Author Topic: PCB Desing, power rectifier advice  (Read 1802 times)

0 Members and 1 Guest are viewing this topic.

Offline AtomTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: it
PCB Desing, power rectifier advice
« on: February 26, 2019, 07:23:22 pm »
Hi to everyone, so now i'm making the pcbs of a bench psu and i'm struggling to make all fit, since i have a max dimensions of 100*100mm..(i want to stay inisde this dimensions so the pcbs will be much cheaper ). for the ones intrested the power supply is the one engineered by Blackdog.

now the real problem is that i like to keep things cool so right now for testing i'm using a kbu6b diode bridge bolted to a generous heatsink, but that heatsink takes place so i need some way to reduce power disspation, i switched to discrete schottky diode bridge, i used 4 mbr20100.. they are cheap and the vf measured is 500mv at even 10A, so they are good, but there is still some power disspation, the thing is that i tought that i could use the pcb as heatsink by soldering the tabs to the copper, by doing so i have also a better path for the cathode of each diode and i can push more current through it (instead of a single to220 pin).

Now we have the schottky diodes on the bottom and the capacitors on the top, the schottky are dumping heat into the pcb , and that is bad for the capacitors that will warm up.

the power supply isn't going to supply more than 5 Amps but if i want i can push it to 8amps, so for calc purposes we can assume a 10A constant load , the capacitance on the output will be 30mF or even 40mf (capacitor of 35*50mm. thats why i don't have space).

so what i'm seeking is an advice on how to keep the diodes cool and also the capacitors. thanks to everyone

keep in mind that in the 3d "render" diodes and caps are placed randomly ..that is kind of what i want to do, but im here for advice
 

Offline keymaster

  • Regular Contributor
  • *
  • Posts: 58
  • Country: gr
Re: PCB Desing, power rectifier advice
« Reply #1 on: February 26, 2019, 08:58:03 pm »
the only way to keep things cool ,is to have a fan to make airflow .
Another option is to use mosfets for the bridge and a controller ,like the lt4320
(https://www.analog.com/media/en/technical-documentation/data-sheets/4320fb.pdf)
(https://eu.mouser.com/new/Analog-Devices/adi-lt4320-controllers/
« Last Edit: February 26, 2019, 09:07:18 pm by keymaster »
 
The following users thanked this post: Atom

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9327
  • Country: fi
Re: PCB Desing, power rectifier advice
« Reply #2 on: February 26, 2019, 09:06:31 pm »
You have probably reached a situation where increasing the efficiency is what you need: throw some hi-tech at it. At low voltages, traditional diode bridges are horribly inefficient. At line frequencies (50 or 60 Hz, I assume), doing synchronous rectification is trivial. You'll find control ICs for this purpose. Last time I checked, they were irritatingly expensive for what they really are, but that may not be a problem - it'll still be cheaper than a heatsink and/or a fan. Cooling things sucks and is bulky - consider it as a last resort and start by minimizing losses. Energy efficiency is then a bonus.

Silicon is cheap today; cheaper than heatsinking. Since your frequency is so low, you won't have switching loss issues, so you can totally oversize your synchronous MOSFET sizes (minimize Rds(on)) to have arbitrarily low voltage drops over them. 10x efficiency improvement over your diode solution is easy.

The capacitors won't heat up. However, such massive capacitance poses another issue: the inrush current might need some controlling, it may risk destroying your diodes (or MOSFETs if you go synchronous).
« Last Edit: February 26, 2019, 09:10:43 pm by Siwastaja »
 
The following users thanked this post: Atom

Offline AtomTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: it
Re: PCB Desing, power rectifier advice
« Reply #3 on: February 27, 2019, 06:03:55 pm »
Sorry for the late reply , yes i was aware of the lt4320 in fact the original designer uses it in the original desing , but it is still pricey for what it is.. i searched for alternatives and everything that i found is the LM74670..http://www.ti.com/lit/ds/snosd08a/snosd08a.pdf
but it even more pricey....so anyone knows other alternatives?

talking now on how to mount the pass transistors , i was thinking to make a board always (100*100max) with the pass transistor on one side and the controls an the other side , i have 2 transistor, and i was thinking of reusing old intel stock heaatsinks , they should be good and i can take away a lot more heat with a metal to metal contact instead of isolating the 2 with siliocon pads. i mean if we watch thermal conductivity we have aluminum 200W/m K vs the 1to5 W/m K of a silicon pad..well i think the choiche here is obvious.

so i was thinking of making a 3mm hole on the heatsink tap it and bolt down the transistor, then i'm going to extend the legs whit some 2mm copper wire to the pcb and solder it there, for keeping the heatsinks in their places i'm going to 3d print some mounts. so any suggestions?

