• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

2SA725/ 726 - Transistor Question

C

cb1975

Member
Hello all, I have searched and searched through the forums but cant find a definitive answer to this question.
Can anyone tell me what would work instead of a 2SA725/2SA726 in the Main Amp board of my sx 525.
I do see where people recommend a 992 but Im sure something else more common and easier to find locally would work as well? KSA1845 or KAS1815 or 2838?
Hoping someone has experience with this so I can get this up and running today.
Thanks !
 
Register to hide this ad
Why are you changing them in the first place, the best replacement is not to change them at all, if there is nothing wrong with them!
 
Hyperion said:
Why are you changing them in the first place, the best replacement is not to change them at all, if there is nothing wrong with them!
Click to expand...

In the forum, in the past, it has been extensively noted the necessity of certain pre-emptive replacements.
especially the 2sa725, 2sa726, 2sc1451 and all their complements.

Example: sometime a while back someone decided to do a recap - JUST recap, ignoring the parts list that called for those transistors to be replaced, and then came back bitching about how NOTHING changed to fix the sound problems that prompted the recap.
I told him to follow the parts list and do the recommended transistors.
He did.
And he was very happy the next time he was on with how great the improvement in the sound was.

Those transistors were made as they were straightening out the technology, and while not EVERY batch showed faults THEN, the faults have had 30 to 40 years of aging to do their dirty work.

NOW, the replacements are the product of much more mature and controlled technology, with DECADES of feedback experience with regards to aging faults so they are a far superior part.

Impulse noise from breakdowns (static and pops) are the LAST symptoms to occur and doesn't mean that the transistor was perfect until the popping or noise started, the prior symptoms and problems just weren't enough to bring the circuit down.
 
Last edited:
Well, I did say IF there was nothing wrong with them, I have encountered those types and not needed to change them - that was all I was saying.
 
it's hard to tell the health of those transistors, the problems don't always smack us in the nose (like a dead or noisy one)
and until it's changed, the deficiency isn't usually glaringly apparent.
of course with what I do, I both buy them in quantity - driving down the price, and do EVERYTHING when it IS here
because it is so hard to get it here, then get into it to work on - you only want to do it once...
 
Is there a list on AK of crappy transistors that should be replaced? ...I've been making a list by hand and it seems to keep growing.
 
markthefixer said:
and until it's changed, the deficiency isn't usually glaringly apparent.
Click to expand...

OK, so I have identified my suspected bad transistors and taken them out, how do I test them to check for this 'deficiency'?

What improvement or difference will I notice when the suspects have been replaced with the approved replacements?

Mark, your first response to my post was as if I had said, "don't change them it isn't worth it, they aren't your problem, you need to look elsewhere" would you care to review that response now, as it is clear what I meant?
 
Last edited:
Hyper - I think I understood what you meant in that first post. Mark may have read it differently, I don't know.

For anyone else who may be wondering about the failures of these devices, from what I have been able to determine through a little experience and some research is that many of these older transistors can suffer from several deficiencies. One such problem is the slow but inevitable deterioration of the lead connections inside the package. This is sometimes characterized by the oxidation or blackening of the external leads. This corrosion can creep into the plastic package and deteriorate the resistance welded leads connecting to the semiconductor substrate. As connections deteriorate they can intermittently break and re-connect due to thermal changes, current levels, and other influences. This causes broken contact (open circuit) which can create static burst noise, crackling, runaway of the connected components, and other problems. Another problem is thermal stress distorting the substrate and/or providing poor heat dissipation, changing junction resistances, and running the component out of specifications.

As a result it is often presumed that these older devices are "ticking time bombs" which probably should be replaced whenever encountered as a general reliability issue. Even if the replacement components are no better than the originals (and they usually are) changing them buys the unit another 30-40 years before the problems rear their ugly heads once again, not unlike the theory behind changing electrolytic caps regularly. In reality, and as Mark pointed out, the new devices being manufactured incorporate the lessons learned from the last several decades and use methods to minimize any such problems.
 
