistors to PREFilter Output ?, the reference design has some FETs to make it compatible for positive looking monitoring signals, i am not going to use PG for any of monitoring, but would just while debugging if PG is built or not, for which i will pull it with a 10k to prefilter vout ? is this correct or wrong.
Shyam,
If you don't use external signal to control EN, tired it to VIN is ok
If PG not used for monitoring, leave it floating.
Others looks well now.
BTW, recommend to follow TPS62913 Inverting Buck Boost EVM User Guide for layout design.
Yuchang
Hi Shyam,
With 350mV of headroom on the positive rail you'll still see very good noise performance. See figure 7 in this app note:
https://www.ti.com/lit/ml/sbva099/sbva099.pdf
I would expect this to work for you.
Thanks,
Stephen
Hi Shyam,
Sorry, I missed the negative LDO review. Based on figure 19 in the TPS7A33 datasheet, I think you'll also have good performance with 350mV of headroom on the negative rails as well.
Thanks,
Stephen
Yuchang Zhang are you sure for inverting mode isnt the Enable pin to be tied to Vout with a 10K ? you meant to say we can it to VIN as like done in normal mode ?
Shyam Daram VIN tied to EN together should be ok for your case, because your Vout=2.58v is small and under VIN UVLO threshold, it still can be powered off normally when VIN drop, if VOUT is higher than UVLO, VIN connected to EN together may cause power off issue, since VIN always high than UVLO and Enable is active as well.
Another concern for VIN tired to EN is the converter will work at lower VIN as long as VIN+VOUT>Vuvlo.
If you think both above impact could be acceptable, VIN and EN tired together directly should be ok.
If not, you can follow TPS62913 Inverting Buck Boost EVM User Guide add level shifter to control enable.
Can you please suggest proper Feed back resistor network R1 and R2 for LDOs both positivie and negative
i tried using Webench, but its not allowing my input voltage and output voltages, throws error saying limit issues.
also i am confused in selecting Vref, you kind help is appreciated.
1. Negative 5V with input -5.35v, i have considered using 332k and 102k for R1, R2 resp.
2. Negative 2.5V with input -2.85v , i have considered using 154k and 133k for R1, R2 resp.
Hi,Yuchang Zhang and Stephen Ziel
Attached is my final power supply portion i hope i am not missing anything psu_portion.pdf
Shyam Daram for 62912, looks ok now.
And by the way, recommend to sperate P/N with different thread, in case missing your question, due to TI has different product line for different production support.
Yuchang
@Stephen Ziel, please support review LDO question.
Shyam Daram Kindly remind please check current limitation could cover your system requirement for -2.85V and -5.35V based on formular 1-8 in TPS62913 Inverting Buck Boost EVM User Guide,
Yuchang
Yuchang Zhang can you elaborate what you meant to say ? below is my requirement chart
Polarity | Voltage | Load Current | |
+ | 2.5 | 400 | mA |
- | 2.5 | 400 | mA |
+ | 5 | 620 | mA |
- | 5 | 330 | mA |
Yuchang Zhang and Stephen Ziel all these attempts were made to select low noise dcd and ldo is to make sure my circuits are less in noise, can you please let me know how to handle the ground now, in my schematic i maintain a analog ground for all opamps, how to seperate it from power ground, can you please kindly suggest good grounding scheme for my opamp circuits, if you can drop me a msg with your mail i wish to share my full schematic to you.
Shyam, for positive Vout, Iout is average of inductor current but for negative Vout, Iout is not average value of inductor current, so we need check the max current ability based on IL, Ipp....which shows details in the IBB reference design document.
For your voltage from the sch, seems -2.5V should be -2.58V and -5V should be -5.35V, right? current should be ok and can be supported by 62912.
Just one more concern is if use 12V to -5.35V, it is too margin, since the VIN/EN rating will be up to 12-(-5.35)=17.35V, and if considering tolerance of VIN and VOUT, it will be worse. Is there possible to use +5V instead 12V for transferring to -5V?
Yuchang
oh... is this happening due to EN being tied to VIN ? i am not crossing VIN limits though
If EN is the pin which is causing issue, why cant i use a potential divider to half to enable EN ?
from your comments i feel only EN is the issue now and 62912 can meet the current requriements, or you want me to replace it with 62913 ? as they are pin compatible.
Shyam,
no, it is not related to EN tied to VIN or not, and VIN/EN has same voltage rating ability, because the reference level of IC is -5.35VOUT not 0V Gound now, so VIN and GND need to stand VIN-(-VOUT) =12+5.35V., If VIN could be changed to 5V, VIN and EN just stand 5+5=10V rating. enough margin from 17V recommendation value.
current ability now is ok, the concern is voltage rating, and 62912 has same rating as 62913 for both VIN and EN.
Yuchang
Yuchang ZhangThis will reduce efficiency and increase board heat if i add one more dcdc, as suggested in original thread Ultra Low Noise Power Chain for Photo Sensitive applications - Power management forum - Power management - TI E2E support forums,
i will go with using TPS63710 , but it doesnt have a output filter approach, can i replace TPS63710 in current design of 62912 ? i shall keep the LC filter in line same and replace teh 62912 with TPS63710
not pin to pin but deisgn wise TPS63710 followed by CLC filter approach and then followed by LDO
Let me know feasiblity please
Shyam, seems 63710 is ok for your application, for LC filter, suggest you submit a new thread with TPS63710 for the product line engineer to review.
Yuchang