I have one of these boards and using it as a subwoofer amp.I´m using it with the 2x40 AC,625VA toroidal transformer.Problem is it use only 15W power.There comes however somehow sub frequency through from it,if using preamp´s and Ht receiver´s gain high.For the amplifier there goes to 2x40AC correctly that´s what i checked.
Any thoughts?
Any thoughts?
We have 3 of these boards. all had problems after some time of minimal use at full frequency. they are low powered and when driven into lower than 4Ω they die (rather violently). we have moved back to Class A/B with higher power and SMPS for supply. seems to be working fine. if anyone has solution to the IRAUD350 and similar, we would like to know. thanks to all.
check out the iraud200 thread .. that's one solution .. the iraud350 could benefit greatly from adding the standard delay cap/resistor into the feedback loop because it's feeding back too quickly .. causes it to overheat. But the iraud350 does not show up with irfb4227 transistors that I recall seeing. Using those transistors requires some adjustments to the circuit that the iraud200 design almost got right ... the iraud200 thread shows how I discovered the weaknesses in the iraud200 circuit and corrected it. I now have two iraud200 boards using irfb4227 and ready for 1kw per channel into 2 ohms
There is a second option I believe .. ljm put out a board that uses irfb4227 transistors .. it buffers them with mosfet drivers while the iraud200 board does not. If you want to do 2 ohm loads, you have to have the irfb4227 transistors. They in turn require some adjustments on the drive circuit, and a heat sink on the 2092 chip .. the iraud200 board uses no gate resistor ... it's direct drive .. however, they took out the delay cap and drive the transistors too fast .. they also put no hf dampening at the output, which allows parasitics to get going ... circuit corrections available
The iraud200 comes in two formats .. one with the irfb4227 transistors .. one with something else .. be sure to get the right transistors for 2 ohm load.
The iraud200 comes in two formats .. one with the irfb4227 transistors .. one with something else .. be sure to get the right transistors for 2 ohm load.
the iraud350? I'm not totally certain .. I never made one of those perform. I tried installing irfB4227 .. (difficult to accomplish) .. I assure you, you can't just do that. A good experiment might be to install the irfB4227, installing the delay cap/resistor pair as described in the iraud200 thread into the feedback loop .. If the 350 has gate resistors, you would have to bypass those with wires .. but even if you did all that, making sure the 2092 chip has a heat sink .. the output inductor would not be prepared for 2 ohm loads, and output dampening may need to be adjusted/installed. Those inductors are not cheap... my best opinion ... forget the iraud350 .. it isn't reasonably convertible to 2 ohm stability even though I think it would be possible. I don't think there's even space for a larger inductor on the board. The end result would be scraggly looking if you could get it to work for 2 ohms.
we would be happy if they worked at 4Ω, but the 2092 blew up and set fire to the speaker. we had been running the amp on dummy 4Ω load for testing with scope. everything looked fine, but when higher power was applied - disaster. the others have had different issues, mostly QC type stuff. would like suggestions for reliable Class D amp and SMPS to power. cost is a consideration but quality and weight are more important to us. thanks.
well, you can get these to run for 4 ohm reasonably, and the solution is to insert the delay cap/resistor as well as the dampener resistor/cap pair on the output like I did for the iraud200. The problem with the iraud350 is that they idiotically decided to get 80 khz by speeding up the 2092 by taking out the delay cap/filter in the feedback loop. This causes the 2092 chip to overheat, even with a heat sink on it. .. in fact, it can overheat the 2092 chip and output transistors, regardless of the load or signal/power usage. The iraud350 I had worked fine for 4 ohms for a while, then it slowly got hotter hotter hotter, which I attribute to slight shifts allowing parasitics to arise .. that's why I suggest also including the RC dampener circuit on the output like I did with the iraud200. I have not looked at the iraud350 board to find its feedback loop location, don't have the boards any more, but there will be a line leading back towards the 2092 from the output stage before the inductor. It will have a high resistance resistor in it like 120k or more. The delay cap/resistor come between the output stage and the high resistor. I don't recall the proper values, check the iraud200 thread for that. Put the small delay resistor in series with the large resistor, connecting it on the other side to the output stage. Where the two resistors meet, connect on the cap. Connect the other side of the cap to ground somewhere. This little add sucks in a tiny bit of charge when the output stage shifts, It takes maybe some tenths of a microsecond for this to happen, causing the feedback loop to slow down. When the feedback loop slows down, the 2092 chip will only have to charge/discharge the transistor gates half as often or so, causing a dramatic reduction in heating for the 2092 and output transistors.
