Hi David,
I think Fred is getting more comfortable daily. That's a very good thing! As for what he is more comfortable with, that's for him to say. Shipping is a real cost we all have to watch for. That normally means buying extra things we need in the future to exceed minimum order levels. Being in Canada, we're very aware of this.
-Chris
Sorry, but you lost me there. If I'm moderating, I'll make that clear. Otherwise, I'm a normal member, just like you. Never worry about that having an effect when speaking with me - please. If anything, I have to watch what I say.
I think Fred is getting more comfortable daily. That's a very good thing! As for what he is more comfortable with, that's for him to say. Shipping is a real cost we all have to watch for. That normally means buying extra things we need in the future to exceed minimum order levels. Being in Canada, we're very aware of this.
-Chris
Hi Pete,
Yes, EDN and other publications have finally clued into these problems over the last few years. The counterfeit parts issue has been here is force for well over 20 years, but the industry didn't really think it had an impact until it began to cost $$.
I have to say that seeing these publications finally waking up was gratifying. Your warnings are certainly valid. At least these days, the chances of getting a fake from the authorized supply chain is far lower than it was a few years ago now.
-Chris
Yes, EDN and other publications have finally clued into these problems over the last few years. The counterfeit parts issue has been here is force for well over 20 years, but the industry didn't really think it had an impact until it began to cost $$.
I have to say that seeing these publications finally waking up was gratifying. Your warnings are certainly valid. At least these days, the chances of getting a fake from the authorized supply chain is far lower than it was a few years ago now.
-Chris
So I'm left with a conundrum. Do I (a) replace just one bad transistor, but with a part of dubious authenticity or do I (b) replace all 8 with a substitute OnSemi part that may, or may not, be an exact match for the specs of the original?
(a) is far cheaper and represents less labor. The risk is an immediate or subsequent failure that destroys more parts.
(b) uses parts that are more trustworthy (on the assumption that a currently manufactured part from an authorized distributor is unlikely to be counterfeit). The risks are technician induced errors and unwanted effects from specs that don't match the original component.
I honestly can't answer the question. Guess I'll sleep on it.
(a) is far cheaper and represents less labor. The risk is an immediate or subsequent failure that destroys more parts.
(b) uses parts that are more trustworthy (on the assumption that a currently manufactured part from an authorized distributor is unlikely to be counterfeit). The risks are technician induced errors and unwanted effects from specs that don't match the original component.
I honestly can't answer the question. Guess I'll sleep on it.
I am indeed getting more comfortable with all this stuff, a fact I attribute entirely to the generosity and patience of the members of this forum. I've learned a ton. And become a very fast de-solderer 🙂
I am also learning daily just how very much I have still have to learn.
HI Fred,
You don't have to have all the output transistors in to get the rest of the amp repaired.
This way you're only putting one or two pair at risk.
When you are ready to do a full load test that's when you will need all the outputs.
I think if you end up with a counterfeit, it will be so poorly matched with the original TRs that you will know. The problem with counterfeits is when you believe they are real and so don't bother to check them. these are very old obsolete transistors and I don't think anyone would bother to counterfeit them, but I could be wrong about that.
David.
You don't have to have all the output transistors in to get the rest of the amp repaired.
This way you're only putting one or two pair at risk.
When you are ready to do a full load test that's when you will need all the outputs.
I think if you end up with a counterfeit, it will be so poorly matched with the original TRs that you will know. The problem with counterfeits is when you believe they are real and so don't bother to check them. these are very old obsolete transistors and I don't think anyone would bother to counterfeit them, but I could be wrong about that.
David.
There is a fairly simple solution if you are comfortable using a hacksaw or dremel tool.
Buy an extra, beta select the one that best matches the ones you have, and cut off the top on the other. Fakes rarely have the correct die size and construction.
This assumes that they look to be from the same lot. You can also cut open the dead one from this amp to use as a reference.
