Replacement for IRFB4227

Jump to Latest
T

Thridel

Member
Joined 2023
#1
Hi.

I have a Behringer NX3000D with a blown output transistor (IRFB 4227). They seem to be impossible to find anywhere else than Ebay and i think they are just counterfeit parts.
Does anyone know a direct replacement i can use which i can source in Europe through Farnell, Elfa, Mouser or Digikey for example or does anyone of you have 2 spares lying around collecting dust? I can find a IRFB4227 in stock but it has other parameters such as Vds, rise and fall time etc. so I guess it won't work with that?

Thanks in advance!
 
#2
IRFB4227 (letters after this are just package designations) are genuine. If you are replacing with genuine parts you don't have to worry about transistor specifications or replacing with equivalents.

This place has them in stock in UK (usually good for genuine parts) ... you can contact them and ask first. You'll maybe need to replace the gate driver IC too.

https://www.cricklewoodelectronics....MIzau_6KOE_QIVXoBQBh1UqAEzEAQYBCABEgLhavD_BwE
 
#3
Thank you for the answer. I meant that the one on Ebay are probably counterfeits, not the type itself 🙂
I will write them and ask. Why would i need to replace the gate driver? Is it to match the new transistors or is it probably damaged as well?
 
#4
Thridel said:
Thank you for the answer. I meant that the one on Ebay are probably counterfeits, not the type itself 🙂
Click to expand...
LOL yes I realised that on re-reading what you wrote...

Thridel said:
Why would i need to replace the gate driver? Is it to match the new transistors or is it probably damaged as well?
Click to expand...
Well I've found the one's on the amp side to be quite resilient but there is the chance that it has been damaged too.

Also check the gate resistors and zener diodes etc. ... can't be too safe, I'm taking it you've still got one working side so the SMPS is still working with all secondary voltages..

The problem with Class D is that if you don't make sure all faults are fixed then the new transistor may not last very long.
 
#5
I have checked everything from the gate drivers to the transistors and nothing seems to be broken afaik, but I'm a noob so I'm not sure. I have already replaced the transistors with ones i bought at Ebay and the cannel works but it gets to hot in a couple of minutes so i think that they don't meet the specs on Rds(on) as genuine parts and thus get a lot warmer.
It could very well be something else but this is by best guess atm (as i said im a noob) I dont have access to a scope either so i cant check the....wave? Channel A works and Channel B measures the same voltages as Channel A at the MOSFET:s. They are 2 at each channel, so a total of 4 in the amp.
 
#6
Thridel said:
the cannel works but it gets to hot in a couple of minutes
Click to expand...
At least nothing has "blown" immediately (that's a good sign) and you've got channel A transistors to compare to ... if B is hotter in comparison then something is still wrong.

Or you haven't clamped or soldered the transistors to the heatsink properly depending on the package (TO-220 or D2PAK).
 
#7
It's a TO-220 type. I have checked and it's no continuity between the transistors and the heatsink so they should be properly insulated.

What do you say about the theory of the Rds(on) resistance being at fault?

If i buy replacement transistors, do i need to have a matched pair? I'm really not sure what that even means, does it make a huge difference?
Is there anything else i can check just with my multimeter?
I found the schematics online for the NX3000 type, they're pretty similar. In my case its the T48 and T49 that is replaced.
 
#8
Did you use new heatsink grease/compound?... Did you put it on both sides of the insulator pad between the transistor and heatsink. Did it have heatsink compound originally (some pads don't). You defo need to have it fixed to heatsink properly.

Transistors from same batch are good enough.

Might be fake transistor... Might be damaged gate driver.

How did you check the zener diodes?
 
#9
I didn't replace the compound between the heatsink and the silca pad but between the pad and transistor i used ordinary computer thermal compound, i guess that works as good? It is just between the pad and transistor, I was very careful not to smear it all over the place.

I measured the resistance as per YouTube and other sites recommendation and also compared the results with channel A and it was the same. I did not desolder them because I'm thinking as long as it measures the same as the good channel it's ok?!

Is there any way to measure the gate driver? I did measure voltage at hi side and low side and it was the same as channel A.

Thank you for bouncing ideas back and forth with me 🙂
 
#10
I think the simplest and best thing to do for your circumstance is to get two new 4227's with same batch code (matching code number under transistor type on component) from a reliable source, fit them carefully and take it from there.
 
#16
Bad news everyone!

I replaced the transistors with new ones bought at Cricklewood electronics but I'm sorry to say it is exactly as before. Getting really hot in a couple of minutes.🙁
I stumbled across this link where it says that it can be a problem with the output filters so I'm thinking maybe something went bang in that area when the short occurred? Any suggestions where I can look for further investigation? Which parts on the schematics IS the output filters?
I'm unable to find any more visual damage on components, and all the diodes I've measured seems fine. Maybe some blown caps that is not showing from the top side?

As I said, I'm out of ideas =/
 
#17
The output filter is the section after the fets and before the speaker.

Do you have oscilloscope? ... it's not going to be easy to get any further troubleshooting info with only multimeter.

How about some high res photo's of the board?

You could also try resistance checking (with amp off) the known good ChA gate driver IC & comparing the readings to the side getting hot, write down the values then compare of:

1) COM to LO
2) LO to VCC
3) COM to VCC
4) VB to HO
5) HO to VS
6) VB to VS

Could also diode test the buffer transistors (google if you don't know how) T34-T37, the fault is most likely a component between the driver IC and mosfets that you haven't tested properly. Most prone to damage are diodes, resistors & transistors. You can test the resistors, capacitors & standard diodes with a multimeter in appropriate modes and compare them to the good side, you cannot however test 15V & 3V Zener diodes with a multimeter but they are cheap and easily replaced.

Troubleshooting would be a lot easier with a scope ... 🙂
 
#19
Hey folks. So I have the exact same problem, exact same amp. I replaced IRFb4227 but upon boot I get a flashing protect light on the channels MOSFET i replaced...it goes away after 20 seconds or so and the channel functions correctly BUT runs hot, you can feel it on the heatsink if you move your finger either side. So you think the 16 pin mosfet driver IC is also bad? I looked it up and its a high voltage device so Im a bit nervous with the scope probe...I don't have the most steady hands. Don't want to short anything...
 
You must log in or register to reply here.
Top Bottom