DC2973A / LTC3337 Test for qualification

Max49 my apologies in the delay in responding to your inquiry.  Mentioning MartyM for assistance. 

Hello,

Thank you for all the information. The LTC3337 operates by measuring the voltage from the BAT_IN pin to the BAT_OUT pin to turn on and off the current. When it turns on any impedance on the input such as Battery ESR, meters, wires, connections, add a little impedance and can cause BAT_IN to drop. When the current is turned on it also starts to charge up the BAT_OUT pin. With BAT_IN dropping and BAT_OUT rising the 40mV hysteresis is reached quicker than desired.  

With the 5V supply setup it is difficult to see what the waveforms look like. Measuring with a scope at the BAT_IN pin and the BAT_OUT pin and finding the difference between the two pins will help. These measurements should be as close to the pins on the part as possible. 

On the 15mA IPEAK plot it can be seen that BAT_IN is changing ~50mV as the current is turned on and off. 

The LTC3337 is a hysteretic device so a ripple will be observed on BAT_OUT as the current is switched on and off. A ripple is also observed on BAT_IN because of ESR*I drop. It is suggested for input sources with high ESR to add more input capacitance to reduce the IR drop on the input. Adding current sense meters on the input also adds ESR and can cause a higher than normal ripple on the input. 

Hi marty, 

i have same issue with impedance value, i use demo circuit 2973A configuration is 
Ipeak :5mA
battery: LiSOCl2 (EVE ER26500)
battery cable : ~15cm
supercap output: 2x5F / 2.7V 


i have try to add mlcc (C3) 10uF, 100uF, 220uF and tantalum 470uF. impedance stil not accurate (Vin_at_on always higer than Vin_at_off) and impedance value always negative 



do you have any suggestion?

Thanks 

Hello everyone,

I have the same resultat that Mustafidz 

It's seems that the coulomb counter works great according to my battery (LS14500) and the Load at the output that i have test.
I attached a document where i have monitored accumulated charges and Impedance variation for a long period.

In a first time, i put a 50 ohms resistor in the ouput to discharge my battery at approximately 72 mA and when my battery fall below 2.0V i change my resistor to 350 ohms for a current around than 10 mA.

According to this discharge, the accumalated charges that i got seem to be ok.

But for the impedance, the values are far to be ok according to the battery.

Hello, 

With high ESR sources it is difficult. If adding extra cap to accommodate a more accurate coulomb counter measurement than the impedance measurement will suffer because it slows down the BAT_IN measurement. 

I did find this note below:

"If the ESR*IPEAK is higher than the hysteresis (typically 50mV) the pulses are too short and the A2D might do wrong conversion. In this case it is possible to add a input capacitor that slow down the BAT_IN Pin voltage and make the pulses longer. But because of the slow rising and falling edge of BAT_IN pin the impedance calculation is not anymore correct.
3337 samples the BAT_IN/BAT_OUT voltage in the 3-5us time after the IPK current pulse starts or stops. Therefore if the ESR*CIN time constant exceed this time the impedance is not correct anymore.

To avoid this situation:
- Chose IPK setting so that ESR*IPEAK< 50mV
- We have 2 fuses available that can increase the hysteresis to 100mV, 150mV or 200mV. This feature is not spell out in the datasheet, but can be offered. We need to understand if costumer really need bigger ESR*IPEAK drop."

The suggestion is to find a battery with an impedance low enough so the IPEAK*BSR is less than the hysteresis. I know this is not always possible not a good answer. Or calculate the affect the added capacitor affects the impedance measurement. I do not know a good way to do that at the moment.