Hello, newbie looking for some suggestion on resurrecting my old Class A/B amplifier after a catastrophic failure.
Here's the tale, so for the best of my luck when I was checking the bias mA on my Samson S700 amplifier for note them down as maintenance routine, somehow when I powered the amp the test alligator clip slides from the output emitter pin touching other pin or the nearest resistor and I quickly turned the unit off after a smelly big smoke cloud, this is the first time this silly thing happens to me tbh, really frustrating situation.
Now the amp channel 1 module has two failed transistors one pair 2SA1492 / 2SC3856, two small 1/2W burned base resistors, tree open/burned 0.47ohms 5W emitter resistors and the unobtanium 2SJ78 / 2SK215 lateral MOSFET's.😱
Since there is no 2SC3856 avail at Mouser, Newark nor Digi-Key and since eBay is full of fakes to bother with, I opted for all the pair 2SA1943N / 2SC5200N(2SA1386A/2SC3519A was pricey) as I've read some threads users using them as replacements and they are a tad better, also listed by the NTE as replacement too.
So I ordered all the parts, 6 pairs 2SA1943N / 2SC5200N and the silicon thermal pads, all the small base/emitter 1/2W resistors, all 12x 0.47ohms 5W emitter output resistors.
However I'm now stuck at trying to find 2 pairs 2SJ78 / 2SK215(or 2SJ79/2SK216) lateral mosfets from any reputable sources to no avail, and the ones I've found under the U.S. location are a bit pricey for being NOS parts if that's even the case.😒
All this frustration made me spent lots of time searching/compare for those laterals mosfets and I almost to execute an absurd experimentation, just replacing the 2SJ78 with -> ECX10P20 and the 2SK215 with -> ECX10N20, I know those laterals are way overkill for the role of pre-drivers but I could not find anything similar in the case TO-220 unfortunately.🤔
So the question is, after re-biasing the amp does this big laterals can work as drivers in place of the unobtanium/obsolete 2SJ78 / 2SK215, and what could be the drawbacks of using such large/overkill laterals instead the smaller with less input capacitance, I guess the amp may be prone to oscillation at high frequencies and/or lower amplifier gain? assuming this overkill laterals can work as drivers of course.
Also is there another/better 2SJ78/2SK215 replacement candidates with similar specs regardless of the gate/drain pin configuration, I'm up for options/experimentation.
Amp channel modules cleanup/preparation started:
Regards
Here's the tale, so for the best of my luck when I was checking the bias mA on my Samson S700 amplifier for note them down as maintenance routine, somehow when I powered the amp the test alligator clip slides from the output emitter pin touching other pin or the nearest resistor and I quickly turned the unit off after a smelly big smoke cloud, this is the first time this silly thing happens to me tbh, really frustrating situation.
Now the amp channel 1 module has two failed transistors one pair 2SA1492 / 2SC3856, two small 1/2W burned base resistors, tree open/burned 0.47ohms 5W emitter resistors and the unobtanium 2SJ78 / 2SK215 lateral MOSFET's.😱
Since there is no 2SC3856 avail at Mouser, Newark nor Digi-Key and since eBay is full of fakes to bother with, I opted for all the pair 2SA1943N / 2SC5200N(2SA1386A/2SC3519A was pricey) as I've read some threads users using them as replacements and they are a tad better, also listed by the NTE as replacement too.
So I ordered all the parts, 6 pairs 2SA1943N / 2SC5200N and the silicon thermal pads, all the small base/emitter 1/2W resistors, all 12x 0.47ohms 5W emitter output resistors.
However I'm now stuck at trying to find 2 pairs 2SJ78 / 2SK215(or 2SJ79/2SK216) lateral mosfets from any reputable sources to no avail, and the ones I've found under the U.S. location are a bit pricey for being NOS parts if that's even the case.😒
All this frustration made me spent lots of time searching/compare for those laterals mosfets and I almost to execute an absurd experimentation, just replacing the 2SJ78 with -> ECX10P20 and the 2SK215 with -> ECX10N20, I know those laterals are way overkill for the role of pre-drivers but I could not find anything similar in the case TO-220 unfortunately.🤔
So the question is, after re-biasing the amp does this big laterals can work as drivers in place of the unobtanium/obsolete 2SJ78 / 2SK215, and what could be the drawbacks of using such large/overkill laterals instead the smaller with less input capacitance, I guess the amp may be prone to oscillation at high frequencies and/or lower amplifier gain? assuming this overkill laterals can work as drivers of course.
Also is there another/better 2SJ78/2SK215 replacement candidates with similar specs regardless of the gate/drain pin configuration, I'm up for options/experimentation.
Amp channel modules cleanup/preparation started:
Regards
Last edited:
I tend to look on ebay for obsolete transistors.
The only vendor for laterals is now Profusion (Exicon) which I suspect you already found.
The only vendor for laterals is now Profusion (Exicon) which I suspect you already found.
Hi, yeah I've spend a bit of time on ebay searching for this NOS J78/K215 parts under the U.S. region but only found one seller for the 2SK215 with an stock illustrative picture resembling my 2SK215 in every little aspect/details, but when I search the seller store neutral/negative feedback I've found keywords like counterfeit, non original and/or fakes under private listings and made me wonder.
As for the 2SJ78 on ebay, I can only find a lower voltage 2SJ76 and its package is color green, the other listings I've found for the J76 either has an one star rating and/or looks like a relabeled generic TO-220 package.
Here some recent findings and my assumptions I think they are bad, but honestly I don't have much experience with the different mosfets behaviors etc.
Here is the pair J78/K215 behavior, in the channel-2(the working one):
2SJ78 - The device turns ON and stay ON after one momentary 9v trigger, diode drop = 512 free, 560 with discharge resistor added.