There are soooo many of these amplifiers its difficult to identify specific types (mine is fitted with IRFB23N20D's), I bought one and have been using it in a sub-woofer (JBL) driving 4R at ~300Wrms daily for 2 years powered by a huge 35-0-35V transformer BUT I had to modify it to begin with, the worst problem is the idiots had attached the 2092 heatsink with silicon, a well known insulator both thermal and electrical and may well be your problem, on top of that the heatsink was squeezed between the two tall caps, not the best way to assure them a long life! anyway basically I re-attached the heatsink with thermally conductive double sided tape and de-soldered/bent the big caps away stabalising the assembly with a bit of silicon. Hope this helps
My iraud350 board that worked (one died instantly) kept working until it somehow got caught in a hf battle with another amplifier circuit and fried. This brings up the other way the iraud350 can die. These boards are very hf active. Anything that fouls up their hf activity can kill the board. Perfect candidate? Another board just like it connected to the same power supply. The iraud200 thread also shows a solution to this problem. Put double ferrite beads on the ground line, and at least single ones on the power lines. This sucessfully protected my iraud200 boards from each other. These boards interfere with radio reception, so if you want to fix that, you will need to do a lot more shielding activity ..Make sure they are enclosed in metal, ferrite beads everywhere, and possibly extra small inductor (5 uH) after the ferrite beads on the output to quiet down the output line
colossusfan, we purchased 2 huge coroidal transformers to power each board. the xformer has dual 50vac 10A outputs and dual 15vac 0.3A outputs and is made by ANTEK. we use one of the 15vac to a speaker protect board, which works great. the boards you have sound like they may be like ours. they are RED and have YUANJING on the silkscreen side. they appear to be simple boards that roughly follow the IR design notes. I have one on the bench now and am tracing out the circuit design.There are soooo many of these amplifiers its difficult to identify specific types (mine is fitted with IRFB23N20D's), I bought one and have been using it in a sub-woofer (JBL) driving 4R at ~300Wrms daily for 2 years powered by a huge 35-0-35V transformer BUT I had to modify it to begin with, the worst problem is the idiots had attached the 2092 heatsink with silicon, a well known insulator both thermal and electrical and may well be your problem, on top of that the heatsink was squeezed between the two tall caps, not the best way to assure them a long life! anyway basically I re-attached the heatsink with thermally conductive double sided tape and de-soldered/bent the big caps away stabalising the assembly with a bit of silicon. Hope this helps
mgshightech responded with ways to calm this bd. down and make it usable and we intend to try these mods to see if this board can be utilized or trashed. cannot spend too much more time with this. we are using Class A/B in our PA cabinets now and they are working well but generate some heat (fans were the answer; efficiency out the window ;-) ). we also replaced other Chinese boards with EV amplifiers that are bi-amped and they are GREAT!!!!
thanks again for the responses...hjp
sure electronics makes quality boards at a higher price. They can be found at mcm electronics, or parts express, and also some times on ebay. LJM claims his amp9 semi-clone works well. I have seen people complain about it. My two iraud200 boards can kick the **** out of 600 watt rms 2 ohm speakers. If I just run one channel so the power supply doesn't wimp out, I can push one of those drivers all around town and get it to smell voice coil in just maybe half a minute or less. The modded iraud200 boards are beastly. I added a lot of extra heat sink for fun and safety .. one on the irs2092, one on the output inductor, and a big one on the main sink. If you need a big heat sink, look for old antminers on ebay (bitcoin miners) you can get massive sinks for just a few dollars that way. I know, class D doesn't put out a lot of heat, but when you are passing 1 kw through 2 little to-220 sized transistors .. it may be a good idea to keep their temperature down because if their temperature rises, so does their on-resistance, possibly causing thermal runaway in a torrent of power.
the iraud200 boards are I believe configured for higher gain than the iraud350 .. I think they use a 200k feedback resistor .. it makes the feedback loop susceptible to noise, so if you wanted to knock down the noise, you could replace the 200k resistor with 120k and drive the board with a little stronger signal.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.