Hello,
Since I am going to be building the Wolverine amplifier with four pair of output transistors and a 1000 VA 50v transformer, I wanted to use the best transistors I could find. It seemed to me the best choice was the Sanken 2SC2922/2SA1216 transistors, except they are no longer available from traditional mainstream suppliers. This led me down an expensive and interesting path and I purchased pairs from seven different sources to compare.
Here are my findings,
Two sets are clearly fake, Shenzhen Sacoh Elec. Company Store and adeleparts2010.
Amps Lab and Sansui Vintage Audio Parts are both claiming to be selling NOS transistors and I have no reason to doubt them.
But what about the other three sets?
Are CZH, Tigerfish and "A New Store" actually selling NOS as well?
Or is there a company in China making "new"copies of the old Sanken chips that are pretty close, not obvious fakes, should/can they be used?
Take a look at this NTE58/59 pair in a similar MT-200 package that I thought only Sanken used,
https://www.nteinc.com/specs/10to99/pdf/nte58.pdf
The data,
CZH Labs/Electronics-Salon, China
18.6 grams each and ~36.3mmx21.5mm for both 2SA1216 and 2SC2922
2SA1216 were "76P" code, hFE 137, 119, 89, 108, 92
2SC2922 were "79P" code, hFE 102, 107, 116, 119, 130
Tigerfish Audio , Taobao, China
18.7 grams each and ~36.3mmx21.5mm for both 2SA1216 and 2SC2922
2SA1216 were "95Y" code, hFE 82, 83, 81, 82, 83, 84, 77, 85, 83
2SC2922 were "94Y" code, hFE 122, 126, 119, 122, 122, 130, 126, 126, 126, 130
“A New Store”, Aliexpress, China
18.7 grams each and ~36.2mmx21.3mm for both 2SA1216 and 2SC2922
2SA1216 were "95Y" code, hFE 97, 97, 98, 97, 96, 97, 98, 98, 98
2SC2922 were "96Y" code, hFE 130, 134, 134, 107, 110, 139, 134, 134, 107, 107
Ampslab, USA
18.2 grams each and ~36.3mmx21.2mm for both 2SA1216 and 2SC2922
2SA1216 were "52Y" code, hFE 78, 83, 46, 70, 83, 86, 75, 71
2SC2922 were "52G" code, hFE 217, 217, 217, 217, 217, 206, 217, 197
Sansui Vintage Audio Parts, Croatia
18.5 grams each and ~36.3mmx21.5mm for both 2SA1216 and 2SC2922
2SA1216 were "86Y" code, hFE 91, 95, 95, 93, 90, 100, 97, 101
2SC2922 were "80Y" code, hFE 134, 139, 126, 122, 126, 139, 130, 110
Shenzhen Sacoh Elec. Company Store, Aliexpress, China
14 grams each and ~36.1mmx21.1mm for both 2SA1216 and 2SC2922
2SA1216 were "53Y" code, hFE 139, 133, 133, 129
2SC2922 were "53Y" code, hFE 138, 137, 106, 106, 109
adeleparts2010, Ebay, China
13 grams each and ~36.1mmx21.1mm for both 2SA1216 and 2SC2922
2SA1216 were "53Y" code, hFE 165, 158, 162, 164
2SC2922 were "53Y" code, hFE 73, 69, 85, 82, 66
* hFE was measured with a DROK 1.8 Inch LCD Multifunction Transistor Tester
https://www.amazon.ca/Multifunction-Transistor-Automatic-Identified-Capacitor/dp/B01MS1FOYM
Here are close up pictures of the Fronts, Backs and Insides of the 2SA1216 transistors from each purchase.
(I only split open the Asian sourced ones, forgot to get extra USA and Croatian sourced ones)
Any comments on how the innards look for the Asian sourced 2SA1216 transistors look? Notice any major differences?
After doing all this, I am totally confused as to which are real and which might be fake and am thinking seriously about just buying MJL4281AG/MJL4302AG from Mouser to be sure I am using legit transistors.
Thanks for looking,
David.
Since I am going to be building the Wolverine amplifier with four pair of output transistors and a 1000 VA 50v transformer, I wanted to use the best transistors I could find. It seemed to me the best choice was the Sanken 2SC2922/2SA1216 transistors, except they are no longer available from traditional mainstream suppliers. This led me down an expensive and interesting path and I purchased pairs from seven different sources to compare.
