I can use a scope just fine. Im just saying, there isnt a lot to grab on to, on these SMD ic's. To be honest, all the DC voltages on the IC seem to check out fine compared to channel B. Not looked at the ac waveforms yet. Have put it back on the shelf for now.
You should know how to scope components like that then? You don't grab the tiny pins, you either probe the next component connected to the pins with the probe tip or you solder a little piece of rigid copper wire to the next component pins ... e.g. you don't connect to the gate driver LO pin on Ch A ... you would grab the leg of R276 (if through hole) or probe with the tip on the LO side of R276 if it's SMD ... or if you want to grab it solder a little piece of rigid copper wire to the SMD component and grab that.
If you want to scope the outputs of these gate drivers I would strongly advise reading up on making floating measurements or your scope could be permanently damaged very quickly. DO NOT just connect the scope ground to COM and probe about until you've read up on and understand what that means then I'll help you with where to connect.
What scope do you have? Don't give up yet, I don't think the OP has a scope so you may be able to help him too.
Give more details of what you have done already ... did you replace both 4227 fets on the problem channel?
Have you tested all components between the driver and fets?
Did you do the IC resistance tests and compare values between channels?
I mainly work on vintage stuff so Im not very often working with SMD.
The 4227s are referenced to ground through the heatsink and I am in UK - our chassis ground goes to earth, not floating like a variac. Plus our houses are fitted with great tripswitches 😆 so I would have thought grounding the scope probe to chassis to be ok although I haven't looked at the schematic yet.
also yes I replaced one 4227 on channel A, the right hand side one, it was dead short in all directions.
Ive tested quite a few components on channel A, I feel like ive tested everything up and downstream of the blown 4227 but yeah I really need to look at the schematic before I look any further.
I hastily bought one of those gate driver IC's when I saw this thread although kind of regretting spending the £10 now .. as it looks generally healthy although have no idea for sure
I appreciate the speedy reply by the way
The 4227s are referenced to ground through the heatsink and I am in UK - our chassis ground goes to earth, not floating like a variac. Plus our houses are fitted with great tripswitches 😆 so I would have thought grounding the scope probe to chassis to be ok although I haven't looked at the schematic yet.
also yes I replaced one 4227 on channel A, the right hand side one, it was dead short in all directions.
Ive tested quite a few components on channel A, I feel like ive tested everything up and downstream of the blown 4227 but yeah I really need to look at the schematic before I look any further.
I hastily bought one of those gate driver IC's when I saw this thread although kind of regretting spending the £10 now .. as it looks generally healthy although have no idea for sure
I appreciate the speedy reply by the way
Eh.. I don't think so... so your saying on your amp the fets are grounded through the heatsink. That would ground the fet drains???
The OP (Thridel) said the fets were insulated from the heatsink.
Are we talking about the same amp here?
So you've not even looked at the schematic? It's always the 1st thing I do... 🤔
Attached is a copy for you:
oh thanks for that. actually i didnt think behringer released it - they didnt release the SM of some other stuff I worked on. I try not to spend too much time on a repair..and I try not to work on behringer stuff at all if I dont have to 🤣 like I say, I work on a lot of older stuff - theyre mostly very simple discrete circuits. I was going from memory of last time I had it open.
now, I got it back on the bench.
yes you're correct, Looking at the schem I see only VSS and some bypass caps ref'd to GND. Source's are at -78v
I just spotted D49 and D50 are labeled differently to my board...the unit I'm working on is "NX3000-D", i wonder if its slightly different. the same zeners on the other channel are also labeled differently.
Either way, because every other component seemed to check out ok, I swapped the ?thought it was good? output 4227 from channel A left hand side (T23 on my board, pictured) to the same position on channel B left hand side, reassembled for testing .. and voila, I get protect light on both sides, as opposed to just protect flashing yet still functioning channel A...Looking at the schem, my guess is that burst of current to -78v that fried all pins on T22 (on my picture) also maimed T23 (my picture) but didnt quite short it, so it still tests ok.
I'm going to replace this tranny i have now swapped to channel B and report back. Perhaps I should just replace all 3 old devices...That 4227 was quite expensive though iirc
edit: Im going to go and delete my mention of the outputs being ref'd to ground , I dont want to mislead anyone. I hate it when I read a forum and someone is completely wrong!
edit ii : looks like its too old to delete...oh well.... I dont write in forums very often, can you tell ? lol
now, I got it back on the bench.
yes you're correct, Looking at the schem I see only VSS and some bypass caps ref'd to GND. Source's are at -78v
I just spotted D49 and D50 are labeled differently to my board...the unit I'm working on is "NX3000-D", i wonder if its slightly different. the same zeners on the other channel are also labeled differently.
Either way, because every other component seemed to check out ok, I swapped the ?thought it was good? output 4227 from channel A left hand side (T23 on my board, pictured) to the same position on channel B left hand side, reassembled for testing .. and voila, I get protect light on both sides, as opposed to just protect flashing yet still functioning channel A...Looking at the schem, my guess is that burst of current to -78v that fried all pins on T22 (on my picture) also maimed T23 (my picture) but didnt quite short it, so it still tests ok.
I'm going to replace this tranny i have now swapped to channel B and report back. Perhaps I should just replace all 3 old devices...That 4227 was quite expensive though iirc
edit: Im going to go and delete my mention of the outputs being ref'd to ground , I dont want to mislead anyone. I hate it when I read a forum and someone is completely wrong!
edit ii : looks like its too old to delete...oh well.... I dont write in forums very often, can you tell ? lol
Attachments
Last edited:
Its the same basic amp circuit as the "D", yes some components will be differently numbered but the part of the circuit in question is the same.
