TIDM-02008

Sam,

Please note that today, July 5th, is a US holiday.  You should expect a reply no later than end of day tomm, July 6th.  Appreciate your patience.

Best,

Matthew 

Hi Sam,

Q1. 'What is the permissible range of that DC voltage change? Can we take that DC voltage to say 5% of the input AC voltage? If not, what is the barrier?'

Can you elaborate on this? Are you trying to regulate dc bus voltage 5% of your ac input? If so, this is impossible. This design is a boost PFC and therefore, your output voltage has to be higher than ac input all the time.

Q2. As long as you have high enough current and voltage rating device, you can build the system only with  4 switches.

Q3. We currently don't have 3-phase PFC design.

Best,

John

Hi John, Thanks for your reply.  Nothing mentioned in the reference design that says it is a boost design. Can you please state where is that stated that in the reference design?  https://www.ti.com/lit/ug/tidud61e/tidud61e.pdf?ts=1626258820119&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FTIDM-02008

Regards, San

Hi John,

The reference document also states the the DC output is programmable. See the reference design --> P.1 -->  Features --> 3rd bullet point. If it is programmable what is the range of the allowable change?

 https://www.ti.com/lit/ug/tidud61e/tidud61e.pdf?ts=1626258820119&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FTIDM-02008

Regards

Sam

John,

The reference document also (Figure 3-63) shows the DC voltage fully regulated at two voltage levels (0 and 400V). So it must be possible to fully regulate that voltage at other levels.

Finally, I received the following answer from one of your colleagues at TI.

Figure 3-61 shows this DC voltage change under transient. Voltage Transient With Non-Linear Voltage Loop, Vin 120 Vrms, 880 W to 0 W Transient, Overshoot 16.8 V. So this is about 4% of the rated DC bus voltage (400V)

Shammin

The answer is not very convincing but i believe he agrees that the DC voltage can be regulated at lower level than the nominal value.

Regards

Sam

Figure 3-63 shows the start up process and that does not mean you can regulate output voltage less than the peak of the input voltage.

If your ac input is 110 V which as peak at around 170 V, how can you regulate boost operation below 170 V? The board was designed to operate up to 260 V input and therefore, the nominal dc bus is set to 380 V which is always higher than the peak of 260 V.

Also, you will see an overshoot in step load transient but not sure how overshoot is related to the dc bus regulation.

Most PFCs are boost PFC and totem-pole bridgeless PFC is under the category of boost PFC.

The boost PFC has the inductor in the input side and it is advantageous to get continuous current.

Best,

John

Again,

The reference document

 https://www.ti.com/lit/ug/tidud61e/tidud61e.pdf?ts=1626258820119&ref_url=https%253A%252F%252Fwww.ti.com%252Ftool%252FTIDM-02008

states

" Programmable Output Voltage, 380-V DC Output Nominal "

You can find this statement in  P.1 -->  Features --> 3rd bullet point of that reference document.

If it is programmable what is the range of the allowable change?

Regards

Sam

For a 110V AC Input (RMS) --> 160V AC input (Peak)  --> Output DC ~= 170V nominal.. we can accept that

For a 260V AC Input (RMS) --> 370V AC input (Peak)  --> Output DC ~= 380V nominal.. we can accept that.too

What does the reference mean by saying programmable DC output?

You can set the output dc bus reference higher than the peak of ac input. (170 V and 380 V for low line and high line respectively).

However, you can't go down to below 5% of ac input because it is boost type PFC.

Ok,...how high can you go? For example for an input single-phase AC volt of 220V what is the output range we can realize? 

What is the voltage level that you want? The power switches are rated around 600 V and it limits the maximum voltage.

Designers typically put 50% margin on the device selection. (Ex: Power switches are rated at 600V for 400V dc bus).

We want to know the limit of the reference design. This has many ramifications for us.

You are saying for an input AC voltage of 220V the output DC voltage can be regulated between 380V DC and 600 VDC and if we choose to have a 50% safety factor the output DC voltage can be regulated between 380VDC and 400 VDC. Is that correct?

What if we use power switches with substantially higher rating (such that we rely on the design limits rather than the power switch limit). In that case is there a theoretical limit above which we cannot go? If yes, is there a down-fall for having a wide range of regulated voltage?

