Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

1kW off-line SMPS claims, NO PFC

Status
Not open for further replies.

CataM

Advanced Member level 4
Joined
Dec 23, 2015
Messages
1,275
Helped
314
Reputation
628
Reaction score
312
Trophy points
83
Location
Madrid, Spain
Activity points
8,409
Hello everyone,

This video is claiming at 11:33 that he got 1 kW output power with the schematic shown at 11:50.

It is hard for me to believe since there is no PFC. Without PFC, the C4 and C3 RMS currents are so large that would blow them up.
Moreover, in the PCB there is no heat sinking nor space for its inclusion.
Furthermore, at 11:33, when she mentions the 1kW power, she states that in order to get the 1kW she added a "limiter follower" , short circuit protection, overvoltage protection, and smooth startup via relay... Nothing about power dissipation improvements.

Does somebody think this can happen like the video is stating ?
 

the "emitter" followers are on the fet drives, 1kW for short term is possible. at 49+49% duty cycle the fets would be soft switched...
 
  • Like
Reactions: CataM

    CataM

    Points: 2
    Helpful Answer Positive Rating
Does somebody think this can happen like the video is stating ?

I guess it is possible at a fixed input voltage and no load regulation; this effectively means fixed duty cycle. The transistor losses can be optimized and you can get efficiency greater than 95% - that means losses reduced to less than 50W at 1kW.

The 50W dissipation can be divided into transistor losses and transformer losses. Just rough idea: transformer losses may be perhaps 5-10W at 1kW and diodes losses may be another 5-10W leaving about 25-30W switching losses - 12-15W per transistor.

It would be good to fit the power transistors and the diodes with heat sink. But they may perhaps work if the output load is 1kW only for a short period.

I do not understand why C3 and C4 would blow up? They dissipate only small power but I anyway suggest 2-3 capacitors in parallel.

I agree including PFC is a very good design habit for all high power devices.
 

The 50W dissipation can be divided into transistor losses and transformer losses. Just rough idea: transformer losses may be perhaps 5-10W at 1kW and diodes losses may be another 5-10W leaving about 25-30W switching losses - 12-15W per transistor.
As you can see, that would not work. If the TO220 package she is using is without heatsink, Rth(ja)=62K/W => temperature increase over ambient=15*62=930ºC

Even with soft switching as Peasy is saying, the maximum power loss of 1 transistor is 2W (without heatsink). No one can dissipate less than 2W at 1kW without at least additional snubber circuitry and very very low Rds(on).

1kW for short term is possible
What does short term mean? How short ?

I do not understand why C3 and C4 would blow up? They dissipate only small power but I anyway suggest 2-3 capacitors in parallel.
It is not due to the power loss. Is the RMS ratings of the cap. Look e.g. at DigiKey how big a cap that can withstand >8 ARMS rated at 400V is and look at hers.
 

I skimmed the video, they show it real quick so who knows what it looked like when they tested it or how long they ran 1kw for.

Yeah TO-220 is only good for 1-2W or so. Maybe double that with very good airflow and maybe double that again for a transient if you're ok pushing twards or above 150C.
 

I agree with CataM. You need two good 470uF capacitors to withstand the ripple current of a 1 kW non-PFC rectifier.
 

Also there are holes on the pcb shown for supporting the heatsinks - which heatsinks are not there...
 

The original circuit was perhaps rated for 50-100W output. This was not clearly stated but appears from the design data. At this power level, heat sinks are not required.

The original circuit has only 1.8K connected to C3 and C4. C3 and C4 cannot have 8A RMS in any case.

The modified circuit has been designed for 0.5 kW output and that can certainly give 1kW for a short time (1-2 s)- I have not calculated.

Even for 500W, the C3 and C4 will not have 8A RMS current. (total current 2.4A - that includes losses and with filter capacitors rated well, about 1/2 will come from the capacitors- say 1.2A)

For many applications where you know the load well, simple designs are better and you can get rid of many features.

You will also have to redesign the transformer for higher load (transformer designed for 500W without sufficient headroom may not supply 1kW even for a couple of seconds).

In the original IBM PC AT power supply, they included a separate PFC board but most cheap computer SMPS simply do not bother (they are often rated at 550-650W).

There are several points where improvements can be made in the original circuit but that is a different story altogether.
 
Last edited:

The original circuit has only 1.8K connected to C3 and C4. C3 and C4 cannot have 8A RMS in any case.
470uF at 1kW gives around 7.6 A RMS. (in the cap)
470uF at 500W gives around 4.4 A RMS. (in the cap)

Even for 500W, the C3 and C4 will not have 8A RMS current
Of course not. I was always referring to the 1kW issue.
 

Hello everyone,

This video is claiming at 11:33 that he got 1 kW output power with the schematic shown at 11:50.

Liquid Nitrogen, anyone? :shock:

Now seriously.
Did anyone notice the thin, thin traces? On a single sided board? How can it withstand a full load current?

The problem with the web in general and Youtube in particular, is that anyone can publish anything they want. If you really want to see some truth and science deprived videos, search "flat earth".
 

470uF at 1kW gives around 7.6 A RMS. (in the cap)
470uF at 500W gives around 4.4 A RMS. (in the cap)

I fail to understand your calculations.

220VAC is RMS and it stays 220V RMS after full wave rectification (just ignore the diode drops)

The controller is simple and fixed at 50% (just ignore the dead time) duty cycle. The RMS value will be reduced: 220/1.4=160V

The RMS current will be 6.25A for 1kW output and ONLY part of this will come from the filter capacitor.

By the way, the original circuit is described at 1:15 in the video and the C3 and C4 are both protected by 1.8k series resistor.

Perhaps you are talking about the circuit at 11:50 but then the value of C3 is not clearly marked.

How did you get 7.6A RMS current in the cap? What was the total RMS current you got?
 

The RMS current will be 6.25A for 1kW output and ONLY part of this will come from the filter capacitor.
160V*6.25A~1kW = Wrong approach. Power Factor surely not even close to 1.

Perhaps you are talking about the circuit at 11:50 but then the value of C3 is not clearly marked.

How did you get 7.6A RMS current in the cap? What was the total RMS current you got?
Yes. That is the schematic I am referring to.
C3=1uF and C4=470uF.

See attached. (I know it is an unregulated converter, but a resistive load gives just slightly less RMS current in the cap)
 

Attachments

  • Constant power load 1kW.rar
    1.1 KB · Views: 71

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top