Yes I have seen fets test as not shorted or open but still not work properly, I have also had to replace the gate driver before.
You want to try swapping the two known good fets to the previous bad channel, i.e. a complete swap and see if the problem completely swaps sides.
Be careful ESD wise these fets are very sensitive to static, so take precautions.
Yes I have seen fets test as not shorted or open but still not work properly, I have also had to replace the gate driver before.
You want to try swapping the two known good fets to the previous bad channel, i.e. a complete swap and see if the problem completely swaps sides.
Be careful ESD wise these fets are very sensitive to static, so take precautions.
In fact take all 4 fets out ... then put only the two known good ones on channel A (I think that was your original working channel) ... leave channel B empty ...
Channel A should function ... again.
Then swap those two to channel B ... then we will see if it is just the fets or something else on channel B (or whatever was the original non functioning channel).
Be careful of ESD.
Channel A should function ... again.
Then swap those two to channel B ... then we will see if it is just the fets or something else on channel B (or whatever was the original non functioning channel).
Be careful of ESD.
I haven't had any time to measure yet but it's nice to see someone else having the same problem, I hope we can solve it together. 🙂
(It's not nice that you have a problem as well but you know what I mean) 😀
(It's not nice that you have a problem as well but you know what I mean) 😀
You could try taking out all fets then try the two good one's on the opposite channel Thridel.
Something else I just realised, is there a software reset for these amps?
Something else I just realised, is there a software reset for these amps?
I was wondering about the software and wether we need to load it to reset it somehow. i should hope and think not but you never know. The nx3000D has a DSP chip with fx and stuff, very digital...it does have an init. all function in the front panel, this didnt seem to do much although may be necessary after all electronics are solved...
I have moved on to other projects for now ill report back during the week folks
I have moved on to other projects for now ill report back during the week folks
I fixed some EV ZLX 15's a few months back & I think I had to reset the software for some reason. It was just a thought... a factory reset can't do any harm.. as I think the protection levels on these are software controlled so it may help. Did you try those two good fets in the other channel baruk?
Usually there is but I've never had one of these DSP models.
Have you tried swapping channels with your two known good output fets?
ok, I got mine running stable. In the end 3 of the original FETs were fine. The first replacement I bought and installed was either bad from the distributor/ counterfeit/ a bad batch/ or I subjected it to ESD, because right off the bat it worked but that side of the heatsink got very hot.
Serves me right for not testing it with a component tester before chucking it in.
I then installed more fresh ones, 3 in fact and just one original, the unit came fully out of protect and worked but heated up very fast and I had to cut power.
so i decided to backtrack and install the 3 originals .. I replaced the new one with another one I bought, and voila, everything came out of protection, fan stayed steady, and heatsink stayed cool. sent some signals in with the generator and all fine. phew. glad that is over, now to return all the components i didnt use.. 🤣
Serves me right for not testing it with a component tester before chucking it in.
I then installed more fresh ones, 3 in fact and just one original, the unit came fully out of protect and worked but heated up very fast and I had to cut power.
so i decided to backtrack and install the 3 originals .. I replaced the new one with another one I bought, and voila, everything came out of protection, fan stayed steady, and heatsink stayed cool. sent some signals in with the generator and all fine. phew. glad that is over, now to return all the components i didnt use.. 🤣
No I haven't had any time to look at it more yet.
Interesting! Can I check my fets before I put them in with just a multimeter and be sure if they are ok or do I need this component tester?
Great stuff, yeah be careful with ESD when working with these fets & IC's. Could have been any one of the other reasons too. Best to keep all paired fets on a channel picked from the same batch or you can even have 2x perfectly good fets getting hot because their characteristics don't match well 😱
I still have high hope for Thridel too... 🙂
You can only check for short or OC with meter. A component tester is better but still only gives a rough idea... best way for repair is replacement with same batch pairs and being very careful with ESD.. 🙂
I've told you how to test already... take the pair of good fets from the good side and try them in the bad side. If that makes that channel OK then you know what the problem is.. 🙂
Thridel, I have a £20 cheap chinese Component tester from ebay, very useful as it gives pinout. and some spec info.
and on a side note - the FETs I bought as replacement had a different VT value - a whole volt difference -3.5v instead of 4.5v
and on a side note - the FETs I bought as replacement had a different VT value - a whole volt difference -3.5v instead of 4.5v
Yeah that kind of voltage threshold difference should be avoided and will have been why it was getting hot,.... so always best to buy in pairs from same batch... 🙂
I use a cheap transistor tester too that I got in kit form for about £10 then upgraded some higher precision parts and used the developed firmware from eevblog...
What type of tester do you use check12? I'm interested.
Also yes, I generally replace in pairs too, but in this case the batch was very important. To be honest I think im going to stop buying on ebay. The seller is claiming they are definetely genuine but Im not sure. I now noticed that Farnell have them, serves me right for not checking in the first place..Because I work on older stuff, there are generally none on the parts suppliers sites.
Also yes, I generally replace in pairs too, but in this case the batch was very important. To be honest I think im going to stop buying on ebay. The seller is claiming they are definetely genuine but Im not sure. I now noticed that Farnell have them, serves me right for not checking in the first place..Because I work on older stuff, there are generally none on the parts suppliers sites.