1kW PFC-Boost - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Power Supplies
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Thread Tools Search this Thread
Old 20th September 2005, 01:51 PM   #11
diyAudio Member
Join Date: Dec 2003
Location: Munich
Next step is the GND foil. It is one origami foil that connects the film cap, the e-caps, the MosFet-source-resitors and the signal GND. Signal GND is the large copper area and no high currents are fed into this GND area. The signal GND foil carries the PFC chip, which is pressed with an additional GND strip onto the GND foil. ==> improved heat dissipation and optimized GND plane design in one step.
Attached Images
File Type: jpg pwr_stage1.jpg (99.6 KB, 1700 views)
  Reply With Quote
Old 20th September 2005, 01:52 PM   #12
diyAudio Member
Join Date: Dec 2003
Location: Munich
And now things are coming ugly for human perception ...all the small components in a three dimensional P2P set up....
You can see the source resistors (light green) providing the shortest connection to PWR_GND.
Behind the light blue resistors you might find the 5 PNP drivers for the MosFets.
And the snubber cap (consisting of four WIMA MKP10 2.2nF/2000V) coming from the drains and connected to the snubber diode which is directly soldered on the top of the PWR_GND.
Attached Images
File Type: jpg pwr_stage2.jpg (98.5 KB, 1592 views)
  Reply With Quote
Old 20th September 2005, 01:54 PM   #13
diyAudio Member
Join Date: Dec 2003
Location: Munich
Here you can also see the discharge resistor of the snubber. The length of its connection is
uncritical (more or less...). In this set up the max. operation frequency at approx. 15W output power is around 100kHz and the snubber losses remain acceptable.

The three black sleeves carry several serial connected resistors, such as the 14x15k etc.
The sleeves are triple layered.
Attached Images
File Type: jpg pwr_stage3.jpg (99.1 KB, 1503 views)
  Reply With Quote
Old 20th September 2005, 02:17 PM   #14
diyAudio Member
Join Date: Dec 2003
Location: Munich
Originally posted by Eva
Sorry, I had forgotten that you were using critical conduction mode. This is nice as to what concerns to switching losses since frequency goes down as current increases thus keeping losses under control.

Have you tried to place a 2.2nF (or maybe 1.5nF) capacitor directly between D and S pins of the switching device? The basic idea under that is to have an intentionally slow gate turn-on so that the capacitor charge peak current is kept low, and a relatively fast gate turn-off (but not fast enough to require a dozen of PNP buffers!). Such a capacitor in conjunction with MOSFET turn-off time and internal drain-source capacitance will limit rise time to approx 100ns even with 12A load, but in a less dissipative way, and shall ringing be present it would happen at fairly low frequencies (so little will be radiated). Note that your particular inductor design nicely allows for slow turn-on since the inductance is high for low currents, so inductor current rise in the first 500ns should be quite small, thus producing low losses.

The boost inductor will tend to resonate with that capacitor but I don't consider that as a problem.

Also, I would try these bead ferrites placed in the inductor side instead of the diode side. You shouldn't slow down diode commutation.

Hi Eva,

yes, I tried the cap directly between drain and source.
But this also slows down the downward sloping speed, which results in several drawbacks.
a) In general slow down sloping causes an increased time without reasonable current in the inductor, means I am leaving critical conduction mode and enter discontinous mode. Down sloping frequency is given by the inductor and the snubber cap and down sloping would become very slow....
b) but I tried it and found that the turn ON current peaks were harder than expected. The chip detects the turn ON timing, when the voltage across the choke is close to zero. The turn ON delay (some 300...600ns) should the hit the turn ON into a low drain source voltage voltage. Well with a large cap directly from drain to source - the turn ON hits into several hundreds volts drain source...
and discharges the caps with a heavy spike.
c) the chip seems to have issues with detecting very slow slopings and missed out several cycles especially at high input voltages (may be this could be fixed by more auxiliary turns on the choke).
d) this method was OK for snubber caps up to about 500pF, but 500pF still resulted in unacceptable fast du/dt at high loads.

Hm, seems like I have to live with the lossy snubber... or set up a completely new design...
But it looks to works acceptable now. As you already mentioned my stepped gap inductor prevents excessive operating frequencies and in combination with the additional power in the snubber (which also reduces fmax) .... it seems like I will typically not exceed 120kHz.
  Reply With Quote
Old 20th September 2005, 04:26 PM   #15
Eva is offline Eva  Spain
diyAudio Member
Eva's Avatar
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
I would consider a dead time of 1us negligible at low operating frequencies like 25Khz to 35Khz. The corresponding periods are between 40us and 28us so 1us of dead time means almost nothing in efficiency terms (but it may solve all your EMI issues).

Also, 1us of deadtime combined with a slow rise time due to capacitor charging will help a lot to reduce maximum switching frequency with low loads. I would definitely take advantage of that fact if I was you, even at the expense of using a different control IC.

PD: The layout looks terrible
  Reply With Quote
Old 20th September 2005, 05:13 PM   #16
diyAudio Member
Join Date: Dec 2003
Location: Munich
Terrible layout? Which layout?

Well, I don't care about "ugly inside".
But reliability might be a drawback, when considering that this PSU should operate inside a vibrating subwoofer...

