TPS25942L: Priority Power Multiplexing Implementation

Hi Gareth,

Can you please elaborate your question in more detail.

Hi Praveen

Both PSU1 and PSU2 have power supplies

But during U27 OVP or OCP, U28 also has no output, Is that right?

Customers using TPS25944L x2 Parallel, Schematic as below

Hi Praveen

U27 FLT and PGOOD are both MOSFET grounded, change to FLT to operate normally,

1. When U27 OVP, U28 OVP is connected to U27 PGOOD, the output is no voltage (PGOOD still has 1V voltage)

2. When U27 OVP, U28 OVP is connected to U27 FLT, the output is U28 output

3.U27 =>MAIN PSU / U28=> BACKUP PSU

If U28 OVP is changed to U27 FLT, when the voltage of U27 is less than the voltage of U28, the output will be fixed by U28 and cannot be switched back to the main power supply of U27, unless U28 fails or the voltage is less than U27

Hi Praveen

PG1 still has 1V voltage when PSU1 OVP (the OVP of a single U27 is tested, the normal PG1 is 0V, as long as U28 is powered on, PG1 will have 1V) resulting in U28 also no output

Thank you

Hi Gareth,

Apologies for the delay in response. I will get back to you on this by tomorrow.

Hi Gareth,

Yes, you are right. If it is an OVP condition on U27 and not a UVLO condition, the PGOOD will still be high and U27 will turn OFF itself as well as U28.

In this case the FLT/ will help in turning ON U28 when U27 has OVP. 

Hi Praveen

Thank you very much

1.PG1 is the open drain pull low, where does 1v come from?

2.Because when U28 is not powered on, U27 has OVP. At this time, PG1 is 0v ......

2. If U28 OVP is changed to U27 FLT, when the voltage of U27 is less than the voltage of U28, the output will be fixed by U28 and cannot be switched back to the main power supply of U27, unless U28 fails or the voltage is less than U27

Hi Gareth,

PG is connected to OVP of U28 and PG1 is pulled up to VOUT. 

Are you seeing 1V on PG1 ? Under what conditions are you seeing 1V on PG1. 

Hi Praveen

Please see attached waveform 

Thank you

Hi Gareth,

Thanks for sharing the waveforms.

• When you say 'PGTH connect to VIN' , does that mean PGTH of U27 is connected to VIN of U27 (DCIN_12_M) ?
• Is PG of U27 connected to OVP of U28 or you are using 
• Are you sharing waveforms with circuit tested in the below configuration ? If not, can you share the schematic for each condition that you have shared waveform for ?

Hi Praveen

Sorry I update the current schematic as EVM

TPS25942 EVM.pdf

schematic is set as follows

removed (R19, R26, D5, D9, C4, C10)
J1 2-3 connect together
R8 short

Waveform as follows

Hi Gareth,

Thanks for confirming the schematic on which you have captured the waveform. I will review and get back to you by 14th Sept.

Hi Gareth,

Looks like you have applied over voltage at the input of Channel 1 and the device has turned OFF its output. This is why the CH1 has turned OFF.

Are VOUT1 and VOUT2 tied together ? What is the voltage applied on Vin2 ?

Hi 

VOUT1 and VOUT2 tied together ?yes

What is the voltage applied on Vin2 ? 12v

Hi Gareth,

If Vin1 goes above OVP and turns OFF, the PG1 should go low and as it is tied to OVP2 the channel 2 TPS25942 should turn ON and power the load.

I don't see any reason why the PG2 should go low. 

Hope VIN1 and VIN2 are independent power supplies. 

Hi Praveen,

 We have independent source for Vin 1 and Vin 2. In our test today on EVM,

1. When we don't tied Vout1 and Vout2 together, the result looks good.

2.  When we tied Vout1 and Vout2 together (J1 is in 2+3, PGTH tied to Vin), we see 1V on PGOOD1 so there's no output on Vout...

3.  If we change J1 sett to 1+2 (PGTH tied to Vout), Vout is swithing in 0V and 12V.

Please help to check it at your side and see how do we have to setup our environment.

Regards,

Gareth

Hi Gareth,

The root cause of the problem is that the PGOOD signal is not pulled low by the controller when there is an OVP event.

The PGOOD continues to be high and reflect the Pullup voltage (Vout1). So, initially when there is an OVP event, the output keeps discharging  anf the PGOOD also follows output as PGOOD is pulled upto Vout. Once the PGOOD-1/OVP-2 reaches less than ~1V, the Channel-2 turns ON and the OUTPUT starts to ramp up. When Vout ramps up, the PGOOD-1 and OVP-2 also ramp up there by turning OFF channel-2. This continues in a cycling process and the VOUT oscillates.

The Priority Muxing configuration mentioned in datasheet is not recommended for input OVP conditions. 

Hi Praveen

Thanks for your assistance

When using one channel , when IC OVP PGOOD is low

Why another pin FLT can be pulled LOW when OVP , But when ch1 returns to normal, ch2 cannot switch to ch1 

ch1 voltage must be greater than ch2 voltage to switch back to ch1

Hi Gareth,

I will get back to you within 2 days from now.

Thanks for your patience and understanding. 

Hi Gareth,

When the primary power supply is present and within limits the primary channel should supply power irrespective of whether it is higher or lower than the secondary power supply. Please refer to waveform captures from datasheet below.

Please verify the connections on your board and try replacing the IC with new IC and re test

Hi  Praveen

Thank you for your support

A-The connection method of waveform captures (Figure 80~83) is PG1 to OVP2 in the datasheet 

    But when it is changed to FLT1 to connect to OVP2   ,When the voltage of U27 is less than the voltage of U28, the output will be fixed by U28 and cannot be switched back to the main power supply of U27, unless U28 fails or the voltage is less than U27

B- When PG1 is connected to DMODE2, it takes a very long time (300ms) for VOUT to switch to PSU2 after PSU1 is turned off.

Will it not be (400uS) when changing to FLT1 and DMODE2? But does DATASHEET indicate the need to connect PG1 to DMODE2?