BQ34Z100-G1: Smoking part and all connected parts.

Ed Thompson

Part Number: BQ34Z100-G1

I have inherited this design. This design has been in use for a few years with no issues that I have heard about. We got a new batch of Lithium-ion batteries and new battery monitor boards. The BQ34Z100-G1 on our monitor board. In the datasheet section 9 we have a 10K for R1, D1 is a BZT52C5V6-TP and Q1 is 2N7002ET1G to get REGIN voltage. Our nominal battery voltage is 20.1V to 17V. Our voltage divider for BAT is 748K over a 16K to GND. Our sense resistor is 0.0008 Ohms. I don't have the part number for the thermistor but it is a passive part averaging 9K to 11K at my room temperature.

During heavy sudden loading we get up to 65 amps spikes for about 50 ms with a peak lasting about 5 ms. We see 7+ amps of current spikes on top of everything that last around 5 ns, these are not continuous. We are seeing a voltage spike of about 1.5 volts lasting for about 10us. We aso see occasional 6 volt voltage spike lasting ~40 ns.

We are seeing the BQ34Z100-G1 burn leaving pits near pins 5 and 6, sometimes these pins melt off. D1 and Q1 from above also burn at the same time. I have not caught anything out of the above description happening when they burn.

I have been working on this problem for weeks and am at a dead end. Do you have any suggestions?

10 months ago

0 Ed Thompson 10 months ago

Prodigy 20 points

This happens on the old batteries with the new monitor board. So it is not the batteries that is the issue.

0 Diego 10 months ago in reply to Ed Thompson

TI__Mastermind 18830 points

Hi,

Has anything in your system changed? Any components changed?

What is the voltage on REGIN during these spikes?

Is it possible for you to share your design schematic?

Regards,

Evan

0 Ed Thompson 10 months ago in reply to Diego

Prodigy 20 points

R16 is now a 0.001 ohm resistor.

We did go to a new board house. No design changes. I will take a new capture tomorrow. From memory I did not see and changes until we got the cascade up to the battery voltage.

0 Diego 10 months ago in reply to Ed Thompson

TI__Mastermind 18830 points

Hi,

How many Watts is the 10k resistor rated for.

Is it possible to swap the Zenner-diode circuit with an LDO?

Do D1 and Q1 comply with this e2e (+) BQ34110: Relates to bq34110 EVM Schematic - Power management forum - Power management - TI E2E support forums

Regards,

Evan

0 Ed Thompson 10 months ago in reply to Diego

Prodigy 20 points

The resistor is a 1/16 W. No visible damage on burned board and still at 10K when isolated.

LDO may be possible in a redesign. As is our production is stopped and shipments are slipping. We hope that we caught all faulty systems before they shipped. This product and design has been shipping for years.

I am waiting for the battery enclosures to be sealed to get the previous requested information. Here is what everything looks like when we just have the battery turned on and only supporting the communication interface.

Green is the voltage at REGIN. It has an average of 3.69V with a pulse of 126 mV.

Yellow is battery power, sitting at 20V.

Blue is across R16, one of the sense resistors.

0 Diego 10 months ago in reply to Ed Thompson

TI__Mastermind 18830 points

Hi,

1/16W resistor with 7A spikes does not sound sufficient.

I recommend adding an LDO if you cannot determine what is causing the problem.

Regards,

Evan

0 Ed Thompson 10 months ago in reply to Diego

Prodigy 20 points

The 7A spikes are noise at sense resistor we have power spikes of about 40A regularly. With the 0.0008 ohm sense resistor we stay under 125 mV up to 154 A. Why would this cause problems with the the current limiting resistor for the fet supplying power to the chip? What wattage for the current limiting resistor do you suggest? Q1 is rated to 310 mA steady state and 1.2A pulsed. Are you suggesting that Q1 is failing first?

0 Adrian 10 months ago in reply to Ed Thompson

TI__Mastermind 33721 points

Hey Ed,

Evan will get back to you by tomorrow.

Regards,

Nick

0 Ed Thompson 10 months ago in reply to Adrian

Prodigy 20 points

Sounds good. I should have the new captures by then.

0 Diego 10 months ago in reply to Ed Thompson

TI__Mastermind 18830 points

Hi Ed,

I suspect either Q1 or D1 is failing.

With the circuit you currently have what voltage is on REGIN?

Regards,

Evan

0 Ed Thompson 10 months ago in reply to Diego

Prodigy 20 points

Green is the voltage at REGIN. It has an average of 3.69V with a pulse of 126 mV. As shown with my post with the scope capture.

0 Diego 10 months ago in reply to Ed Thompson

TI__Mastermind 18830 points

Hi Ed,

When the part melted at the REGIN Pin, this occurred with 3.69V and 126mV on REGIN. This is inside the specs for the part and should not cause the part to melt.

Regards,

Evan