I agree that 1us dead time would be fine.
But 2.2nF with 1.1mH will result in a natural resonance of 100kHz.
So theoretical sloping may take around 5us then, practically even more.
Already in the current design I am observing slow down sloping of around 3us....

But point b) & c) where my main reasons.
I also agree that for turn ON I could increase the impedance.
This would allow to bring down the turn ON current peaks, but
turn ON losses would be still defined by the hard switched energy.
For turn OFF, I expected to be able to work without PNP drivers,
but the low impedance was the only way which worked out without ringing in real life. Even with 2.2nF subber, this is the limit where I again observe starting issues during MosFet turn OFF.
With smaller snubber caps I simply did not manage to get a proper turn OFF, even not with that massive PNP driver.
My high heavy inductive load seems to interact with the MosFets
and gate drive in a very inconvinient way.

Hm,... should I really design a constant frequency PFC similar like yours?

Especially with your magnetic snubber this seems to be a nice solution that could handle wide range input, low load / full load,
and general high power much easier.

Or should I simply use my proto as long as it works?
...feeling some contradiction of my reliability demands vs
available DIY time...
  Reply With Quote
Old 20th September 2005, 05:30 PM   #17
Eva is offline Eva  Spain
diyAudio Member
Eva's Avatar
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
Try an IGBT instead of a MOSFET, you may like the result.

Don't think that my magnetic snubber is going to solve all your problems, it may actually make things worse since it produces a nice 600V spike at turn-off (always within the avalanche ratings). It will probably cause a lot of trouble with MOSFETs, but IGBTs provide inherently limited rise times... and low capacitances... try them...
  Reply With Quote
Old 20th September 2005, 07:01 PM   #18
diyAudio Member
Join Date: Dec 2002
Location: Slovenia

sorry, I also was under impression that your design is continious mode. The only experience I had with L6562 is very small flyback capacitor charger powered from 12V. Even so, the chip seemed a little touchy to me, especially when compared with very robust UC3843 from the same manufacturer. I rather use UC3843 as controller for discontinious flybacks. Frequency control requires addition of only one external npn from the reference to the CT pin, with R-C series connected from the base of the npn to the drain of the switching mosfet.

I also agree with Eva about IGBT. I once designed two transistor (asymmetrical half bridge) laboratory flyback capacitor charger working in discontinious mode. I used large IR mosfet modules, but they failed regularly, until I exchanged them for first generation of IGBTs in TO247 case. Peak current was 70A and charger is still working today.

Best regards,

Jaka Racman
  Reply With Quote
Old 21st September 2005, 05:51 PM   #19
diyAudio Member
Join Date: Dec 2003
Location: Munich
Thanks to you both ! !

Great hint ! Sometimes I need some external help
to overcome my blindness....
I also think that an IGBT might be an interesting option.
Everybody is blaming them about their slow switching properties and especially their current tail.... But in my application this really could be
an advantage.
It might take again some time, but I plan to try it.

...hoping to come back with new results soon...
  Reply With Quote
Old 24th September 2005, 02:48 PM   #20
diyAudio Member
Join Date: Dec 2003
Location: Munich
IGBT seem to be quite extraordinary for amateurs.
XXXXX has tons of them in their catalogue, but in the shop they had nothing, because "nobody is interested in buying that...".
Well, in finally he found one single IGBT. G4BC20U from International Rectifier, 600V/13A. Not powerfull enough for application, but worth a trial....
( Our professional electronic shops in Munich are closed on Saturday and currently I am not able to go shopping instead of work during the week......)
G4BC20U showed less ringing effects, but it was not possible to use a high driving impedance that would result in reasonable du/dt. During my ongoing tests I blowed the small IGBT and it took also the TDA4862 into death. Well, - I saw that an IGBT seems to match better to my requirements, but would not fully solve my issues.

I decided to improve the gate drive of my MosFets and go on with them. The new TDA4862 was more touchy than my first one (may be a fake?).
I tried one of the old L6560, which is now working alright after some adjustments.
Improvement of gate drive: With the previous 27 Ohms between Chip and PNP driver was still some ringing at full power. It was clearly measurable that the output of the chip did not ring, while the base of the PNP already showed slight ringing and the MosFet gates again worse. Now with 10 Ohms this is fine and also the voltage and current wave forms of my MosFets are looking alright even at full power.

The white trace is the Drain voltage vs GND, 200V/Div.
The coloured trace is the current: 6,4A/Div.
Time base: 5us / Div.
Attached Images
File Type: jpg boost797w.jpg (98.1 KB, 1320 views)
  Reply With Quote


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

Similar Threads
Thread Thread Starter Forum Replies Last Post
11.5-12.8v to 13.8-14.4v boost star882 Power Supplies 16 19th May 2010 03:13 AM
[enquiry] Treble Boost/High frequency Boost circuitry elnino86 Parts 2 18th December 2008 12:33 PM
1kW Boost allanon77 Power Supplies 17 4th February 2008 12:18 PM
1W Boost allanon77 Power Supplies 1 22nd January 2008 09:13 AM
EQ boost and Q......help Javaman Multi-Way 1 20th January 2002 02:15 PM

New To Site? Need Help?

All times are GMT. The time now is 10:19 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2