2SK215 - The device turns ON and slowly fades OFF after one momentary 9v trigger, diode drop = 260 free, 534 with discharge resistor added.
Here is the pair J78/K215 behavior, in the bad channel-1(the failed one):
2SJ78 - The device turns ON and slowly fades OFF after momentary 9v trigger, Diode drop = 560 free, 558 with discharge resistor added.
2SK215 - The device turns ON and instantly fades OFF after one momentary 9v trigger, Diode drop = 251 free, 531 with discharge resistor added.
I have to note that both channels where working fine previously and that the BIAS seems ok on both channels as well.
Also all the lateral mosfets works identical(instant ON/OFF) after the ~10k gate resistor placed to keep the device instantly off after trigger absence.
Maybe I'm just acting overly paranoid and starting to guess that this is a common lateral mosfet behavior, and it may change when they aged apparently.
Here are some illustrative images of the originals:
Also I'm not sure if this lateral devices get hot whet the amplifier is pushed but since they are old and almost extinct, I will put some small heatsinks on all of my Samson amplifiers as this laterals are standing just freely on all of them.
Regards
I would have used the Exicon lowest power lateral's with correct voltage.
They dont look too different in size. The Exicons will be TO247 although gate capacitance might be higher.
I tend to use 390R gate resistors with the Exicons.
They dont look too different in size. The Exicons will be TO247 although gate capacitance might be higher.
I tend to use 390R gate resistors with the Exicons.
Hi, thanks for that suggestion, I will look at the Exicon website for they lateral cross reference list and look at the lower power devices indeed.
The amplifier has +/- 82v rails so as a guess the suitable lowest voltage device may be the 160v ones but will do more research on this as well.
Regards
Hey mate, not sure if it helps but Wagner Australia sell the 2SK216. I'm not a transistor guru or anything but this one appears on the same datasheet with a higher drain to source breakdown voltage. Not sure if it's suitable for your application. I've bought from Wagner plenty of times and they are legit.
Very good. It says they offer international shipping so you should be OK. If not, I can proxy for you if you're willing to trust some random internet stranger lol.
You can try here as wel: https://www.diyaudio.com/community/threads/nos-hitachi-2sj79-2sk216.333943/
Hi Bensen, thanks a lot for pointing that NOS thread in regards the laterals.
Between I was in more deep research in regards this enhanced lateral behavior and to my surprise they all worked good in the recommended/equivalent circuit despite the gate capacitor being different in all of the J78/K215 devices, all of them where switching ON/OFF as in the intended application and all of them showed about 12~14 Ohms when turned on, even activated them with mV ranges up to 9v, so I decided to test them in the remaining amp module as test bed and both pairs worked as intended and the amp BIAS where between 7~8mV across both output emitters and the output offset voltage was below 2.0mV range.
This will not stop me to test those large Exicon laterals as drivers regardless, though I'm looking also on those "SOCL 50X TEF" driver/protection boards around YT as I have already the chassis and power supply at hand to play with.
Regards
Small update!
After some time and bucks investment in a lot of parts, the amplifier modules finish/test went well at first glance.
Unfortunately after a careful inspection and cold testing before power-up, the amplifier module 1 failed and blown PNP finals, this time more as I made them 10 finals per module.
So disappointed already but I have only two options in mind before bin the amp, either getting one of those DIY amp PCB's(from gerber files) which are newer amp designs and reuse the heatsinks, or for less money just get lazy and put two ICEpower 500A modules since the rails voltage in this chassis is +/-80v, and start accepting Class D regardless.
P.S. Guess adapting the ICEpower modules to the chassis may be simpler and cheaper than redo the whole Class A/B amplifier modules again, plus the efficiency is welcome as long as it sound good.
After some time and bucks investment in a lot of parts, the amplifier modules finish/test went well at first glance.
Unfortunately after a careful inspection and cold testing before power-up, the amplifier module 1 failed and blown PNP finals, this time more as I made them 10 finals per module.
So disappointed already but I have only two options in mind before bin the amp, either getting one of those DIY amp PCB's(from gerber files) which are newer amp designs and reuse the heatsinks, or for less money just get lazy and put two ICEpower 500A modules since the rails voltage in this chassis is +/-80v, and start accepting Class D regardless.
P.S. Guess adapting the ICEpower modules to the chassis may be simpler and cheaper than redo the whole Class A/B amplifier modules again, plus the efficiency is welcome as long as it sound good.
Hello, I am having problems obtaining those FETs as well, according to the data sheet, they would be 180V, 0.5A, perhaps someone can tell us if it would be possible to replace them with MOSFETs from the IRF family that are much easier to obtain. Attached diagram. Thank you
Hi, you may try Wagner as they carry the higher voltage ones 2SJ79 / 2SK216 as denoted previously, or even try the Exicon ones but they are TO247 package case and need some mods, but didn't tried them actually so can't confirm they may work at all, can even be dangerous.
I also have several of those Samson's amps 2x stock/working S2000, one stock/working S500 and one recently damaged S700.
P.S.
Since those remaining enhanced/laterals Fet's price keeps rising and for the price on getting a small lot of them for spares, I was considering replacing the whole amplifier board with a newer design without those Fet's like the Samson SXxxxx series which circuit are way simpler, or just getting a ready designed circuit pcb, however this may require ordering a minimal amount of PCB's, lot of parts plus time and patience(search SOCL 504 TEF circuits for premade layouts).
OTOH with less money one can easily adapt to an Samson S700 two ICEpower500A Class D modules using the existing power supply(80v), and re-using the input buffers, vol pots etc. with minimal mods.
I attached the Samson SX1200(SX3200 was the biggest on that series) schematic for circuit/topology comparison.
Regards