AUIR324x for 18V (cut-off = 12.5V, max full charge = 21V) battery switching

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
lock attach
Attachments are accessible only for community members.
xavier1111
Level 1
Level 1
5 replies posted 10 sign-ins 5 sign-ins

Hi community, hi Infineon support,

we are planning to use the AUIR324x IC gate driver for a battery switching architecture with current constraints of

15A continous, 20A for 2mns, 30A for 10s. We have been testing the "AUIR3241S/AUIR3242S - Board Back to Back 12V" with the batteries and it seems to switch correctly between the 2 batteries.

However at this time we have 2 questions :

- technical : attached TEK0000 (GPIO, Load - resistive electronic 150W load - voltage switching from BAT1=13V to BAT2=20V in ~30us) we can see a first intermediate voltage level for 10us then a small peak at 25V before reaching 20V. Could you explain the signal ?

- commercial : we don't see any stock for AUIR324x IC chip and we want to design a PCB around September. What is the expected planning for procurement of this chip ?

Xavier.

 

0 Likes
1 Solution

Hello Xavier, 

I would like to update you on the simulation results. 
I could recreate the measurement results on the simulation test bench and the analysis as follows

Test bench

Vishnu_Nambrath_0-1659443075678.png

  • Load 10 Ohm
  • Delay between the GPIO signals = - 100ns (Why it is negative? First GPIO is turned off after 100ns from turn on of second GPIO. This create an overlap in the GPIO signals as well)
  • First battery 12V and second battery 20V

Waveform 

Vishnu_Nambrath_1-1659443406011.png

  • GPIO signals are in the 3rd graph and curser shows the -100ns delay
  • Gate source voltage of both eval board is shown in the second graph and it shows an overlap here as well
  • During this overlapping period, both eval board MOSFETs are on and creating a cross current between 12V battery and 20V battery
  • This cross current is the reason behind the strange load voltage

Solution : Increase the delay between the turn on an off of the eval boards.
The below simulation shows with a positive delay of 2us between the GPIO signal

Vishnu_Nambrath_2-1659444016950.png

  • No cross current 
  • But there is certain period where load is not supplied by any of the batteries. 

Are you okay to have this dip in the load voltage? if not you can have a capacitor parallel to the load which can supply the load current during the transition. 

Max delays involved with the gate driver is also available in the datasheet as shown below. 
This will help you to optimize the delay between the GPIO signals in your application. 

Vishnu_Nambrath_3-1659444372924.png

Vishnu_Nambrath_4-1659444497831.png

Hope it is clear to you. If not please feel free to ask. 

Regards 

Vishnu N

 

 

 

 

 



View solution in original post

0 Likes
9 Replies
Vishnu_Nambrath
Moderator
Moderator
Moderator
50 solutions authored 100 replies posted 5 likes given

Hello Xavier, 

Thanks for posting into our community. 

Technical: You have switched from 13V to 20V in 30us and electronic load is also involved. Can you please share the schematics/block diagram for the circuit for better understanding? Where are the signals probed? uC GPIO and source pin of the power MOSFET right?

Commercial: Procurement has to be discussed with your distributors.  
AUIR3241/ AUIR3242 are active and preferred parts.  If you are looking for any other parts, you can check the product status on our product page. 

 

Regards 
Vishnu N

 

 

0 Likes
xavier1111
Level 1
Level 1
5 replies posted 10 sign-ins 5 sign-ins

Hi Vishnu,

Thanks for your prompt reply. Below my follow-ups


Technical : please find attached a bloc diagram extract where you can understand that BAT1(13V) and BAT2(20V) are connected to each AUIR324x board, GPIO controls for each board come directly from a uC output 3.3V logic, The GPIO are switched on/off in opposite at the same moment (from a clock cycle point of view at the uC 100ns difference between the switch of each GPIO), the load is connected to the PLUG2/5 pins. The load is an electronic 150W resistive load (like this one https://www.diymore.cc/products/150w-ldm-digital-electronic-load-discharge-capacity-tester-usb-voltm...) allowing up to 7-10A max. The signals probes are GPIO signal for one battery and the other signal is the LOAD voltage signal (basically on MH33).

