LED error of DLP4710 chipset

Rossinag

Other Parts Discussed in Thread: DLP4710, DLPA3005

Hello All,

Thank you very much for your support.

1. We would like to know what kind of LED condition induces the "Byte 2 -LED Status" error of the "Read System Status(D1h)" command.

Is it possible to provide this information?

2. In our testing of DLP4710 chipset with DLPA3005, LED does go off in the case of short-circuit of the anode/cathode.

This suggests us that there seem to be kind of a short protection function in the DLPA3005 device.

1) Is this the assumed function in the DLPA3005?

2) If yes, Is it possible to describe summary of this function?

3) Can we know this occurrence by the control command or something?

Best regards,

Rossinag

over 9 years ago

0 Nadine Estermann over 9 years ago

TI__Genius 10495 points

Hello Rossinag,

The LED Status will show an error when no LEDs are connected. Are you currently seeing this error on your system ?

Regarding the fault condition I will check internally with our team and will get back to you as soon as possible. In general a short between the anode and the cathode is not a valid condition. Do you have any use case for that ?

Best regards,

Nadine

0 Rossinag over 9 years ago in reply to Nadine Estermann

Expert 1705 points

Hello Nadine,

Thank you for your reply.

Please let me try to describe.The short anode/cathode LED situation does not happen in usual operation.

In order to do the any test by our customer, this phenomenon was found in their abnormal circumstances testing.

So they would like to know the mechanism of the LED going off in case of the short between anode and cathode.

Best regards,

Rossinag

0 Nadine Estermann over 9 years ago in reply to Rossinag

TI__Genius 10495 points

Hello Rossinag,

In case anode and cathode are shorted together a PG(Power Good) error will occur and the DLPA3005 will shutdown. Please see copied information about a PG error from the DLPA3005 datasheet.

7.3.2.5 Illumination Monitoring

The illumination block is continuously monitored for system failures to prevent damage to the DLPA3005 and LEDs. Several possible failures are monitored such as a broken control loop and a too high or too low output voltage VLED. The overall illumination fault bit is in register 0x0C (ILLUM_FAULT). If any of the below failures occur, the ILLUM_FAULT bit may be set high:

• ILLUM_BC1_PG_FAULT

• ILLUM_BC1_OV_FAULT

Where, PG= Power Good and OV= Over Voltage

7.3.2.5.1 Power Good

Both the Illumination driver and the Illumination LDO have a power good indication. The power good for the driver indicates if the output voltage (VLED) is within a defined window indicating that the LED current has reached the set point. If for some reason the LED current cannot be controlled to the intended value, this fault occurs. Subsequently, bit ILLUM_BC1_PG_FAULT in register 0x27 is set high. The illumination LDO output voltage is also monitored. When the power good of the LDO is asserted it implies that the LDO voltage is below a pre-defined minimum of 80% (rising) or 60% (falling) edge. The power good indication for the LDO is in register 0x27 (V5V5_LDO_ILLUM_PG_FAULT).

Please let me know if you have any further questions.

Best regards,

Nadine

0 Rossinag over 9 years ago in reply to Nadine Estermann

Expert 1705 points

Hello Nadine,

Thank you very much for your help.

Please let me ask you following.

We confirmed that it looks like DLPA3005 does not shutdown in case the anode/cathode is shorted. Because the LED turns on immediately after the short situation is removed. Also the ILLUM_BC1_PG_FAULT does not indicate the error. This can be confirmed by LightCrafterDisplay4710 EVM. There seem to be something protection function is working.

Would you please provide comment again?

Best regards,

Rossinag

0 Rossinag over 9 years ago in reply to Rossinag

Expert 1705 points

Hello Nadine,
Thank you for your support.
I added question in my previous reply.
Would you please find it?

Best regards,
Rossinag

0 Nadine Estermann over 9 years ago in reply to Rossinag

TI__Genius 10495 points

Hello Rossinag,

Sorry for the delay.

When did you read the register ? The register will be reseted after the fault condition is gone.

Best regards,

Nadine

0 Rossinag over 9 years ago in reply to Nadine Estermann

Expert 1705 points

Hello Nadine,

Thank you very much for your comment.
We read the register 0x0C "Main Status Register" of DLPA3005 when LED was shorting and the ILLUM_FAULT showed "0".
Also the register 0x0D "Interrupt Mask Register" showed 10100101, this means that the ILLUM_FAULT_MASK=0
"0: Not masked for ILLUM_FAULT interrupt" .

For our test we shorted the anode and cathode by using tweezers. At this short situation by the tweezers LED turned off.
And LED turned on when we removed the tweezers from anode and cathode. Therefore it looks like that the DLPA3005 did not shutdown when LED was in short situation. Would you please provide advice?

Best regards,
Rossinag

0 Nadine Estermann over 9 years ago in reply to Rossinag

TI__Genius 10495 points

Hello Rossinag,

Your observation is correct.

In case the LED is shorted, the VLED control loop controls VLED down such that the current (through the tweezer short) is equal to the set current again.

Note that the control loop controls VLED such that the intended voltage across RLIM is obtained (ISET x RLIM).

You can also think of the tweezer as an extreme case of a minimum LED forward voltage, i.e. about 0V.

Once the tweezer is removed again VLED needs to go up again (by the LED forward voltage) to have ISET running again.

