Voltage ripple when TemperatureEnable[Source] is on

Hi Martin,

Are the spikes in the measurement occurring at 2 second intervals? Temperature measurements on the thermistors are taken every 2 seconds, so there will be a pulse on the TS1 pin at this interval. Is only Cell4 showing this ripple? From the gg.csv file, it looks like you are using a 7 cell configuration, so Cell4 is the top cell of the bottom cell group.

So there will be a thermistor load every 2s and normally TS1 will draw current from the bottom group supply pin (VC5X) - this should pull down on VC5X slightly, but should normally not affect the cell voltage since it has a separate filter resistor. If you look at Figure 13 in the BQ769x0 datasheet - do you have the Rf and Cf components for the VC5X pin installed (typically these are 1k and 10uF)?

Regards,

Matt

Hi Matt,

thank you for your fast response. Yes the ripple is every 2 seconds.

We have a capacitor of 10µF connected from VC5x to Ground. And a connection of the VC5x pin to the "VC5x" label in figure 13, but we don't have the resistor Rf which connects VC5x to Label "A". VC5X is also connected with a diode to VC5 which is connected with a 1k resistor to the cell four

Hi Martin,

Can you try adding the Rf to the board to see if it improves the ripple? 

Thanks,

Matt

I will try to add one

Hi Matt,

I tried to add the resistor, but it did not change anything.

Cheers,
Martin

Hi Martin,

I would have expected the missing Rf to cause the behavior. The missing Rf lets the VC5X supply pull down to supply current through the diode and the resistor for VC5, which will change the voltage across the VC5 resistor. 

Another possibility is the diode could conduct, but a standard diode should not conduct under these conditions. TS1 should pull around 3.3V/10k = 330 uA. Are you using a Schottky diode by chance?

Regards,

Matt 

Hi Matt,

we have just a normal diode with 0.7V.

you mean 3.3V/(2*10k) because of the internal 10k resistor and the ntc?

Hi Martin,

The load current during TS1 measurement will add a current pulse every 2 seconds from the BAT pin through the 3.3V regulator and out the TS1 pin.  At 25C this will be about 3.3V/20k = 165 uA. At hot the thermistor resistance drops and the current would approach the shorted thermistor condition of 3.3V/10k = 330 uA.

Matt

Hi Matt,

thank you for the explanation.

I checked out the schematics of the evaluation module of the BQ769x

https://www.ti.com/lit/ug/slvu925c/slvu925c.pdf?ts=1617021909400&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FBQ76940EVM

there is also no Rf and Cf.

But if Rf is 1k then the voltage drop is about 0.3V so the voltage after Rf is larger than VC5 - Diode voltage so the diode should not conduct.

Hi Matt,

thank you for the explanation.

I checked out the schematics of the evaluation module of the BQ769x

https://www.ti.com/lit/ug/slvu925c/slvu925c.pdf?ts=1617021909400&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FBQ76940EVM

there is also no Rf and Cf.

But if Rf is 1k then the voltage drop is about 0.3V so the voltage after Rf is larger than VC5 - Diode voltage so the diode should not conduct.

I compared again and I've seen that the problem occurs anymore on cell4, but only on cell 7 and cell1. I added a screenshot to the first post, because for some reason I was not able to add it to this post

Thank you Matt, I will test it out as soon as possible

Thank you Matt for your help. With Rf at VC5x and Rf at VBat the results look good.