Commercial : I am glad to know this an active and preferred part and I will follow-up with the local distributor in our country.

Regards, Xavier.

0 Likes

Hi Vishnu,

technical : did you get my last message ?

commercial : I am now in contact with local dealers, they announce 56weeks lead-time for AUIR324x parts - I hope that I could get samples from Infineon or dealers and shorten this LT.

Best, Xavier.

 

 

0 Likes

Hello Xavier, 

Thanks for sharing the block diagram. I understood your connections. 

Can you try using a high wattage resistor load that draws 7A-10A?  I am suspecting the spike near the end of transition in the load voltage is due to electronic load. 

Regarding the flat region at the load voltage, I am trying in the simulation. 
I shall share my findings soon. 

Regards 
Vishnu N

0 Likes

Hello Xavier, 

I would like to update you on the simulation results. 
I could recreate the measurement results on the simulation test bench and the analysis as follows

Test bench

Vishnu_Nambrath_0-1659443075678.png

  • Load 10 Ohm
  • Delay between the GPIO signals = - 100ns (Why it is negative? First GPIO is turned off after 100ns from turn on of second GPIO. This create an overlap in the GPIO signals as well)
  • First battery 12V and second battery 20V

Waveform 

Vishnu_Nambrath_1-1659443406011.png

  • GPIO signals are in the 3rd graph and curser shows the -100ns delay
  • Gate source voltage of both eval board is shown in the second graph and it shows an overlap here as well
  • During this overlapping period, both eval board MOSFETs are on and creating a cross current between 12V battery and 20V battery
  • This cross current is the reason behind the strange load voltage

Solution : Increase the delay between the turn on an off of the eval boards.
The below simulation shows with a positive delay of 2us between the GPIO signal

Vishnu_Nambrath_2-1659444016950.png

  • No cross current 
  • But there is certain period where load is not supplied by any of the batteries. 

Are you okay to have this dip in the load voltage? if not you can have a capacitor parallel to the load which can supply the load current during the transition. 

Max delays involved with the gate driver is also available in the datasheet as shown below. 
This will help you to optimize the delay between the GPIO signals in your application. 

Vishnu_Nambrath_3-1659444372924.png

Vishnu_Nambrath_4-1659444497831.png

Hope it is clear to you. If not please feel free to ask. 

Regards 

Vishnu N

 

 

 

 

 



0 Likes

Dear Vishnu,
(sorry for my delay to reply) thanks for your precise and clear simulation report.

1.Is there a way to run this simulation on my side and edit some settings (Infineon online simulation tool?) my self ? Do you have a file to share ?
2.Concerning the idea of cutting-off the first eval board and then starting after a 2us, I need to test it with the actual boards to see the drop in terms of output voltage because we need a continuous voltage. Maybe a LC filter on the output (considering the load as an R-load) would prevent drop and smooth the output voltage. We then need to calculate the right dv/dt to estimate the values of L and C with this case.
What is your feeling about an LC filter on the output ? Can we add that to the simulation ?

Much regards,
Xavier.

0 Likes

Hello Xavier, 

I will send you the SIMetrix design file via chat. 

Why LC filter? Simply a capacitor can buffer the load current during the battery transition period, right?  I have added a capacitor also to the test bench. 
The concern here is, You need to take care to minimize the oscillations in the test bench and inrush current to the capacitor. It should not cross the absolute maximum ratings of the devices. 
One more drawback to the design is, that the load will be supplied even after switching off the MOSFETs until the discharge of the output capacitor. 

Regards 
Vishnu N


0 Likes

Hello Vishnu,

1.Thanks for your SIMetrix design file, I will give try in a free demo version of SIMextrix

2.For the LC filter, during our tests we used a R-type motor load and when switching from 13V to 20V we observed a glitch on the motor load which is due to the sudden switch from 13V to 20V. Therefore to smooth the transition we are thinking about adding a LC filter to control the dv/dt. Do you have another proposition to smooth the transition between V1=13V to V2=20V ?

Regards, Xavier.

0 Likes

Hello Xavier, 

Please find the attached schematic file. You can download the models from the Infineon website and install them in the SIMetrix before running the simulation. 

Coming to your second question, LC also should be ok. 

Regards 
Vishnu N

 

0 Likes