"I am indeed getting more comfortable with all this stuff, a fact I attribute entirely to the generosity and patience of the members of this forum. I've learned a ton. And become a very fast de-solderer
I am also learning daily just how very much I have still have to learn."
We all started here Fred.
I'm am still learning new things to this very day. I enjoy that part more than doing.
David.
I am also learning daily just how very much I have still have to learn."
We all started here Fred.
I'm am still learning new things to this very day. I enjoy that part more than doing.
David.
I've been working on a JBL 6290 amp for the last month or so. This amp uses Sanken 2SA1494/2SC3858 for the drivers and outputs. The one channel that had been repaired previously (not by me) had several outputs shorted so I ordered new ones from MCM Electronics, 5 of each. Installed them and the amp seemed to run fine on the bench UNTIL I ran it well into clipping then pop it was dead again. I just now cracked one of the new transistors open to find it a fake as described on the ESP site. Then I cracked open one the transistors that I removed, it was a real one, huge die compared to the P.O.S. I got from MCM. MCM is now off my list for 2SXXXX transistors. Where do you get real 2S XXXX. How is "B & D Electronics" for real transistors, I think I saw them mentioned on the forum.
Any one have any JBL6290 experience? There doesn't seem to be much info out there. I found the owner's manual/schematic but the one I have has had some updates to it that are not on the schematic. Didn't mean to hijack this thread but I thought the fake transistors were pertinent.
Craig
Any one have any JBL6290 experience? There doesn't seem to be much info out there. I found the owner's manual/schematic but the one I have has had some updates to it that are not on the schematic. Didn't mean to hijack this thread but I thought the fake transistors were pertinent.
Craig
Fred I have an working power amp board from a powered mixer that was abandoned in my basement. The band didn't want to fix it and no one came to pick it up. The board is loaded with MJ21195 and MJ21196. These transistor are the next generation from the MJ15022 and MJ15023. The biggest difference spec wise is the beta is higher.
Have a look at the data sheet at OnSemi and see if these would work for you.
If you want them I'll pull the ones from the working board and give them to you for free.
This amp has never been repaired before and so the outputs are definitely original.
Yorkville has a huge stockpile of these transistor. These ones are stamped Motorola and made in Mexico, not china.
Let me know and I'll take them off the board for you.
David.
Have a look at the data sheet at OnSemi and see if these would work for you.
If you want them I'll pull the ones from the working board and give them to you for free.
This amp has never been repaired before and so the outputs are definitely original.
Yorkville has a huge stockpile of these transistor. These ones are stamped Motorola and made in Mexico, not china.
Let me know and I'll take them off the board for you.
David.
Hi Pete,
Opening one up is a good way, but some fakes use a large die device that has a lower VCE rating. Your method is about the best tehre is though.
I measure the beta and E-B capacitance and look for the readings to be in a normal range for the part. But even before that step, just looking at the printing and case style will speak volumes.
Someone once tried to sell me a 2SCxxx number in a T0-3 case. The case was clearly the old Motorola types with the thick aluminum flanges.
Hi David,
Hi Craig,
No problem. Just use a modern equivalent and buy from a proper distributor. Sorry to hear you were burned.
-Chris
Opening one up is a good way, but some fakes use a large die device that has a lower VCE rating. Your method is about the best tehre is though.
I measure the beta and E-B capacitance and look for the readings to be in a normal range for the part. But even before that step, just looking at the printing and case style will speak volumes.
Someone once tried to sell me a 2SCxxx number in a T0-3 case. The case was clearly the old Motorola types with the thick aluminum flanges.
Hi David,
There was a problem with people returning bad devices that were actually fakes for warranty. That is about the only way that their parts could become contaminated. Every company ran into that problem. Digikey now looks at lot numbers and date codes, no more problems.
Hi Craig,
No problem. Just use a modern equivalent and buy from a proper distributor. Sorry to hear you were burned.