Here are my findings,
Two sets are clearly fake, Shenzhen Sacoh Elec. Company Store and adeleparts2010.
Amps Lab and Sansui Vintage Audio Parts are both claiming to be selling NOS transistors and I have no reason to doubt them.
But what about the other three sets?
Are CZH, Tigerfish and "A New Store" actually selling NOS as well?
Or is there a company in China making "new"copies of the old Sanken chips that are pretty close, not obvious fakes, should/can they be used?
Take a look at this NTE58/59 pair in a similar MT-200 package that I thought only Sanken used,
https://www.nteinc.com/specs/10to99/pdf/nte58.pdf
The data,
CZH Labs/Electronics-Salon, China
18.6 grams each and ~36.3mmx21.5mm for both 2SA1216 and 2SC2922
2SA1216 were "76P" code, hFE 137, 119, 89, 108, 92
2SC2922 were "79P" code, hFE 102, 107, 116, 119, 130
Tigerfish Audio , Taobao, China
18.7 grams each and ~36.3mmx21.5mm for both 2SA1216 and 2SC2922
2SA1216 were "95Y" code, hFE 82, 83, 81, 82, 83, 84, 77, 85, 83
2SC2922 were "94Y" code, hFE 122, 126, 119, 122, 122, 130, 126, 126, 126, 130
“A New Store”, Aliexpress, China
18.7 grams each and ~36.2mmx21.3mm for both 2SA1216 and 2SC2922
2SA1216 were "95Y" code, hFE 97, 97, 98, 97, 96, 97, 98, 98, 98
2SC2922 were "96Y" code, hFE 130, 134, 134, 107, 110, 139, 134, 134, 107, 107
Ampslab, USA
18.2 grams each and ~36.3mmx21.2mm for both 2SA1216 and 2SC2922
2SA1216 were "52Y" code, hFE 78, 83, 46, 70, 83, 86, 75, 71
2SC2922 were "52G" code, hFE 217, 217, 217, 217, 217, 206, 217, 197
Sansui Vintage Audio Parts, Croatia
18.5 grams each and ~36.3mmx21.5mm for both 2SA1216 and 2SC2922
2SA1216 were "86Y" code, hFE 91, 95, 95, 93, 90, 100, 97, 101
2SC2922 were "80Y" code, hFE 134, 139, 126, 122, 126, 139, 130, 110
Shenzhen Sacoh Elec. Company Store, Aliexpress, China
14 grams each and ~36.1mmx21.1mm for both 2SA1216 and 2SC2922
2SA1216 were "53Y" code, hFE 139, 133, 133, 129
2SC2922 were "53Y" code, hFE 138, 137, 106, 106, 109
adeleparts2010, Ebay, China
13 grams each and ~36.1mmx21.1mm for both 2SA1216 and 2SC2922
2SA1216 were "53Y" code, hFE 165, 158, 162, 164
2SC2922 were "53Y" code, hFE 73, 69, 85, 82, 66
* hFE was measured with a DROK 1.8 Inch LCD Multifunction Transistor Tester
https://www.amazon.ca/Multifunction-Transistor-Automatic-Identified-Capacitor/dp/B01MS1FOYM
Here are close up pictures of the Fronts, Backs and Insides of the 2SA1216 transistors from each purchase.
(I only split open the Asian sourced ones, forgot to get extra USA and Croatian sourced ones)
Any comments on how the innards look for the Asian sourced 2SA1216 transistors look? Notice any major differences?
After doing all this, I am totally confused as to which are real and which might be fake and am thinking seriously about just buying MJL4281AG/MJL4302AG from Mouser to be sure I am using legit transistors.
Thanks for looking,
David.
I have a set of real 80s/90s that I'm saving for an upcoming Naim clone, I'm going to look at it tonight
Obviously even some China stores have some genuine NOS. But don’t count on that to always be true - when CZH runs out of their current NOS you may^H^H^H will get fakes from then on from them. Because that’s all they will be able to get.
NTE58/59 were at one time genuine Sanken devices. They just BUY OEM devices, re-badge them and sell at a premium. When they can’t get them from Sanken anymore (which is now) they buy from wherever they CAN. Which is probably ISC, and we know how that ends.