 

Offline AtomTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: it
Re: PCB Desing, power rectifier advice
« Reply #4 on: February 27, 2019, 09:41:32 pm »
So i searched the net for a bit and this is what i found ..http://projectcircuit4u.blogspot.com/2010/04/power-mosfet-active-bridge-rectifier.html
i tried it on a breadboard and it works.. 20mohm p channel and 6mohm n channel, 1000uf output capacitance and a 1 amp load...the mosfet stayed cool as they should be ... any apparent flaws that a more experienced eye could catch?

i'll solder the mosfet to a copper pcb and do some more power testing tomorrow let me know of any flaws on the circuit :-+
a discrete mosfet bridge even if it doesn't reach the lt4320 standard could be very good for "budget builds"...i mean 10 or so resistor and an old quad opamp
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9327
  • Country: fi
Re: PCB Desing, power rectifier advice
« Reply #5 on: February 28, 2019, 06:28:11 am »
To prevent short-circuit shoot-through currents, it's paramount that the switches turn off before there is significant reverse current.

If you think about a diode, it does exactly that, measuring voltage drop over itself, just leaving 0.7V of "safety margin".

Every active circuit must leave some margin, which directly relates to the losses. But in essence, when you see that you soon need to turn off the transistors, you need to do it a bit earlier. At this point, the FETs change their behavior to standard (not schottky) diodes (which they are internally, in  this configuration), and the losses rise for a short time.

50Hz makes this circuit easier, because you can do most of your analysis in simple DC domain, and ignore switching time as a delay element.

The amount of safety margin they need to leave depends on their error margin: resistor tolerances, and worst case input offset voltages of the comparators. They recommend minimizing the resistor tolerances down to even 0.1% - I would also recommend choosing a comparator based on worst case input offset voltage + worst case input bias current times Rinput. I don't know how important this optimization is, you'd need to actually calculate it over the worst case variations to find out.

I did something similar when I had to design a discrete efuse circuit (which monitors voltage drop over a MOSFET pair); integrated parts were unsuitable. I made an Excel spreadsheet where I collected all the worst case minimum, maximum values of all components (beware: it can be a combination of parameters, like input offset V and bias I for a comparator, or tolerance + tempco for a resistor), and calculated the voltage thresholds at the combinations of min, max. The end result was that my 50A nominal e-fuse switches off between around 35A and 80A; such a high span was to be expected. The lab prototype performs close to that 50A, but if you want to make a robust product over unit and environmental variation, you need to look at the full range.
 
The following users thanked this post: Atom

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: PCB Desing, power rectifier advice
« Reply #6 on: February 28, 2019, 07:02:13 pm »
Your options are:
- Pay the measly five bucks for the controller,
- Slap on a bigger heatsink and fan (easily $20 in parts, plus labor),
- Engineer something yourself that might be cheaper in parts cost, but takes exponentially more labor to realize.  Is your time really discounted to nearly zero?*

*Now, if you're treating it as a learning exercise, it may very well be discounted all the way to negative, i.e., you're willing to spend time on it, regardless of the eventual outcome.  That's a different matter. :)

Give or take forum advice, but again, you get what you pay for.

From the high level view -- it should be no surprise that specialty chips like that exist.  They're not widely applicable, but they do one job very well, and you will be hard pressed to do any better, even given a lot of time to come up with alternatives!

That said, I am aware of one approach that uses saturable reactors to control AC gate drives.  Commutation is passive (determined by load current), very dumb and reliable.  This is the guy, but the links have since deceased:
https://groups.google.com/forum/#!original/sci.electronics.design/K3ybRPxdGkQ/YnIM3ipNBAAJ
Mind, that's another not-uncommon tradeoff: replacing semiconductors (e.g., bridge controller) with custom wound transformers.  No free lunch. ;)

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 
The following users thanked this post: Atom

Offline AtomTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: it
Re: PCB Desing, power rectifier advice
« Reply #7 on: March 04, 2019, 12:11:41 am »
So i managed to make everything fit, in the end i've gone for the schottky with a little heatsink, and by doing some calculations the temperatures would stay around 70 deegres, that is totally fine, mainly because the lt4320 would cost 7 euros.
so here there is the pcb and the 3d render let me know if there is something that shouldn't be done pcb wise, the pcb isnt totally finished , i need to squeeze othe parts on the pcb. any suggestion on the connector for taking out the power from the board ? i wanted to use a 2*11 2.54 pitch angled header, so i would have 10 pins for the + 10 pins for the - and the 2 remaining pins for control of the mosfet. the schematic is the same as the one on the first post.
« Last Edit: March 04, 2019, 12:18:12 am by Atom »
 

Offline AtomTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: it
Re: PCB Desing, power rectifier advice
« Reply #8 on: March 04, 2019, 12:13:05 am »
After i have finalized the bottom desing i'll mirror it to the top and add a tons of vias for extra current capability, is this a good idea?
also the dioeds will have a little of airflow coming from the fan of the main pass tranasistors.
i noticed a problem, here there are the correct pcbs
« Last Edit: March 04, 2019, 12:17:51 am by Atom »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf