LM74930-Q1: PS ORing 50V350A

Hi Stas,

Your understanding is correct that higher number of MOSFETS will take more time to switch. If it's okay for you, you can use it. If you do not want to use RSENSE to measure current, we don't have any controller to measure using RDS_ON. You will have to implement with external circuit. Now if you don't really want current related protection from the controller, you can consider using LM74800-Q1 as well. You can tackle back EMF with it.

Your application is challenging with respect to compact size and high current majorly. From controller POV, I don't see any problem here.

Regards,

Shiven Dhir

I just wonder if it is normal practice to use up to 10 mosfets in paralel, maybe there is a problem doing so, but due to growth of rds_on of temperature, system looks self balancing. LM74800 looks nice, it is drone, i cannot have overcurrent protection, it will fall. My main concern is will this solution work with inductive loads (motors). Main consumer are motors, maybe only 10A to system. 12s battery is like 50V, I consider put 60V mosfet, also LM74800 is up to 65v. I am not really experienced guy in such big energy, is 10v gap enough? Considering on other side there is Li battery ready to eat backemf. Not sure I even will be able to use load dump, because if i select mosfet for 200v i think there will be like 50 in parallel for this current.

Hi Stas, 

10 MOSFETs although seems very high but is achievable. Although you can fulfil your current requirement with a smaller number of FETs also, just need stronger FETs, since you are tight on space, I think you should choose a smaller number of FETs.

I am unsure what you meant by "able to use load dump" "because if I select MOSFET for 200V I think there will be like 50 in parallel for this current."

Do you mean that just be sure, you are considering 200V load dump configuration, if there are any transients? What does he following line mean?

"50 in parallel for this current"

Regards,

Shiven Dhir

Hi!

There is a "load dump" feature to prevent high voltage spikes on "load drop" event.

200V MOSFETs have about 5 times hifger Rdson, so to implement "load dump protection" I would need to use 5 times more MOSFETs.

If I use 10 parallel 60V MOSFETs so I will need to use 50 parallel 200V MOSFETs for load dump.

Also I more curious about will this save me?

I mean I want to control 50V300A with 60V MOSFETs and ORing controller. Will it work?

How should I calculate voltage on FETs?

How should I protect them against spikes?

Is there any manual I can read to understand better how to control inductive loads?

Stas

Hi Stas,

While controlling inductive loads, you will see spikes during turn off of the motor. Datasheet talks about this and recommends adding a schottky at the output to clamp the negative side. To select input protection, you need to find out the level of transient first (if any)

Controlling 50 MOSFETs does not sound idle. Stick to lesser number of FETs and add transient suppressor instead. 

Regards,

Shiven Dhir

But shottky is against negative voltage? Against spikes I would need TVS? Overvoltage will just break shottky, no?

As I understand Schottky D2 is needed in case I turn off power supply while motor running, so inductive load will take current from ground.

And TVS D1 is needed in reverse situation: motor cut off and switch open. It will lead to increase voltage on input and TVS will eat this spike.

Please correct me if I am wrong.

I found out 200+V MOSFETs able to carry 15A, 20 MOSFETs in parallel is ok?

Or better use ORing controllers in parrallel? Like 4xLM's with 5 MOSFET's each?

Can I estimate level of transient before board production?

Like if I have 50V system should I start with 60V MOSFETs or it is better to start with 80V?

Or you suggest measure transient with direct connection from battery to motor?

I expect it will dramatically change due to Rdson added in net, because wire connection is like microohms and right now I expect to see no transient with battery direct connection.

I see some numbers of transient for 12/24V. Are there the same numbers for 48V or something like this?

Hi Stas,

Yes, overvoltage can break schottky. For that you can set OVP feature in the device.

TVS D1 is needed to clamp negative spike within the abs-max rating of the controller. 

Further, I will let our high-power expert comment on this.

Regards,

Shiven Dhir

Hi Stas,

I don't think you really need reverse current blocking feature here, so you need not consider an Ideal diode. Please consider TPS4811-Q1 once. You can start with 85V FETs and with time, improvements can be done, 85V should be a good starting point.  

TPS4811-Q1 data sheet, product information and support | TI.com

Regards,

Shiven Dhir

Hi!

Thanks for estimation.

As I understand LM7480 can work as Ideal diode and Power switch according to pin setup.

TPS4811 as I understand can act only as power switch.

I can setup one PS with higher voltage to be power switch

And second PS with lower voltage to be as diode in case of temporary peak current and so voltage drop on PS1.

One PS is Li Battery and second PS is hydrogen fuel cell, so it obviously some alternator inside and that is  why I ask about load dump.

85V is ok, but common transistor have grade 80V and 100V, I would start with 80V if you think it can succeed.

Thanks

Stas

Hello Shiven and Stas,
I am not sure if it is appropriate to jump into your post, but i have a very similar application and would also like to determine the best solution. Shiven, please let me know if not appropriate to be here and i can create a new topic.

Application is for a mobile robot power distribution board with the following components:

• 2 x 2kW 48V batteries in parallel
• 250 Amp continuous operation
• inductive loads (6 ESC and Motors)
• most likely using IAUT300N10S5N015 with a gate charge of 216 nC each

I have considered both the LM74930-Q1 and  the TPS4811-Q1 for this application. My consideration of the LM74930-Q1 is that i would like to consider using it in a dual ideal diode ORing configuration as in this image. Both batteries would be at the same voltage so that i could hot-swap one battery and therefore remove the requirement to stop robot for charging.

The following App Note was also inspiration for the design 

https://www.ti.com/lit/an/sluaan7/sluaan7.pdf?ts=1715647314737 

Requirements:
Primary Battery disconnect:

• Max 60VDC operating voltage
• <10us short-circuit protection (bidirectional)
• common source bidirectional switch support
• Bidirectional current monitoring
• Adjustable current limit (overcurrent protection)
• Adjustable overvoltage protection
• reverse polarity protection
• short circuit protection
• Pre-charge circuit

Load (High-Side) Switching:

• 4 x 50 Amp Motors (TPS4811-Q1)
• 2 x 20 Amp motors(TPS4811-Q1)

Can you please provide some insights into the benefits of using the TPS4811-Q1 over the LM74930-Q1 for my application?

Thanks for the prompt reply Rakesh, I will create a new thread for further discussions.

Thanks Zoran

Your task looks familiar to mine, I will also subscribe to  your net.

Stas