0
\$\begingroup\$

I want to run a 2G Modem HL6528RD. I provide it with 3.6 V, got from the following circuit :

enter image description here

I finally manage to make it work but the problem is that if R108 is mounted, the modem doesn’t work, (the modem's power supply drops by 1V due to current peaks of up to 2A and causes the modem to restart).

However, if I replace R108 by an inductor with high enough value (2.2uH for exemple), the voltage hardly changes when current peaks occur.

Could someone explain to me, what is causing this phenomenon?

Thanks for reading my question

\$\endgroup\$
  • 1
    \$\begingroup\$ R108 is supposed to be zero ohms...just a jumper. What value did you use? \$\endgroup\$ – Elliot Alderson Jan 28 at 16:56
  • 1
    \$\begingroup\$ Did you design your own PCB for the modem, or are you using a purchased module? \$\endgroup\$ – JRE Jan 28 at 19:09
  • \$\begingroup\$ R108 is a zero ohm resistance and musn't have any influence on the circuit ... Yes I designed my own PCB for the modem \$\endgroup\$ – A.David Jan 29 at 8:21
2
\$\begingroup\$

The ESR of input and output Caps is important since a step load can have a rise time shorter than the bandwidth, BW of any high speed linear loop. 0.35/BW = 10~90% rise time.

Consider a 0.1A/us as slow and 1A/us and >1A/us as faster than what the loop can regulate without help from a low ESR cap. This controls the step load voltage drop Istep*ESR. Thus for 2A and say 200 mV error you need an effective series resistance, ESR of 100 mΩ max at all times.

Decoupling the load from the LDO also has an impact since the value of a step load drop in voltage becomes ΔV=L dI/dt Thus you can estimate your dI/dt from the effective ΔV and L you chose.

These ESR impedance ratios and their slew rates impact the Step response directly.

In 8.2.2.4 "The TPS7A7300 is evaluated using an X5R-type, 10-μF ceramic capacitor. TI highly recommends the X5R and X7R type capacitors because they have minimal variation in value and ESR over temperature. Maximum ESR must be less than 1 Ω."

Your problem appears to be excessive ESR in caps.

\$\endgroup\$
  • \$\begingroup\$ All capacitors are X5R-type ceramic capacitors, so the ESR are insignificant. C92 reference is T520D337M006ATE015, it has a 15mΩ ESR \$\endgroup\$ – A.David Jan 29 at 8:41
  • \$\begingroup\$ If you have a voltage drop, it can be explained by conduction or induction impedance ratios. So if you supply accurate photos of signal, layout and BOM, a more precise answer can be given \$\endgroup\$ – Sunnyskyguy EE75 Jan 29 at 14:15
  • \$\begingroup\$ For example the SNS loop vs load loop and how well it senses load V and why the load caps do not support the transient. wire Inductance? \$\endgroup\$ – Sunnyskyguy EE75 Jan 29 at 14:31

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.