Hi all,
I'm keeping myself entertained these days with a NAD 2200 project (on a preliminary observation, I like the quality at which it's put together). At turn on, Protection and Overload lights come on.
Upon innards inspection, I see a scorched area on the right channel - pictures enclosed - which at first glance made me think Q303 has shorted (which would toast R309 and the 50V rated C309). I initially only replaced Q323 (used KSA992) and Q307 (used 2N5401), plus R309, R323, and C309. Some of these were carbonized/melted, or shattered (one of the transistors). My running hypothesis was that Q303 has shorted B-C and exposed C309, R209 (and Q307) to about 100V and torched them. But for some reason I didn't replace Q303 at this step. I fired it up, and got R309 burned up again. I have a 100V rated poly cap for C309, so I think it should be fine.
Realized my lack of diligence and got Q303 replaced (used 2N5550), but lost another R309 upon firing up. I am assuming whatever happened here (Q303 measures dead short BE, and low resistance CE - acts as a junction) took down more parts, and so somehow I still have a high drop over R309.
My conundrum is that the catastrophic failure doesn't really allow me to analyze the stages and figure it out. My plan is to start pulling legs of parts around there to check them (especially the diodes all around there), which would be all offline work.
Thank you for thinking on this with me.
I'm keeping myself entertained these days with a NAD 2200 project (on a preliminary observation, I like the quality at which it's put together). At turn on, Protection and Overload lights come on.
Upon innards inspection, I see a scorched area on the right channel - pictures enclosed - which at first glance made me think Q303 has shorted (which would toast R309 and the 50V rated C309). I initially only replaced Q323 (used KSA992) and Q307 (used 2N5401), plus R309, R323, and C309. Some of these were carbonized/melted, or shattered (one of the transistors). My running hypothesis was that Q303 has shorted B-C and exposed C309, R209 (and Q307) to about 100V and torched them. But for some reason I didn't replace Q303 at this step. I fired it up, and got R309 burned up again. I have a 100V rated poly cap for C309, so I think it should be fine.
Realized my lack of diligence and got Q303 replaced (used 2N5550), but lost another R309 upon firing up. I am assuming whatever happened here (Q303 measures dead short BE, and low resistance CE - acts as a junction) took down more parts, and so somehow I still have a high drop over R309.
My conundrum is that the catastrophic failure doesn't really allow me to analyze the stages and figure it out. My plan is to start pulling legs of parts around there to check them (especially the diodes all around there), which would be all offline work.
Thank you for thinking on this with me.
Attachments
Last edited:
Its just a matter of testing every component before changing anything to save wiping out new components. It doesnt look like a massive amount to test.
I have got away with transistor testing in circuit.
You can work out expected resistor values from other components around them in circuit diagram. Resistors tend to o/c anyway. Caps tend to go short.
Debugging amps can be fun. Someone sold me a faulty Maplin amp that took me hours to fix. In the end someone had been at it before me and put in a npn instead of a pnp transistor ! The number had rubbed off and it of course diode checked ok.
I didnt think of checking correct diode polarity.
I have got away with transistor testing in circuit.
You can work out expected resistor values from other components around them in circuit diagram. Resistors tend to o/c anyway. Caps tend to go short.
Debugging amps can be fun. Someone sold me a faulty Maplin amp that took me hours to fix. In the end someone had been at it before me and put in a npn instead of a pnp transistor ! The number had rubbed off and it of course diode checked ok.
I didnt think of checking correct diode polarity.
Add some series resistors between the PSU and the amp board to limit the current - then you have a chance of measuring voltages. something like 500 -- 1k 2W resistors in each rail might be a starting point.
Since that amp can handle 2 ohm loads the PSU is quite capable, unfettered, of destroying small components in milliseconds.
I have the luxury of owning a thermal camera which is ace for commissioning circuits as you can spot unexpected rapid heating of components well before the tell take smell or smoke. Well worth it for this kind of problem.
Since that amp can handle 2 ohm loads the PSU is quite capable, unfettered, of destroying small components in milliseconds.
