Reference design RDR-838: Role of MOSFET Q3

Hi Pratik,

Thank you for reaching out to Power Integrations and for your interest in our devices.

MOSFET Q3 in RDR 838 is called a bus switch, it functions as a series pass FET that connects and disconnects the output of the converter to the receptacle of the USB. Basically, when MOSFET Q3 is on it connects the output of flyback converter to the type c receptable and when it is off, the bus voltage discharges through R23 and D5. VBUS is a standard requirement for safety as the output of the flyback converter should be disconnected when no communication is made through the CC pin. 

For alternative VBUS FET, here are the standard guidelines for choosing:

A low-cost, n-channel MOSFET with at least 30 V rating and sufficient continuous drain current rating can be used. 
A low RDS(ON) (< 10 mΩ) is also preferred to reduce conduction losses. 
VB/D pin gate drive is typically 4.5 V with respect to VOUT so a gate voltage threshold voltage range (VGS(TH)) of 1.5 V to 2.5 V are recommended.

Here is an alternative solution with respect to the output requirements of RDR 838

AON7318 - AON7318.pdf (aosmd.com)

Thank you

Hello PI Exert,

Thank you for the response related to the role and selection bus switch.
In the reference design RDR-838,  AON7520 is used (typical gate threshold voltage 0.85V) and is obsolete.

In line with your suggestions, I have used AON7418 (typical gate threshold voltage 1.8V), which is very close to your suggested AON7318 (typical gate threshold voltage 1.7V)
Still, I am able to get the output at the USB connector only when the Q3 bus switch is shorted on the PCB board. It seems AON7418 is not a  suitable alternative.

Is the gate threshold voltage responsible factor? 
Please suggest a suitable solution for the bus switch operation.

Thanks and Regards,
Pratik N
 

Hello PI team,

Kindly let me know if there are any updates in the above query.

Regards,
Pratik N

Hi Pratik,

Apologies for the late response.

FYR for the terms used:

VBUS_IN - Node prior the VBUS Switch

VBUS_OUT - Node after the VBUS Switch

May I clarify whether your VBUS_IN have any voltage when you connected the Bus Switch you used. Additionally, is your VBUS_OUT is 0V or do you see any low voltages. This is to check whether the issue is just the VBUS Switch being only partially on.

If you found that the VBUS_IN is still regulating at 5V while there is low/no voltage at the VBUS_OUT,   kindly check the ff and ensure that the faults mentioned below are not present, this is assuming that no latch-off or auto-restart are observed during power up: 

1.  CC1and CC2 pins are shorted to each other.
2.  VB/D pin is open.
3. GND and CC Line are shorted to each other.
4. BPS and CC Line

Kindly check all the pins and ensure there are no open pins or pin-to-pin short.

I will be waiting for your response.

Thank you.

Hello PI Expert,
1) I have 2 boards available with me. It was found that the VBUS_IN is still regulating at 5V while there is no voltage at the VBUS_OUT.
So, I have checked both the boards for the above 4 suggested cases and they are found ok (none of the above issues detected)

2) Later, I found out that the start-up issue was mainly due to the electronic load operating in CC mode. Both the boards worked with CR mode.

3) Further, I observed that one of my board works fine for entire load range in CR mode during start-up. Here, the transformer used was fabricated from a standard manufacturer.
While, the other board works fine during start-up at selective light load condition (near about 60 ohms or higher at an output voltage of 15V). Here, the transformer used was fabricated in-house.

The question is does the transformer fabrication affect this start up operation?? Please clarify on the same.

Thanks and Regards,
Pratik

Hi Pratik,

Would it be possible for you to raise a ticket to our Online Support using the link below:

PowerPros Online Support | Power Integrations

This is so that I can request for the full details such as schematic, transformer construction etc. personally to further investigate why the design is behaving that way.

Let me know if there will be issues or concerns with raising a ticket to our Online Support.

Thank you.

Hi Pi Expert.
Thanks for your quick response. I will raise a ticket for this, after I have a new transformer sample fabricated and tested for the same.

In the initial release:
Actually, no changes were made to the schematic and layout for the tested boards.
Also, the transformer fabrication was same as suggested for RDR-838.

Regards,
Pratik 

Hi Pratik,

Thank you for the response as well.

We will be waiting for your ticket on our side and let's discuss further details there.

Thank you for your understanding.

Regards