Counterfeit 2SC2912 (and others)

[davada]

Originally Posted by PB2

I'm not sure if others are interested but I am since I scratch built one.

Maybe you did get some fake parts - hard to say.

I thinking maybe had to do with the layout of the kit. If your scratch built one has worked for years.

David.

[anatech]

Hi Fred,
Yes, the MCM part is an older production part, but real. It's interesting to see really old parts being sold by MCM. Not what I would expect. I guess the real question is .. where did those older parts come from?

Hi David,
Yes, those designs could be very unstable. There is gain in the output stage in many of these designs. Bypass capacitors are critically important with these. Use non-inductive emitter resistors as well. You may even need to use base stoppers with some.

Hi Pete,
I've worked on quite a few similar amplifiers, and they all share the same tendency to oscillate. The stacked output stage is even more susceptible than a normal E-F stage.

-Chris

[MCMElec]

Originally Posted by PB2

I have to wonder how you can say this given the number of reports of fakes from them here on the board. Fred and I have wasted a lot of time on his particular problem, it is not a minor thing IMO. Fred's driver board is probably hacked, not because of his ability, but because he had to swap parts so many times. Consider that in two separate orders he got fakes? Next he will have to waste a part that he paid for to hack open and build some confidence that the other is real. I will certainly buy first from a dealer that I trust since I don't like to waste my time and risk my projects on this sort of situation.

I welcome a solution from Mr. Minix that doesn't involve using customers for test of his inventory. Certainly there is a huge difference between we provide a quality product from the start, and we stand behind our product and will take back parts that customers determine are fake or defective.

Perhaps I said it wrong, but what I meant to communicate is that if there is a problem with a product (ICs and CAPs do have a certain defective rate that we cannot control or test for), I would certainly prefer people to contact us directly and allow us to set it right before they assume that we are in the business of ripping people off and then make those statements on websites. We stock over 40,000 unique parts and on some items we process multiple receivings every month. I can't remain competitve and have enough staff to fully test and inspect every product that we receive. We have to rely on our supply chain and then verify where possible or if we have specific problems. Our buyer for these product categories has issued an alert to all of our suppliers of these products and we will begin inspecting receipts from the supplier of the items referenced herein for a period of time. Our tech support people and customer service people have been asked to relay any information they get back to our buyers quickly so that we can attempt to isolate any problem and rectify it.

MCM is part of Premier Farnell and therefore is a sister company of Newark, Farnell, etc. We are located in 23 countries throughout the world and we all have a lot of pride in serving our customers well. Even being as big as we are, we have had this problem in all divisions of the company. It's not an excuse, it's just a bit of a reality that this happens and we try to avoid it -- however, we also have to balance how much we can spend to root it out 100%. Part of our unique selling proposition at MCM is having hard-to-find parts. That means that we also work through brokers and agents to source things our customers need and this creates some adidtional challenges.

I appreciate being part of the discussion.

[fredlf]

I don't believe that MCM is in the business of purposefully ripping people off, and I hope I never implied any such thing. I will say that they have been very responsive to me and my issue, and will be refunding me for the bad parts and even for a set of probes that melted while testing said bad parts. They have been proactive (they called me first!) and very apologetic.

I appreciate that they are stocking hard-to-find parts, and that this brings with it issues of authenticity (Chris, I suspect that many of these "old" parts are actually "re-manufactures" that are made by third parties, hopefully under license or some agreement with the OEM).

That said, does getting back the handful of dollars that is the cost of the parts come close to compensating me, and everyone else here, for all the time and effort (not to mention cost of other parts purchased for troubleshooting)? No, of course not. Not even close. And yes, as Pete pointed out, my input board is knackered from all de-soldering and re-soldering (see the attached pic of a repair I had to make to a worn-out trace on Q4, one of the bad parts).

Obviously, it's not reasonable to expect MCM to reimburse me for any of that. Nor could they reimburse me for the self-doubt and frustration that I experienced trying to figure out what could be wrong. Seems to me the solution has to come from avoiding the problem in the first place.

I would gladly pay a premium for a part that is 100% guaranteed tested and genuine. I bet other people would too. Doing those tests would also allow MCM to offer parts matching, a service many DIYers would appreciate greatly. I would also pay more for a guarantee that covered more than parts replacement. For example, Kryptonite bike locks have long had a warranty that, in the event of a theft when their lock was being used, would replace not just the lock but the stolen bicycle (up to a certain amount) as well. Could a similar warranty work for select electronic parts?

I'm sure there are other potential solutions, but I don't really know this business very well. Anyway, that's Mr Minix's job! All I do know for sure is that the current model doesn't work very well for consumers like me.

[russel]

Originally Posted by fredlf

I would gladly pay a premium for a part that is 100% guaranteed tested and genuine.e.

That is what the military do. You do get a complete audit trail and fully tested parts but is it worth the cost for what we do?, most likely not.

For example

N2369A JANTX2N2369A Military Transistor


Price: $22.20
Quantity in Basket: None

430 available for immediate delivery
Product code: T452PD

Volume Pricing
Quantity Price Each
1-9 $22.20
10-99 $20.20
100+ $18.20

Quantity:



View shipping costs




Military Transistors
NPN
TO-18 Package
Gold Plated Leads

Maximum Ratings:
Power Dissipation @ TA=25°C: 360mW
Power Dissipation @ TC=25°C: 1.2W
Collector-Base Voltage: 40V
Collector-Emitter Voltage: 15V
Emitter-Base Voltage: 4.5V
Operating & Storage Temperature: -65°C to 200°C

Specs:
DC Current Gain @ IC=10mA: 40 min ; 120 max
Output Capacitance: 4.0pF
Turn-on time: 12ns
turn-off time: 18ns

Manufactured by: Raytheon
Part Number: JANTX2369A

or


2N3716 JANTX2N3716 4A 80V Military Transistor


Price: $225.75
Quantity in Basket: None

7 available for immediate delivery
Product code: T458PD


Quantity:



View shipping costs

[fredlf]

Knowing what I do now, I would pay $20 each for a matched, guaranteed pair of critical drivers or a diff. pair. However, $225 is a little steep (though now I know where that $350 military toilet seat in the 80's came from!).

A full audit trail seems a bit much. Maybe there's a middle ground?

[sonidos]

This is quite interesting. A couple of you know I bought a Heathkit IT-18 beta tester and finally got it working.

Cutting to the chase, I have several Sanyo 2SD1047 transistors that I purchased from two sources, MCM and BDent. I purchased these sets last Fall and going back through them, the ones from MCM have a different silkscreen and measure below the specs on the datasheet. Specs say 100 to 200, they measure around 50. The BDent units I got in the fall measure around 120, so within spec.

But I just bought a dozen more from BDent. They look like the real mccoy, but they all measure around 65. What gives?

[davada]

Originally Posted by sonidos

This is quite interesting. A couple of you know I bought a Heathkit IT-18 beta tester and finally got it working.

Cutting to the chase, I have several Sanyo 2SD1047 transistors that I purchased from two sources, MCM and BDent. I purchased these sets last Fall and going back through them, the ones from MCM have a different silkscreen and measure below the specs on the datasheet. Specs say 100 to 200, they measure around 50. The BDent units I got in the fall measure around 120, so within spec.

But I just bought a dozen more from BDent. They look like the real mccoy, but they all measure around 65. What gives?

Is there any letters after the part number?

[davada]

WS data sheet says Hfe 55 min 160 max.

[davada]

Sanyo data sheet says 60-200 and Mospec says 20-200.

I guess it depends on who the manufacture is and there may different beta grades.