I have the luxury of owning a thermal camera which is ace for commissioning circuits as you can spot unexpected rapid heating of components well before the tell take smell or smoke. Well worth it for this kind of problem.
I have fixed a number of NAD2200's with very similar problems to this. You mention is is the right channel, yet Q303 is on the left channel- so the numbers i will use are the left channel. The -ve rail seems to fail at this point as per your picture. The main culprits often are:
diodes- ZD301/ZD303, D309/D311
Caps- it is wise to replace them all at some point
Transistors-Q305/ Q307, Q301/Q303, Q333 and Q335- check all the output transistors with a multimeter and look for shorts
Resistors- R375,373,307,309, 333,361, 359 are prone to failure- especially R333 and R361
One technique, (learned from the good folks here), is to connect the collectors of Q301 and Q303 which will cold bias the amp, and turn off the output stage- this helped me sort out a 2200 that was pulling too much current.
Check the voltages from the PS- the ribbon cables can be removed, but be careful as if you insert them back the caps will still have a charge and sparks will occur.
If the overload detector is always on then check Q214 and Q338 and associated components
I have attached a slide which shows what each section is doing- try to break the diagnosis into up to Q212 and then following on from this.
diodes- ZD301/ZD303, D309/D311
Caps- it is wise to replace them all at some point
Transistors-Q305/ Q307, Q301/Q303, Q333 and Q335- check all the output transistors with a multimeter and look for shorts
Resistors- R375,373,307,309, 333,361, 359 are prone to failure- especially R333 and R361
One technique, (learned from the good folks here), is to connect the collectors of Q301 and Q303 which will cold bias the amp, and turn off the output stage- this helped me sort out a 2200 that was pulling too much current.
Check the voltages from the PS- the ribbon cables can be removed, but be careful as if you insert them back the caps will still have a charge and sparks will occur.
If the overload detector is always on then check Q214 and Q338 and associated components
I have attached a slide which shows what each section is doing- try to break the diagnosis into up to Q212 and then following on from this.
Thank you all, and pwdiya12 - pretty close on target there. BTW - yes, all this is happening on the L channel, you are correct.
Dove further in and there's more toasty parts than I initially estimated. Dead were R325, R327, R363, R309 (already mentioned), ZD303, and - until my desoldering station had a mishap, and, as if that wasn't enough, my soldering station's handheld disintegrated on me tonight - I also figured Q315 is busted (CE short). For these 2SA1302/2SC3281s, my initial research pointed to the MJW1302A/MJW3281A, which seem drop-in replacements - can anyone confirm this?
Thank you all.
Dove further in and there's more toasty parts than I initially estimated. Dead were R325, R327, R363, R309 (already mentioned), ZD303, and - until my desoldering station had a mishap, and, as if that wasn't enough, my soldering station's handheld disintegrated on me tonight - I also figured Q315 is busted (CE short). For these 2SA1302/2SC3281s, my initial research pointed to the MJW1302A/MJW3281A, which seem drop-in replacements - can anyone confirm this?
Thank you all.
2SA1302 =MJW1302A --yes sight difference in build otherwise spec.looks the same --ditto MJW3281A.
Most hobby soldering stations are not up to industrial standards and don't like continuous use, okay for a while .
Watch out ! --hobby stations are built to a price , you can have three or more versions having the same outer case but internals in cheap versions are rubbish and --yes I can prove that personally.
If you can afford it buy a commercial station you don't need the full blown versions just the basic models are better than the general public types.
Most hobby soldering stations are not up to industrial standards and don't like continuous use, okay for a while .
Watch out ! --hobby stations are built to a price , you can have three or more versions having the same outer case but internals in cheap versions are rubbish and --yes I can prove that personally.
If you can afford it buy a commercial station you don't need the full blown versions just the basic models are better than the general public types.