Please let me know if you have any further questions.

Thank you,

Nadine

0 Rossinag over 9 years ago in reply to Nadine Estermann

Expert 1705 points

Hello Nadine,

Thank you very much for your support.

>In case the LED is shorted, the VLED control loop controls VLED down >such that the current (through the tweezer short)

>is equal to the set >current again.

> Note that the control loop controls VLED such that the intended voltage >across RLIM is obtained (ISET x RLIM).

I found a block diagram of "Figure 4. Illumination Control Loop" in the www.tij.co.jp/.../dlpa3005.pdf.

But ,sorry, since I could not understand , may I ask you following?

Does the "ISET" mean the current for the LED?

Why does the VLED control loop control VLED down when the LED is shorted by the tweezer?

Best regards,

Rossinag

0 Nadine Estermann over 9 years ago in reply to Rossinag

TI__Genius 10495 points

Hello Rossinag,

The control loops still has RLIM in the loop so it is not a short to ground condition. You can also think of the tweezer as an extreme case of a minimum LED forward voltage, i.e. about 0V.

The voltage across RLIM is compared with the current setting from the IDAC registers and the loop regulates the current to its set value. ISET is the LED current you would like to achieve in your system.

Since the forward voltage of the LED is in this case 0V the LED voltage VLED will be controlled down to achieve the intended voltage across the RLIM resistor.

Please let me know if you have any further questions.

Best regards,

Nadine

0 Rossinag over 9 years ago in reply to Nadine Estermann

Expert 1705 points

Hello Nadine,

It looks like following sentence is the keys.

Note that the control loop controls VLED such that the intended voltage across RLIM is obtained (ISET x RLIM).

7.3.2 Illumination

The illumination function includes all blocks needed to generate light for the DLP system. In order to accurately set the current through the LEDs, a control loop is used (Figure 4). The intended LED current is set through IDAC[9:0]. The Illumination driver controls the LED anode voltage VLED and as a result a current will flow through one of the LEDs. The LED current is measured via the voltage across sense resistor RLIM. Based on the difference between the actual and intended current, the loop controls the output of the buck converter (VLED) higher or lower.

7.3.2.1 Programmable Gain Block

The current through the LEDs is determined by a digital number stored in the respective IDAC registers 0x03h to 0x08h. These registers determine the LED current which is measured through the sense resistor RLIM. The voltage across RLIM is compared with the current setting from the IDAC registers and the loop regulates the current to its set value.

Regarding the intended voltage..

Does the Control Loop make the actual current(measured by RLIM) to the intended current(set by IDAC)?

Or does the Control Loop make the intended current setting of the IDAC to the actual current(measured by RLIM)?

In case of the VLED=0V(an extreme case of LED short condition), does the voltage through the RLIM show 0V? And then the IDAC will be controlled to low. Is this correct understanding?

Best regards,

Rossinag

0 Nadine Estermann over 9 years ago in reply to Rossinag

TI__Genius 10495 points

Hello Rossinag,

The illumination driver controls the LED anode voltage VLED and therefore the amount of current flowing through the LED and RLIM. Depending on the voltage across the sense resistor the LED anode voltage will increase or decrease depending on the LED current you would like to achieve.

In the IDAC register you define which amount of current you would like to achieve. The DLPA3005 will not change this register it only changes VLED to achieve the desired current.

In case of a LED current of 16A the voltage across RLIM (RLIM=0.009 Ohm) should be 144mV. The control loop will try to achieve this voltage across the RLIM resistor by increasing or decreasing the VLED.

In case that you short anode and cathode with a tweezer the voltage across the tweezer is nearly 0 V. So the DLPA3005 tries to regulate VLED down to achieve the desired voltage across RLIM.

After removing the tweezer the resistance in the loop increases and VLED will increase again.

Please let me know if you have any further questions.

Best regards,

Nadine

0 Rossinag over 9 years ago in reply to Nadine Estermann

Expert 1705 points

Hello Nadine,

Thank you very much for your support.

We understand that the DLPA3005 tries to regulate VLED down to achieve the desired voltage across RLIM ,in case we short anode and cathode with a tweezer.

I got one more question from customer. Please let me try to describe in below.

They monitored the amount of the current for the system board of the DLP4710 chip set by monitoring factory power supply display.

And they observed that the amount of the current was decreased when they forced to make LED short situation. Theoretically they thought the current which was generated by the ILLUMINATION DRIVER should keep the current(for RLIM@150mV) in this case. But the amount of the current was decreased.

From this phenomenon, there seem to be a function of current reducing in the DLPA3005. Is it possible to explain this?

Thank you and best regards,

Rossinag

0 Nadine Estermann over 9 years ago in reply to Rossinag

TI__Genius 10495 points

Hello Rossinag,

The DLPA3005 is using a buck converter for regulating VLED to achieve the desired LED current. Since the VLED voltage will be regulated down to nearly 150mV the ON time of the regulator is smaller than in the normal LED use case. Which means the OFF time is larger which saves power.

You can also refer to following document: www.ti.com/.../slva057.pdf

Best regards,

Nadine

0 Rossinag over 9 years ago in reply to Nadine Estermann

Expert 1705 points

Hello Nadine,

We understand it. Thank you very much for your description and kind support.

Best regards,
Rossinag

DLP®︎ products

DLP products forum

LED error of DLP4710 chipset