TPS25982: Issues encountered during startup and restart.

Hi Guanming,

I apologies for missing your previous reply. Let's continue our discussion here. 

I am first replying to old questions and then new ones. If the old queries are resolved, please ignore them.

1. After adjusting the output resistance, the startup will not reach the upper limit of protection in this case. So, what does the resistor here do? It seems that this resistor is necessary.

A: Our eFuse during steady state is like a resistor (pass FET is in linear region). So whatever resistive load you apply, it will draw some constant current. If the load is capacitive, it will draw inrush current and then will stop drawing current unless there is some dvdt.  

2. If we want to simulate the surge part, how should we set the environment? What modifications do we need to make to TPS259827LEVM?

A: I don't fully understand what you mean by it. If you want to test the surges, you will have surge profile so you can apply it to the device. I think there was some confusion regarding the device operation or my understanding of your question.

3.In this case 2, if the electronic load is switched to CC mode and then the device is turned on, the startup will still fail, and the waveform is as shown in the response below.
So, can I think of CC mode as a large capacitor? When the capacitance is large and there is no RL, is it easy to trigger the thermal protection?
But I also adjusted Rlstart, but it didn't achieve the effect I wanted. What could be the problem? Has the value of Cout reached the upper limit? The maximum value of Cout is in SPEC. Is there any place I can refer to?

4.Is there any Inrush test data for TPS25982 that I can refer to?

A: Let me share some resources for Q3 and Q4.

Datasheet section 9.2.2.5 helps in setting the dvdt value. It takes both example (only capacitive load and both cap and res load).
Device in case of high current goes into current limiting mode and in this mode, you will have power dissipation at device of (Vin - Vout)*ILIM. With this power dissipation, device in some time will go to thermal shutdown. This keeps the MOSFET in SOA. 

In case of high cap load, inrush current and startup time will determine if the device goes to thermal shutdown or not. Our design calculator (as previously shared) will help in checking if the device can perform glitch free startup or not. 

External_Solving customer issues by discovering unnoticed eFuse application design and evaluation facts.pptx

Please refer to slide number 6 onwards for startup and 17 onwards for CC mode vs CR mode in e-load

guanming ji said:

Q1:

We left the electronic load on all the time and startup we saw the current rise to 12A and then stay level.
How will this IC start up when there is a heavy load at startup?
Will dvdt be effective when started in this mode?

Yes, dvdt will be effective in case of high capacitive load up to certain limit. dvdt won't help for current drawn by resistive load.

guanming ji said:

Q2:

Use electronic load mode: cc to simulate the backend load. Is this testing method correct? I would like to ask about the correct way to test this ic, or how to verify this ic.

As previously mentioned, CC mode is not recommended. Please go through slide attached above. It will also explain the issues with the CC mode and also tell the correct way to test the device. 

guanming ji said:

Q3:

Same as the continuation problem of Q1, if you perform a short circuit after starting, you will find that it cannot automatically restart. In this case, it may be that the PD consumes too much power and triggers the SOA protection when restarting?
At this time, VIN: 20V, the back-end electronic load is continuously turned on, and 12A is set.

Need more clarification. Are you performing the output short after the startup is done or during the startup? Are you keeping the short for long time and is device trying retry when short is still applied? 

guanming ji said:

Q4:

If the back-end load is very large, is it suitable to use an IC like TPS25982 to design?
If the back-end load of the design circuit is large, are there any recommended ICs that you can refer to?

Can you tell me some value. which load and how much are you applying?

Best Regards,
Arush

Hi，Arush Gupta

Thank you very much for your reply and for the report on the correct usage of efuse.

Let me ask some more questions.

TPS25982 Questions and Supplements to Questions.pdf

Finally let me thank you again for your reply and look forward to hearing from you.

Hi Guanming,

TPS25982 Questions and Supplements to Questions_7may.pdf

Best Regards,
Arush

Hi，Arush

Thank you very much for your reply.

Based on your reply, I have a summary of DVDT. Can you help me confirm if it is correct?
1.dvdt is invalid during Current Limited Start-up.
2.dvdt should be set according to the actual load. In CR and CC Mode, I may not be able to verify this function.

2.

Based on Q5, let me rephrase that and ask again.
If CC and CR modes are used to simulate back-end load, can this excel be used to calculate and verify?

Finally let me thank you again for your reply.

Hi Guanming,

Can you confirm your operating conditions (12V, 14A continuous) and schematic. I can verify it and suggest any changes if required. 

Yes, during current limited startup, dvdt won't control the slew rate. Current limited startup will only happen when the current is more than the limit or device is retrying from the fault. 

What is your actual load?

Best Regards,
Arush

Hi，Arush

Based on your slides, I modified the test method of TPS25982.
I first used CR Mode to test, and everything was normal until the Auto-retry function discovered several problems during testing.

TPS25982 Auto-retry Questions .pdf

Our design is actually about 2000uF, Vin 12V, Iout 8A, I have actually connected it to my design and verified the functionality and it works great.

