Pass DIY Addict
Joined 2000
Paid Member
To go along with the electrical measurements, SoundHappy posted some images that can serve as a quick visual reference.
Toshiba datasheets aside, the fact that Marantz did used non Toshiba looking jfets in their HDAM is undeniable. Go buy a set and open it up to see for yourself, the cd players are cheap on the used market.
I don't believe that Marantz will risk their reputation by using fakes, so those jfets inside are the real deal too !
I don't believe that Marantz will risk their reputation by using fakes, so those jfets inside are the real deal too !
Post a picture of the 2SK170s from your HDAM, while you are at it measure the dimensions.
Making bold statements that it is undeniable doesn't hold water unless you have any tangible proof to support this claim.
Fake semiconductors are routinely sold on eBay. So fake that even a Chinese tester can be used to detect an "outside the datasheet" issue like here, gate leakage current (~10nA).
Attachments
Hi Aridace
Sorry, I have a couple of Qs:
1. Given your pic shows 'hFE' & 'Uf' ... how do you get gate leakage current of ~10nA?
2. What is the tester you are using - in particular, where would I buy one from?
Thanks,
Andy
This is silly JFET is enhanced 367mV so gate current is forward not leakage (10nA is believable at that low a forward bias) and running at Idss so "HFE" reads 220,000!
The tester (LCR-T4 ESR Meter Transistor Tester, look on Google) runs several mA drain/collector current. In this case that required more than 10 nA gate current whereas according to the datasheet, Idss is between 6 and 12 mA. So I measured Idss too and it was 2.6 mA - way outside the specs for a genuine 2SK170. With plenty of comparable, genuine jfets at my disposal the conclusion was easy. BTW the tester above won't measure symmetric jfets like BF246, BF247, E430 - it displays such devices as 2 diodes with threshold voltage for each. But although the P8000 is symmetric too, it gets measured correctly and gives identical values with drain and source interchanged.
Do you mean Texas Instruments P8000? The one in to-220?
Where did you get these, or better: when? They were EOL-ed
some 35 years ago. I still have a few P8002 in a BD138-like package,
same chip, mechanically somewhat smaller. Nothing to write home about.
Gm = 20 mS @ 55 mA, a 50 Ohm termination in grounded grid for
ring mixers etc.
I have never ever seen a data sheet of them.
Where did you get these, or better: when? They were EOL-ed
some 35 years ago. I still have a few P8002 in a BD138-like package,
same chip, mechanically somewhat smaller. Nothing to write home about.
Gm = 20 mS @ 55 mA, a 50 Ohm termination in grounded grid for
ring mixers etc.
I have never ever seen a data sheet of them.
Last edited:
Do you mean Texas Instruments P8000? The one in to-220?
Where did you get these, or better: when? They were EOL-ed
some 35 years ago. I still have a few P8002 in a BD138-like package,
same chip, mechanically somewhat smaller. Nothing to write home about.
Gm = 20 mS @ 55 mA, a 50 Ohm termination in grounded grid for
ring mixers etc.
I have never ever seen a data sheet of them.
Correct, the P8000 was TO-220 and even the pinning was application-friendly (grounded gate) with the gate at the middle. They were also useful in a DBM (10 KHz... 30MHz) until I discovered that provided proper cooling, the same performance was possible with BF246 or BF247. When I was working at Elektor it was easy to order even the most exotic parts.
I have one of those cheap testers as well. Connecting the device you are testing by touching the leads to the SOT pads for surface mount yields different results than using the ZIF. I purchased a few 2SK170 from ampslab that showed as BJT using the ZIF, and JFET using the sot pads. Idss tests were all in range except a few slightly under 6ma. Is ampslab selling legit 2SK170?
When Idss was measured from the group of 20 pcs there was one at 5.3 and two at 5.9. This was using the test described in Borbely article measuring in mA directly (not V drop across resistor).
Used a Keithley 2231A-30-3 power supply and Keithley 2110 DMM. Measured the CCS error of the power supply as +1.6ma across a resistor after supply had been on for 30min. So if the supply was set to 5mA, a measurement of 5mA+1.6mA=6.6mA was seen on DMM. The voltage was measured to be accurate to 10.00 on DMM. I read the calibration manual for the power supply and the voltage is set using y=mx+b method, so that is where I figured the current offset to be slope m ( I am not sure I could be very wrong).
So not sure about these parts. Perhaps the other tests should be done? Or is the cheap China tester just that - a cheap China tester?
No difference between ZIF or PAD use here when preventing contact of fingers with the leads. The datasheet shows 3 classes of the 2SK170: "GR" with Idss 2.5...6.5 mA, "BL" with Idss 6...12 mA and "V" with Idss 10...20 mA. The devices I received are stamped "BL" but have Idss not above 2.6 mA IOW the lower end of "GR". So even if they are jfets, they couldn't be used to replace the original Toshiba "BL" devices. The eBay seller gave a full refund.
According to this post in the "Replacement For Toshiba 2SK170/2SJ74" thread of the Pass forum, their stock can't be trusted.
Just buy them from B & D Enterprises while they have them in stock. They are Authorized Distributor for Japanese transistor makes like Toshiba, SanKen, FUJI Electric, Shindengen, etc.
Thanks very much for this info.
Thanks, ej.
Have not heard of them but I will go looking.
Andy