Regards

Sam

Q1.You are saying for an input AC voltage of 220V the output DC voltage can be regulated between 380V DC and 600 VDC and if we choose to have a 50% safety factor the output DC voltage can be regulated between 380VDC and 400 VDC. Is that correct?

-Yes

Q2.What if we use power switches with substantially higher rating (such that we rely on the design limits rather than the power switch limit). In that case is there a theoretical limit above which we cannot go? If yes, is there a down-fall for having a wide range of regulated voltage?

-If you use higher rating power devices, they typically have higher Rds_on and Eon/Eoff which mean you will see more losses. 

Thanks for letting know about the losses. We might accept the extra losses as we are keen on wide range of DC voltage regulation in some applications.

Can you please address the issue of output DC voltage range without delving into other issues?

Regards

Sam

Sam,

Apologies for the lack of reply from TI on this side.  John is currently OOO, I'm working to get someone else assigned here to reply back.

Best regards,
Matthew

Sam,

I have replied to your post, the design is only tested for 380V-400V range. The reason why we put programmable is that the SW is capable to change it, but we necessarily have not checked the HW for the breaking point, nor guarantee it.

Just for your reference, the design using 450V DC Bus Caps, and this will be the most limiting and i will not go any higher than 420V, because the DC bus cap will blow if the design is used as-is. 

We document the nominal conditions under which the design is tested, this is a programmable solution with digital control and our customers are free to design with 500V rated caps and increase the DC Link voltages. 

-Manish 

Thank you Manish

Now we know that;

- One limit/constraint on the regulated output DC voltage is the rating of the power switches

- 2nd limit on the regulated DC output is the losses (which increase as increase the rating of the power switches)

- 3rd limit on that output is the rating of the capacitors.

We certainly know that we have to ensure that the ratings these components (and all other components) have to be considered, and we will do that. We certainly know also that you only documented what you already tested and we appreciate that.

The question remains;

Does the topology (or the design concept) itself places a limit on the variation range of the DC side?

For example if the input AC voltage is 220 VAC can we get a regulated DC output voltage= 1000VDC (assuming we rate all the components properly)? If not, what is the constraint assuming we will re-design the inductor and every other component we need to re-design.

We are not asking for any test results apart from the ones documented.

Regards

Sam

This is the 8th time we pose the same question with no decent answer. We get answers for a spectrum of issues apart from our issue.

The question is;

In regard to the programmability of the DC out of this reference design what is the topology limit.

We know that;

- One limit/constraint on the regulated output DC voltage is the rating of the power switches

- a 2nd limit on the regulated DC output is the losses (which increase as increase the rating of the power switches)

- a 3rd limit on that output is the rating of the capacitors.

We certainly know that we have to ensure that the ratings these components (and all other components) have to be considered, and we will do that. We certainly know also that you only documented what you already tested and we appreciate that. Further, we also know that if we change the DC output of that design we might have to re-design resistors, capacitors or inductors.

The question remains;

Does the topology (or the design concept) itself places a limit on the variation range of the DC side?

For example if the input AC voltage is 220 VAC can we get a regulated DC output voltage= 1000VDC (assuming we rate all the components properly and assuming we will re-design the inductors and every every other component we need to be re-designed)? If not, what is the constraint?

We are seeking topology limit. We are not  seeking components limits NOR we are asking for any test results apart from the ones documented.

Regards

Sam

Sam,

This design is a boost type PFC and theoretically dc bus can be set to 1000 V.

Best,

John

Thanks John

Can you please refer us to any TI reference, book or paper that proves that the theoretical limit is 1000V?

Further, can you please refer us to any TI rectifier + PFC reference design, where the design is essentially a Buck converter?

Regards

Sam

400 V dc bus is most commonly used in the industrial and  automotive applications. I don't know if there are any docs that limit dc bus to 1000 V.

There are no C2000 reference design but here's the link for buck PFC 

https://www.ti.com/lit/ds/slusak8a/slusak8a.pdf?ts=1629378637436&ref_url=https%253A%252F%252Fwww.google.com%252F

https://www.ti.com/seclit/ml/slup264/slup264.pdf

https://www.tij.co.jp/jp/lit/ug/tiduas3a/tiduas3a.pdf?ts=1629378643749&ref_url=https%253A%252F%252Fwww.google.com%252F

Best,

John