-Chris
Hi David,
Those parts will be fine in the Adcom. I have used those in Adcom rebuilds before and nothing but good ever came of it.
-Chris
Those parts will be fine in the Adcom. I have used those in Adcom rebuilds before and nothing but good ever came of it.
-Chris
My vote would be for the MJ21195/96 also based on data sheet specs. Better hfe and nearly double the SOA of the original part. The higher hfe is important in this amp since it does not have a triple output stage and the original part did also have high hfe.
Wow, very generous. If Chris is confident they would be a suitable substitute, I would gratefully take you up on your offer. And I'll gladly paypal you something for the shipping and trouble. I'l PM you my address.
Presumably I can't just drop one of these in, but have to replace all 8 (no mixing p/n's in a channel, right?).
I do have a couple 424's coming from MCM, so it will be instructive to saw one open and see what we see.
As an old bike mechanic, a hacksaw is a tool I'm already very comfortable with!
That makes three votes (and a combined, what, 80-90 years or so of experience), so I'm convinced!
In the meanwhile, I'll put everything back together but only connect up one pair of outputs.
Should I still leave out the diff. pair on the input board or do we want to just test the complete circuit with everything installed?
NB that the drivers (Q11 and 12) are still the poorly matched 30/31 set. I'll work on understanding/building up Chris's driver matching jig. Maybe that should be the subject of a new thread?
I won't be too concerned right now about the drivers. That's more of a final tweak.
The VAS will just have to work a little bit harder on driving the side with the lower beta driver. Not a big deal right now. As for the diff pair, I'll leave that to whom offered you the advice.
David.
The VAS will just have to work a little bit harder on driving the side with the lower beta driver. Not a big deal right now. As for the diff pair, I'll leave that to whom offered you the advice.
David.
Yes leave out the diff pair since otherwise there is no easy way to do the
leakage test.
Something to consider about matching. PNP and NPN devices are very
different from a physics standpoint. Note the huge difference in the
outputs. It would be completely reasonable to deliberately mis-match
the drivers to make up for the difference in the outputs. You'd want the
products of the driver-output betas to be closer. Also, beta is dependent
on current, Vce, temp - many things. They operate at 80Vce, will you
be matching them at 50 to 80V and 10 to 15mA? Or 1.5V in the IT-18?
Also, the drive to the output is not symmetric, the VAS can and will supply
up to about 15mA if needed to the drivers, pulling the other way, the
current source only supplies 4 mA max.
I think a new thread is a good idea for the matching.
leakage test.
Something to consider about matching. PNP and NPN devices are very
different from a physics standpoint. Note the huge difference in the
outputs. It would be completely reasonable to deliberately mis-match
the drivers to make up for the difference in the outputs. You'd want the
products of the driver-output betas to be closer. Also, beta is dependent
on current, Vce, temp - many things. They operate at 80Vce, will you
be matching them at 50 to 80V and 10 to 15mA? Or 1.5V in the IT-18?
Also, the drive to the output is not symmetric, the VAS can and will supply
up to about 15mA if needed to the drivers, pulling the other way, the
current source only supplies 4 mA max.
I think a new thread is a good idea for the matching.
I emailed MCM about what I found, they are going to have UPS pick them up and refund my money, that makes me very happy. As far as using another transistor that is not possible without reworking the heatsink, MT200 style transistors don't leave much wiggle room as far as replacements go. Can't find new insulators for those either. This amp had some NTE/ECGs in it also, emailed NTE about insulators, no go there either. Going to try B&D as they purchase directly from Sanken and guarantee they will be originals, prices are better than MCM to boot. No JBL6290 experts out there?
Craig
Craig
Your comments about fakes are welcome here, thanks. Why don't you start a new thread about the JBL amp, perhaps an expert will pop up. I'll take a look at it if you post schematics.
They are going to take back the cut open parts? I suppose that is reasonable since it is proof of them being fake.