Some functions are not currently used by design. For example: Auto-retry function
But we will still test this feature to prevent anything we didn't notice if we need to use it in the future.

Let me end by thanking you again for your reply.

Hi Guanming,

I am out of office today. I will get back to you on Tuesday. Thank you for understanding.

Best Regards,
Arush

Hi，Arush

If you need to add anything, please let me know.
I feel very strange about this problem of not being able to restart.
Maybe there is something wrong with my operation and understanding.
Or are there any possibilities that I can try?
Thanks.

Hi Guanming,

Thank you for your patience.

Can you tell me what are the values of dvdt cap which you used. Increasing the dvdt cap makes the start slower which can help here. Dvdt pin open is the fastest start rate so that won't help here.

I see that constant 3A load during startup. Did you face any issue during normal startup with this much load or you turned on the load after the device performed the initial startup? 

It looks like this is very high load and device will struggle due to constant current load during startup. If you enable the load after the device startup (using PG pin, enable the load) the device should be able to handle the load. 

Best Regards,
Arush

Hi，Arush

Regarding your first question.
The current configuration of Dvdt pin is 0.82nF.

--

Regarding your second question.
Let me mention here again that CR Mode was used during testing.
No problems will be encountered when powering up with load.
The problem of being unable to restart occurs only after short circuit protection.

--

Regarding your third narrative.
Do you mean that in the automatic restart, because it starts in the current limit mode, it may not be able to carry the load and start successfully?

--

Let me describe the problem again,
1. In CR Mode, I use the setting configuration of VIN 20V, 3A. Under this configuration, when the power is first turned on, it can carry the load and start successfully. Initial startup will be no problem and the TPS25982 appears to be working fine.

2.While Load remains on, short circuit protection is triggered.
After the short circuit is removed, theoretically it should be possible to successfully start with the load. But it will actually trigger the protection and lead to an infinite restart state.

--

Thank you for your reply.

Understood, so for same load device is turning on during initial startup but fails to do so after restart. 

Yes.
Do you have any ideas?
I saw that the power at startup is also very low, and I don't quite understand that it can succeed the first time, but the Auto-retry after a short circuit will fail. Both are the same configuration.

Thank you for your reply.

Hi Guanming,

I think the device shouldn't startup with this load condition as continuous current drawn by resistor is high and capacitor will also cause high inrush current. Let me try this startup and auto retry test in our lab with EVM. I will apply 3A CR load along with 2mF cap at output and will check for startup and autoretry after short circuit check. Please give me time till end of week as I can only perform test based on availability of test bench. 

Best Regards,
Arush

Hi，Arush

I'm sorry, maybe I didn't explain something clearly and caused you to misunderstand.
The current testing problems all occur on the public version of EVM, and the backend has always been tested using electronic load.

--

Let me talk about my testing method here.(Vin:20V，Iout:3A)
1. Before starting, turn on Load (CR Mode 3A) and turn on the power supply. The load can be successfully powered on.
2. Press the short-circuit button of the electronic load. After entering short-circuit protection, cancel the short-circuit button.
3. Load remains open at this time. I saw in this step that after entering protection, Auto-retry cannot be restarted.

-

If you need more information from me, I will get back to you as soon as possible.
Thank you for your reply.

You also have a capacitor load of 2mF along with 3A CR load. Am I correct?

Best Regards,
Arush

Hi，Arush

Are you talking about the circuit diagram on the public version of EVM?

I roughly understand the problem you are talking about. When I just re-examined the public version of the EVM line,
I found that there is a large capacitor of 1800uf in the official document, but in fact the piece I got did not have this capacitor.
The actual capacitance is only 221uF.

Thank you for your reply.

Oh, so your application requires 2mF cap, but you are not using it in the mentioned auto retry test.

You can just remove the electrolytic cap (there will still be a ceramic cap at output side of EVM) and perform this test again. It should work. Please share some waveforms if it doesn't work.  Waveform with longer time scale so that retry are visible.

Best Regards,
Arush

Hi，Arush

I'll test it as soon as possible and attach the waveforms along with it.
If it's convenient for you, please help me test it to see if I can have the same status.
If possible, could you please send me a copy of the environment configuration and test waveform diagram during your test?
--
I will get back to you as soon as possible after testing.

Hi Guanming,

Sure, if it doesn't work in your setup then I can test it on my end. Another issue is that I don't have this variant. I have other variant device which have over voltage clamp so won't be able to test it for 20V. I have ordered some devices and should get them by early next week. 

Best Regards,
Arush

Hi，Arush

Oh!! Thank you very much for the direction. When I tested it later, it looked almost the same as the result I calculated using excel tool.

The pdf below contains some test waveforms and my inferences. I would like to ask you to help me check whether it is as I guessed.

TPS25982 Auto-retry related tests.pdf

Thank you for your reply.

Guanming.

Hi Guanming,

I will get back to you by EOD or Tuesday Morning (IST). I am stuck with some other bench testing. Thank you for your understanding. 

Best Regards,
Arush