Duncan - frankly, I could use some pinpointed recommendations, or hands-on experience, etc. I am in the US, so that's a factor on what regards sourcing (AMZ, ePay, etc.). I've been abusing this Radioshack station for years now, and it's been slowly coming apart (I am surprised it lasted this long - an Atten built, apparently, whatever that means). In think tonight was its final act (I reassembled it a little while back, but it didn't feel it comes back together very well). I am at the point where I may well benefit of a used industrial or semi-something (pro?) new station, given the utilization I am giving it (though that may decrease as things go back to normal).
BUT: maybe age, maybe utilization, maybe frustration - I think it's time for me I go for something reliable in replacing this (a bit more $ though). My H. desoldering station was a similar pursuit, though I am a bit disappointed with its performance and reliability this far.
If you dont want to spend too much money Rax have a look at these entry models -
Weller vs Hakko: The Best Soldering Irons & Stations - VSearch
Both should be easily available in the USA --a comment from Weller-
What Type of Soldering Iron Do I Need? | Weller Tools
And Hakko USA-
American Hakko Products, Inc. Home of superior quality soldering, desoldering, rework, BGA, and fume extraction systems and hand tools for the electronics manufacturing industry.
Weller vs Hakko: The Best Soldering Irons & Stations - VSearch
Both should be easily available in the USA --a comment from Weller-
What Type of Soldering Iron Do I Need? | Weller Tools
And Hakko USA-
American Hakko Products, Inc. Home of superior quality soldering, desoldering, rework, BGA, and fume extraction systems and hand tools for the electronics manufacturing industry.
On the tools - the desoldering tool I have is a Hakko (FR-301). Pretty solid (I have one of the cheap stations too, one pretty well reviewed, and the suction on the Hakko is in a different magnitude). I've been disappointed to get a clogged tip though, after just 3-4 months of use. This isn't daily, heavy/professional use, though may be a bit heavier than the usual diyer's.
But definitely, Hakko is in my target, and I've also been looking at Pace and maybe even JBC or Metcal, though, as you get there, they disappear quickly from "view" ($). I've more recently became of the mind that with tools, one should spend "more" than they afford. Buy used, NIB, etc. But a good, reliable tool, is pretty much priceless.
But definitely, Hakko is in my target, and I've also been looking at Pace and maybe even JBC or Metcal, though, as you get there, they disappear quickly from "view" ($). I've more recently became of the mind that with tools, one should spend "more" than they afford. Buy used, NIB, etc. But a good, reliable tool, is pretty much priceless.
Further on the amp work.
Boy, was I naive to think replacing Q303 and a couple other parts was going to fix this. The more I investigate, the more I find failed parts - the L channel is a mess.
First, from the parts I just put in, the only fail I could identify is Q323 (KSA992) - not catastrophic, but the stress put on it made junctions measure too low in resistance to be good (don't recall which; one measurement was off). The others seem to have survived OK.
But part of the original deflagration took offline most power transistors on that channel (Q315, Q319 have CE shorts, and Q313 has basically all legs welded together internally). Still checking parts, but at this point I have the replacement MJWs coming from my usual source. My current plan is to detect as much of the obvious fails as possible and then reassess whether I should trust that it wouldn't go into major short mode if I kick it on. Chances are that, replacing the failed power transistors would get protection circuitry out of commission and then repair would be smoother. I think these were my major culprits. I started recapping selectively - where too handy to pass on - but upon getting it to run OK, I will definitely recap the whole thing.
In a nutshell, somebody may have used this for actual welding, which we all know it can do
Boy, was I naive to think replacing Q303 and a couple other parts was going to fix this. The more I investigate, the more I find failed parts - the L channel is a mess.
First, from the parts I just put in, the only fail I could identify is Q323 (KSA992) - not catastrophic, but the stress put on it made junctions measure too low in resistance to be good (don't recall which; one measurement was off). The others seem to have survived OK.
But part of the original deflagration took offline most power transistors on that channel (Q315, Q319 have CE shorts, and Q313 has basically all legs welded together internally). Still checking parts, but at this point I have the replacement MJWs coming from my usual source. My current plan is to detect as much of the obvious fails as possible and then reassess whether I should trust that it wouldn't go into major short mode if I kick it on. Chances are that, replacing the failed power transistors would get protection circuitry out of commission and then repair would be smoother. I think these were my major culprits. I started recapping selectively - where too handy to pass on - but upon getting it to run OK, I will definitely recap the whole thing.