SoundOfSound said:
Is there a list on AK of crappy transistors that should be replaced? ...I've been making a list by hand and it seems to keep growing.
Click to expand...

There may be a standing list, but I am unsure of where to find it. I have compiled a cross reference for quick access, and this is some of what I have. Anyone is welcome to correct and/or augment it as it currently stands.

2SA725, 2SA726 Shot noise
2SA809, Poor heat dissipation
2SC458, 2SC871, 2SC1000, 2SC1312, 2SC1313, Leakage, noise, static
2SC1318, Thermal stress, thermal flex
2SC1344, 2SC1451, Static burst, Open circuit faults
 
my head is really out of the game right now, and while I don't see /understand what Watthour read, I am no stranger to apologies.

What may have appeared to be a personal attack, was intended primarily as damage control.

For any personal attack, intentional, unintentional or unconscious that translated through, I humbly apologize.

I have rethought and reworded the first lines of that post.


Watthour - nice list - head is so out of it I can't think of any additions.

In addition to stuff creeping in from the outside world to hurt a transistor, there is the stuff IN it already from variations in the doping / junction formation process that are slow poisons. It is inevitable, it is just the degree to which they have crept in through the various quality control windows in place in the manufacturing process at that time.
I know - while it wasn't semiconductor junctions I was after, I used to have process controls that formed a photocathode in second generation night vision tubes - which involved high vacuum high temperature vapor depositations of alkali elements and the reactions between them.
 
Last edited:
Hyperion said:
OK, so I have identified my suspected bad transistors and taken them out, how do I test them to check for this 'deficiency'?


?
Click to expand...
You would need a good curve tracer that runs them at power.

They are expensive and the old ones have a limit in that you view the curve and guesstimate the location on the screen and do math by hand.

SO, we just replace known problem transistors when we find them.

Just like caps, a true cap checker is expensive and you need to remove the cap to test it and since its out and caps are cheap.... just put in new ones.
 
Last edited by a moderator:
Thank you Mark, I really appreciate the rewording as I know you have been having a rough time one way and another.

Many thanks to Watthour for his responses and clarification.
 
Last edited:
While I have noted those transistors breaking down under load i.e. curve trace analysis, more often than not even if they pass scrutiny its the intermittent nature of them failing from heat during operation, the position of the moon etc etc so imho when I come across them they get pulled and replaced on general principles, Faster to do that than to pull and check (the only real way to verify their operation due to other potential circuit influences) rather than fuss with them. Like others, I also order small signal transistors in mass quantities to a. have them on hand and b. ability to gain match them when required.
Side note, hang in there Mark :)
-Lee
 
Well, I was seeking clarification as although I know those transistors CAN be faulty and cause noise problems and the like, I wasn't aware that every single one is a ticking time bomb and should be replaced on sight.

Thinking back, I did remove quite a few 2SA726 and 2SC2240 transistors from an amplifier that I own. But only because I believed them to be counterfeit, and also they were contrary to the transistor type numbers quoted on the schematic, they weren't making any noises. Using instead KSA992 and KSC1845, the difference afterwards was really quite dramatic.

The old adage 'you live and learn' is at play here I think.

Thank you for your post Lee ;)
 
Last edited:
Its not the 5 minutes to replace them, its the hour of dis-assmbly to get at the pesky things :)
My own sx-939 had one start to fail a couple of years ago and the occasional burst of static at rather annoying volume levels would cause me to jump, of course it happened whenever I happened to be elbows deep in a live circuit checking voltages.
 
Overundr1 said:
of course it happened whenever I happened to be elbows deep in a live circuit checking voltages.
Click to expand...

That's as good a reason to get rid of them as any :D - and that's horrible isn't it, any loud noise or something unexpected - just when you are 'in there' doing your stuff!
 
You must log in or register to reply here.
Back
Top Bottom