Tool/software:
Hello,
Im running into some stability issues with the TPS7H3301-SP and I am trying to improve the overall stability during read and write operations of DDR2.
The transients are very fast and seem to cause some "ringing" on the VTT/Vo when there is a transition that happens.
Here is the a scope shot of when there is a data read compared to the VTT output.
This is the current schematic, but I change R4015 from 36.5k to 499 Ohms to help improve the performance after running some simulations.
And here are the following numbers I used to calculate the Resistor value for VTTSNS:
- Ceramic capacitance ESR Values not taken into account
- 300904107106KC ESR = 5.83mΩ @ 100kHz
- Tantalum capacitance ESR Value taken was worst case, not nominal which would be more accurate for response timing
- CWR29FH227KCH Max ESR = 180mΩ @ 100kHz
- CWR29FH227KCH Nom ESR = 64.05mΩ @ 100kHz
- Calculation to used calculate the 36.5k resistor
- Total ESR including Ceramic capacitor: 1.364mΩ
- Time constant is 1.364mΩ * 700uF = 0.954us
- 1nF => 0.954u / 1nF = 954Ω.
- 499Ω provides slightly better performance.
Can you help verify if my calculation is correct? When I use the simulation files, 499Ω shows faster performance values than the previous 36.5k. But it looks like 499 ohms is now causing it to be under damped (Gain margin of 40dB) versus the 36.5k which is over damped (Gain margin of 70dB). Im fixed on the capacitors that I have to use for the design.
Thank you
Albert