NTE58/59 were at one time genuine Sanken devices. They just BUY OEM devices, re-badge them and sell at a premium. When they can’t get them from Sanken anymore (which is now) they buy from wherever they CAN. Which is probably ISC, and we know how that ends.
Here is an extreme close up of the 2SA1216 from CZH and "A New Store", Can anyone spot any differences?
The rebuilds from "ISC" were investigated with professional equipment by an user in a German community. They behaved well and are electrically equal respectively better then the Sanken originals. They have an equal die size but are manufactured directly on the MT-200 case without the TO3P building block between die and case. If I have time I will take a look if I could find the topic were the investigation was done.
Also, there´s another thread here:
https://www.diyaudio.com/community/threads/counterfeit-transistors.115281/
https://www.diyaudio.com/community/threads/counterfeit-transistors.115281/
I would be very curious to read that investigation.
Do you recall what forum or web site it was on?
Thank you,
David.
Yes, I would be curious too. Especially whether the testing was limited to data sheet parameters only. The Sanken MT200 LAPT types were VERY VERY often used well above ratings (as were 2SC3281’s and 2SD424’s) by receiver manufacturers (and many low end pro amps) because they KNEW the transistors would take it long enough to get past the warranty period. You’d see single pairs operating at +/70 or more volts where there clearly isn’t enough SOA. Real Sanken parts don’t blow up immediately the first time you turn it up to 11 - it takes a couple years/months of use for the receiver to die. If you put in something that just barely makes spec, you‘ve got one time use and then the outputs are blown again. I HAVE seen this happen many times over with NTEs as replacements, and once got bit myself with a bank of Mospec 2N6609 being run on +/-85 volts Where original Motorola from the 80’s is still working in the other channel. Parts from that same batch had no heartache with +/-60 volt operation in my ancient QSC Model 41 (Which originally had the horrid MJ15016’s when I got it back this time - that took out the drivers because of the low gain - something you really have to watch out for in second source parts too). It’s been bounced from user to user for 40 years and fixed a half dozen times by the look of it so no idea what it came with from the factory.
Doesn't look like there is any important ones. Datasheet states that weight approx.18.4 grams. both are close to it. and die size is close to 5.5x5.5mm. But to make sure measure linearity of h21. Take few points from 100mA to 5A (pay attention to temperature). More-less linear h21 in that range (±5-7 units) will indicate quite high probability that transistor is genuine.
The big difference between SanKen and ISC was charge storage and capacitance. Measuring turn-off time under high di/dt should be a good test.
I put the transistor pair in a class AB output stage, made it output several amps at 20kHz, and checked base and emitter currents in both transistors.. Details in the link in post #11.
In fact, many times. The products sold by merchants will directly tell you that this is fake.
But in fact, users will buy it.
For example, the sale of genuine products is $5. The fake sale is $2.
In fact, most people will buy $2 directly.
last. People who sell $5 have no choice. Only to sell fake goods for $2.
Sorry to dredge up an older thread, but what did you end up doing? @DaveFred?
Did you find a way to verify your transistors well enough to have peace of mind?
I'm in a similar situation. I'd like "the ultimate outputs" and ended up with 2SC2922/2SA1216 pairs from B&D Enterprise (Bdent.com). Their weights are correct. They look exactly right on the outside. And when I crack them open, the insides look exactly like yours. The chips are 5x5mm exactly.
B&D sent them with mica insulators, which is handy and (if I remember correctly), Saken shipped their MT200 devices with them. So that makes me feel better about them, I guess?
Did you find a way to verify your transistors well enough to have peace of mind?
I'm in a similar situation. I'd like "the ultimate outputs" and ended up with 2SC2922/2SA1216 pairs from B&D Enterprise (Bdent.com). Their weights are correct. They look exactly right on the outside. And when I crack them open, the insides look exactly like yours. The chips are 5x5mm exactly.
B&D sent them with mica insulators, which is handy and (if I remember correctly), Saken shipped their MT200 devices with them. So that makes me feel better about them, I guess?
I value your work very much @peufeu , really excellent and sensible material. Very useful, informative.
(Fortunately most of us will never need several amps at 20KHz. But I agree with the testing method, to really expose the shortcomings of the outputs).