10 March 2021 at 8:54 #10283
Following a discussion that developed in the comments to one of my articles, I created this forum to have an easier view of all the interventions in chronological order.10 March 2021 at 8:56 #10285
COPY OF THE ORIGINAL TEXT
high power with IR2153D
good evening everyone. I'm sorry for my poor Italian, my present. my name is Petro Ruta. reading Mr.. Amilcare, I had an idea to reuse components of a full-bridge smps, that by now he had lost the control card, built with family circa l6598. These smps are certainly Chinese-made, some components have numbers crossed out, but I can provide approx 3 kw, unfortunately a long time ago I needed spare parts to fix other smps and I took some components from them.
now I have bought some class of amplifier boards that use high voltages +/- 70volt and require a lot of current about 10ampere per branch. it is iraudamp2000 card with two channels with irs2092 ic and irfp4227 mosfet. very very powerful, and they can also work as a bridge. the seller declares about 350watt rms with load 8 ohm per channel and bridge approx 1200, as he declares 700 watts per channel on load 4 ohm e 2000 bridge. I have tested with a 8ohm load bridged on a woofer b&c 18pzb100 is the result is fabulous. so I decided to complete the work by building the subwoofer. now that i'm done i found a problem, the first is to find a toroidal transformer of at least 2kw with outputs 50+50 and auxiliaries for crossovers and accessories is very rare and very expensive. looking in the cellar i found these old 3kw smps boards but i don't remember the heat of the components i took off so it became impossible to make it work. this article by Mr. Amilcare was inspirational, I decided to dismantle old smps and build a new one by reusing the new pieces, but having no experience, I ask you for help. this is a schematic of a smps that might suit my case, but I need help as I don't know very well how a smps works and how to find the right compromise with recovered pieces10 March 2021 at 8:58 #1028710 March 2021 at 9:09 #10289
For the moment, I'll summarize the simple improvements to prevent MOSFETs from overheating
R16 and R18 both from 100 ohm should be brought to 1K so as to make the two transistors intervene quickly to quickly saturate the MOSFETs.
R22 you can omit it because the MOSFETs once triggered cease to draw current, the intent is to make them trigger as quickly as possible.
In the transition between on and off and vice versa the MOSFET dissipates a large energy, the shorter this time, the lower the average energy dissipated by the MOSFET in normal operation.
Resistor R22 by limiting the current at the gate lengthens the time in which the MOSFET dissipates energy.10 March 2021 at 14:44 #10299
today proposes a new modification to separate ac dc module for ir2151d from pilot transistors
10 March 2021 at 22:01 #10301stony stoneParticipant
This modification makes it possible to power the transistor drivers from a high voltage power line. Such small ACDC module only works for ir2151. But at this point you can also replace it with zener diode some resistor and a transistor11 March 2021 at 7:55 #10307
I liked the first pattern much more, simple linear and efficient.
For the second I have doubts that at the start you have enough energy to start switching the MOSFET.12 March 2021 at 9:31 #10309stony stoneParticipant
ok I proceed with the first scheme without r22 and with r16 r18 from 1k.
I compose everything on orcad and attach the layout
- You must be logged in to reply to this topic.