Totally other JFET, relabeled as a 2SK170. On a real Toshiba the "1" is just a stripe, on a fake it is a stripe with a little roof on it
But why do they work exactly like the original ones, despite from one batch that measures high?
How could they be "totally different" If Idss, V pinchoff, Rds, and yfs are exactly like the original Parts?
I had real fake 2sk170 before and they measured totally different from the original.
If my actual batches are not original, how can they measure as they do?
They are not in the same range as genuine parts. "BL" will have an Idss between 6-12mA, and "V" will have Idss of 10-20. Every single one of yours would have left the Toshiba factory as a "V".
https://alltransistors.com/pdfdatasheet_toshiba/2sk170.pdf
.
If they really did this i would be glad. Cause when i look at the datasheet of the bf862 it could be a nice replacement, cause my rail Voltage does not exceed 20V in my actual projects
They are not in the same range as genuine parts. "BL" will have an Idss between 6-12mA, and "V" will have Idss of 10-20. Every single one of yours would have left the Toshiba factory as a "V".
.
That was the reason that got me asking you friendly folks.
But: only one batch of 10pcs. that came directly from China has this out of spec measurements.
I sourced 110 similar looking "2sk170BL 5A" at different eBay sellers and they all measure well within the BL range. See attached pic.
I will have to get me a Voltage source that can put out more than 24V and then er will see. I will measure bandwith also one rainy day.
First i keep on matching, this stuff sounds no different then the almost unavailable originals.
Thanks for all your responses, very helpful.
Attachments
quick update: i contacted the ebay seller where i bought the best measuring batch, asked him about my observations. He instantly refunded my money without asking me to send the jfets back. I guess he did that to avoid a bad rating on ebay.
Is there anyone with a curve tracer / ability to do in-depth-test to those jfets? Maybe we can figure out what kind of chips are hidden in this TO-92 case. I will be happy to ship some samples of my fake 2sk170s.
I hope i dont get kicked off the forum just because i am willing to experiment with fake 2sk170
Is there anyone with a curve tracer / ability to do in-depth-test to those jfets? Maybe we can figure out what kind of chips are hidden in this TO-92 case. I will be happy to ship some samples of my fake 2sk170s.
I hope i dont get kicked off the forum just because i am willing to experiment with fake 2sk170
Hello,
They need not be relabeled something else.
They may be counterfeit.
Even the real deal has wide ranging parameters. Not every pair that you take out of the bag will give matched performance.
I suppose that you can get matched pairs out a bag of counterfeit transistors.
LSK170 is claiming "or equal performance".
Noise performance is important to me.
If you claim "or equal" performance including distortion, at specified gain and noise. You need test results; then you can start claiming "or equal".
Thanks DT
They need not be relabeled something else.
They may be counterfeit.
Even the real deal has wide ranging parameters. Not every pair that you take out of the bag will give matched performance.
I suppose that you can get matched pairs out a bag of counterfeit transistors.
LSK170 is claiming "or equal performance".
Noise performance is important to me.
If you claim "or equal" performance including distortion, at specified gain and noise. You need test results; then you can start claiming "or equal".
Thanks DT
Yes, you are absolutely right. Of course the noise performance is much worse than the original 2sk170 performes. I plugged two matched pairs of the fakes into my jfet phono (glad i use sockets) and noise dramatically increased. I guess those fake jfets have a much higher bandwith than the 2sk170, because it sounds like i am receiving radio signals when i crank up the volume on my amp with the jfet preamp connected but no record playing.
And even if the fake 2sk170 would measure equal to the original in any way (and not only in idss and vp), there is still something fishy about using fake transistors. Because they will be produced badly and cheap and may tend to fail. One batch of fakes come with one type of chip, the next one may come with a totally different chip.
Still i am not tempted to throw them away.. I will try to measure to get more information and maybe identify the type of jfet that is really build inside those fake 2sk170. Maybe i can use them in unity gain buffer stages or as constant current sources.
Looks good, but to understand just how good you might want to show the region of interest (-0.5 - 0.5 (V) Vgs and 9-12 (mA) Id, in this case) with higher graphical resolution.
Alweit has always been a great source of matched FETs and I'm not questioning the quality (highly recommended!!!).
I'm just pointing out that a 1-2% Idss difference would be invisible in the chosen graphical representation - and I know somebody would ask for less difference before calling it "incredibly good"
You are the champ
What is the standard error across independent measurements?
What is the standard error across independent measurements?
0.3% Idss and 3% Yfs, at controlled temperature, of course.
Patrick
