UCC28056: Both AC and DC input operation

Hi Doi-san,

UCC28056 should operate as a traditional boost converter if it is fed with a DC input voltage.

Best Regards,
Ben Lough

Hi Ben-san,

Are you checked DC voltage range?
My customer said,

UCC 28056 evaluated by AC input so far
For the EVM set, it was set to activate the COMP terminal of UCC 28056 remotely with the DC input voltage detection IC from the outside, it was set to start detection at about 80 V, but the input voltage that can operate is over 90 V.
I confirmed that it operates at the no load level.
This is because the starting voltage setting inside the IC becomes 90 volts in terms of DC input even if the comp terminal input level of this IC is open.
Therefore, even if the SW operation to boost conversion with DC 110V is possible, we do not think that it can not be used for devices that can not set activation guarantee at input lower limit (DC 88V) or less.
It is judged that there is no merit to evaluate it over time as it can not adjust the startup setting outside the IC.
Is there any possibility of using this IC under this condition as well?
If not, we would like to introduce again if there is a device that can start up by external setting at DC 110V (DC 88V ~ DC 143V) or less voltage.
When there is nothing, there is no excuse, but we will consider it with other company's products.
Please teach again.

Hi Doi-san,

Changing the divider ratio of the resistor chain between the MOSFET drain and the ZCD/CS pin will adjust the start threshold.

Best Regards,
Ben Lough

Hi Ben-san,

Now still exist problem.

UCC28056 and UCC256302 combination problem at DC input

Plpblem 1. See attached pdf.
Whether hunting of 5 V output is caused by malfunction of PFC or LLC has not yet been determined.
 
Ploblem 2.
At the same time, application evaluation with DC 110V input compatible is required at the same time,
and for the time being, I tried to operate the input startup voltage circuitally and operate it with DC input,
but the input current burst mode setting level of the example is within UCC28056
Because it is fixed from the setting, burst operation is forced up to about 30% of the maximum rated output of LLC (about 100 W).

Do you have countermeasure?

UCC28056 and UCC256302 combination problem at DC input.pdf

Hi Doi-san,

In figure 1, it looks like VCC drops to VCCrestart threshold and then the LLC HV startup function re-enables, charges up VCC again and when VCC reaches 26V, starts switching again. The depletion of the VCC cap is likely the cause of problem 1. You could do a couple things to address this:
1. increase the bias winding turns ratio to have a larger VCC voltage
2. reduce the startup time so that the bias winding replenishes the VCC voltage before VCC has depleted
3. increase the VCC capacitance to have more charge available during startup.

Could you share how the enable remote start was implemented? It might be affecting the control loop and making the PFC unstable which would look like burst mode.

Best Regards,
Ben Lough

Hi Ben-san,

Please look at the attached pdf.

They can show the" DC input gradually rise problem" is occur even both EVMs are original status.

They have two LLC EVMs. One is original status EVM.

If you can, try this same way.

UCC28056&UCC256302 DC in voltage gradually up.pdf

Hi Doi-san,

The issue is caused by VCC dropping too quickly. I would suggest adding more capacitance to VCC instead of RVCC. Reducing the startup time (reduce the switching frequency of the LLC) would also help get the bias winding voltage to rise faster.

I have not seen this particular issue when pairing UCC28056 and UCC25630-1 EVMs in my own testing. I can try this testing and share with you but it will take some time to acquire the EVMs. In the meantime, could you share how the HV pin was interfaced to the UCC28056 EVM? Normally we take diodes D1 and D2 off the UCC25630-1EVM and solder the anodes to pin 2 and pin 3 of the diode bridge on the UCC28056 EVM, connect the cathodes together and connect the cathodes to TP10 of the UCC25630-1EVM. Just want to understand if perhaps there is a difference in the EVM configuration.

Best Regards,
Ben Lough

Hi Ben-san,

Mey be, are you misunderstanding device?

IC name is UCC256302.

So I think HV terminal of UCC256302 shoud be connected to Buss voltage through 5kohms.

It is described in 7.2.2.16 of the data sheet.

Does someone tried the combination UCC28056+UCC256302?

Doi-san,

UCC256302 can either be fed by the AC line through diodes or from the PFC bus. Either is acceptable.Are you connecting the HV pin resistor string to the PFC bus?

Best Regards,

Ben Lough

Hi Ben-san,

You know, Korea and Japan tests of lightning surge is hard.

PFC bus to HV pin is better.

And, customer said, ( I said they are very strict )

It is confirmed that the condition under which this malfunction occurs is slightly related to the startup boosting of the PFC output and the start of the LLC activation (charge / discharge cycle) and occurs only at a limited timing.
This is the case when the condition that the VCC of the LLC reaches the discharge end voltage (UVLO) after the step-up reaches the LLC bulk detection level (start of LLC start) is met.

For Japan design only, they try all the timing test  using a lot of  time to find ppm level malfunction. This is Japan quality.

Hi Doi-san,

Could you tell me the turns ratio between Primary to Auxiliary winding for the transformer your customer is using? It could be possible to increase this turns ratio to operate with a higher VCC voltage. The benefit of this is during startup, the bias winding voltage will reach 11V faster so the amount of time during startup the LLC IC is relying on the stored capacitor charge will be smaller. This would let you avoid the hunting you are seeing without having to increase the VCC capacitance.

Best Regards,
Ben Lough

Hi Ben-san,

You can see pdf I had up load.
This phenomenon can be found, both EVM are original condition , no change except UCC256302 and it's HV connection.
Next week they will try your three recommendations about UCC256302 Vcc Up.
I takes more and more time to confirm what is best solution for them.
I think you will try the same thing by using original EVMs.

Hi Ben-san,

Thank you for your solution.

But my customer said," It's a IC design problem".

Customer said it was proved that it is not logically reasonable as a countermeasure from the following trouble observation and explanation.

Please check te attached PDF.

And give me another solution.

UCC28056 and UCC256302 combination problem.pdf

Transformer.pdf