In a nutshell, somebody may have used this for actual welding, which we all know it can do
Last edited:
(I hope I'm not throwing off the tread with this OT conversation on tools... If yes, moderators please alert me)
Duncan,
I've provisionally narrowed it down to the Hakko FX-951 and Pace ADS200. From what I'm reading the Pace has a much better build (though I can't see them reviewed side by side, so take with a grain of salt). Though EEVblog puts some nuance in that.
Being cartridge irons, a consideration is cost of cartridges/tips. It seems the Pace makes them really cheap, which is very reassuring (being both afforable and genuine). For the Hakko, I find a plethora of generic ones on eBay, which only reassures me on $, but not quality.
Although I mentioned Hakko USA they are also made in Japan so that might be a reason for the bigger range on the web.
The Pace is also available in the UK and outputs -120W whereas the Hakko is a bit lower .
I have never spent time investigating Pace but what always reassures me in cases like this is it supplies to the US/UK government.
I buy ex.UK government stock usually ex.military test equipment sold by a government contractor of good reputation ( equipment tested before dispatch ) .
I like the construction of the Pace it looks industrial and solid .
It took me some years in earlier life to realize you only get what you pay for so I saved up for good quality equipment -HP/Tektronix/etc even my Variac I bought from the UK government contractor is a USA Zenith make and "built like a tank ".
The Hakko is 75W for the FX-951 and is actually dearer than the Pace --tips are not supplied.
Here is a comparison of top makes including Pace -
The best soldering stations - review: Hakko, Metcal, Pace, Ersa, JBC, Edsyn, Xytronics, Weller, Den-On, Goot..
Pace was a UK company originally (set top boxes but supplied to many big UK/USA companies ) now owned by Arris -parent- CommScope (USA conglomerate ) no surprise Farnell UK is now owned by a big USA company as well.
The Pace is also available in the UK and outputs -120W whereas the Hakko is a bit lower .
I have never spent time investigating Pace but what always reassures me in cases like this is it supplies to the US/UK government.
I buy ex.UK government stock usually ex.military test equipment sold by a government contractor of good reputation ( equipment tested before dispatch ) .
I like the construction of the Pace it looks industrial and solid .
It took me some years in earlier life to realize you only get what you pay for so I saved up for good quality equipment -HP/Tektronix/etc even my Variac I bought from the UK government contractor is a USA Zenith make and "built like a tank ".
The Hakko is 75W for the FX-951 and is actually dearer than the Pace --tips are not supplied.
Here is a comparison of top makes including Pace -
The best soldering stations - review: Hakko, Metcal, Pace, Ersa, JBC, Edsyn, Xytronics, Weller, Den-On, Goot..
Pace was a UK company originally (set top boxes but supplied to many big UK/USA companies ) now owned by Arris -parent- CommScope (USA conglomerate ) no surprise Farnell UK is now owned by a big USA company as well.
(still waiting for parts... some of the largest US shippers seem to be in complete disarray, btw. Never seen anything like it)
Edsyn has become a real contender. It's hard to find good reviews/comparisons - thank you, Duncan, or the above, btw - and so what I'm still unsure of is if I want to be focusing on cartridge-based irons, or just very reliable ones. The latter focus would include Edsyn, as I don't think they make cartridge-based irons. Edsyns, which seem to last forever, seem pretty plentiful and affordable on the used market. For a soldering station, finding used would be the perfect route for cost/return ratio.
Edsyn has become a real contender. It's hard to find good reviews/comparisons - thank you, Duncan, or the above, btw - and so what I'm still unsure of is if I want to be focusing on cartridge-based irons, or just very reliable ones. The latter focus would include Edsyn, as I don't think they make cartridge-based irons. Edsyns, which seem to last forever, seem pretty plentiful and affordable on the used market. For a soldering station, finding used would be the perfect route for cost/return ratio.