2SK1058 from Futurlec - beware!

[Analog]

Some of my students purchased the 2SK1058 (n-chan) / 2SJ162 (p-chan) complimentary MOSFETs from Futurlec for an audio amp project. I thought it was weird that the 2SK1058 was a little over a $1, and the compliment 2SJ162 to it was over $7, both from Futurlec. So after they came in, the amplifier did not work, set even to its lowest bias. I personally tested the removed MOSFETs, as well as other spares that were never used. Every single 1058 acted like it was always on, regardless of its Vgs. The 162s acted appropriately for a p-channel MOSFET.

The most basic test consisted of a short between gate and source, a 15 ohm 10W resistor from drain to +10V, and source grounded. The MOSFET acted as if it was on, with a Vds of ~0.7V! I couldn't turn the MOSFET off at all. I checked another five of these and they were all like that.

I guess it makes sense that they sell the n-channel version so cheaply, in fact, nobody in the industry comes close! They seem to be junk, so buyers beware!

 

[Speakerguy]

Hrm. I got some IRF1310's there for cheap. I thought it might have been too cheap. Hmm. I'll have to try them and see I guess.

 

[HarveyH42]

Have you Emailed Futurlec with this? Doubt they test/check each and every component. It would do them and everyone else a service to let them know about these parts. If they got a large quantity, they would have a chance to correct the problem before it gets too big. They might also replace your parts, give a refund or credit. Might take some time and patience, they aren't the quickest or most talkative, but have only read a few post where they absolutely refused to do anything. I maybe order from 3-4 times a year, about 5 years now, never a quality issue, usually exceed my expectations. If they were crooked, pretty sure I'd have seen something suspicious over the years...

Students... Any chance they might have gotten their grubby hands in the parts bin? Haven't used a lot of fets, my instructor in High School scared the crap out of me with all the talk about how fragile they are, and how easy they are to damage, all the special care in handling and installing, ect... Not sure how real all that was, but have been finding that they aren't so bad (perhaps I'm over cautious).

 

[Speakerguy]

FET's these days are pretty rugged. I've tossed them around a lot without any caution regarding ESD and I haven't ever had a FET fail that wasn't in response to me catching it on fire. I think this is one of those things they more or less fixed within the last decade.

 

[Oznog]

Not unusual considering 2SK1058 is an ENHANCEMENT Mode MOSFET thus they have different threshold characteristics from your conventional, garden-variety MOSFET and may be slightly "on" even at 0V Vgs. A negative Vgs (positive for P-ch) will ensure they are completely off. Review the spec sheet.

2SK1058 has zener protection diodes built into its gate, it is difficult to damage them with static.

 

[Analog]

Not unusual considering 2SK1058 is an ENHANCEMENT Mode MOSFET thus they have different threshold characteristics from your conventional, garden-variety MOSFET and may be slightly "on" even at 0V Vgs. A negative Vgs (positive for P-ch) will ensure they are completely off. Review the spec sheet.

2SK1058 has zener protection diodes built into its gate, it is difficult to damage them with static.

Oznog,
Thanks for the reply, and I agree with you on the slightly "on" aspect. However, if you looked at what I did, you'd see that they are definitely not just slightly on, but very on. Below is the transfer characteristic of the N-channel 1058 from the data sheet. As you can see, it should have an Id of maybe a few milliamps at Vgs=0. The parts that I tested, including 5 virgin parts that came out of the anti-static packaging, were exhibiting an Id of ~950mA at Vgs=0.

The 2nd graph is one that I did for the P-channel 162. As you can see, it follows the appropriate curve for a P - channel, noting - of course, that the polarities are reversed for everything. It's Id<10mA at Vgs=0.

I'm just saying its suspicious that their part is sooo cheap - I can't find the 2SK1058 anywhere else under $7, whilst they are charging <$2.

If anything else looks funny on my analysis, please let me know. Thanks!

 

[audioguru]

Not unusual considering 2SK1058 is an ENHANCEMENT Mode MOSFET thus they have different threshold characteristics from your conventional, garden-variety MOSFET and may be slightly "on" even at 0V Vgs. A negative Vgs (positive for P-ch) will ensure they are completely off. Review the spec sheet.

I couldn't find a datasheet for an old 2SK1058 Mosfet. Maybe they are war surplus (which war?)

Nearly all garden variety Mosfets are ENHANCEMENT mode.

An old N-channel IRF540 Mosfet has a max leakage current of only 25uA at its max voltage when its Vgs is zero. Like thousands of other Mosfets its theshold voltage is a minimum of 2.0V when it conducts a max current of only 250uA.

 

[HarveyH42]

FET's these days are pretty rugged. I've tossed them around a lot without any caution regarding ESD and I haven't ever had a FET fail that wasn't in response to me catching it on fire. I think this is one of those things they more or less fixed within the last decade.

Thanks. High School was 26 years ago. It's just something that really stuck with me. He was one of my favorite instructors, tolerated a lot of our 'fooling around', didn't treat us like kids, wasn't always serious or business-like. The things he did stress, weren't forgotten.

 

[Oznog]

Ah true. Most are Enhancement mode, I was trying to point out that the 2SK1058 has the unusual feature of no significant threshold knee.
This is probably a silly question, but are you sure you had zero volts on Vgs (shorted), not just left floating, and didn't get the pins mixed up?

~950mA is certainly beyond the acceptable limits. The people to contact would be the part manufacturer. Future Electronics generally deals in parts from active manufacturers, not surplus or no-name, but I wouldn't talk to Future, you'd just get a sales person with no ability to investigate your problem. The mfg should have a representative who should investigate bad parts.

I could not find 2sk1058 on Future Electronics website at all.

 

[Analog]

Ah true. Most are Enhancement mode, I was trying to point out that the 2SK1058 has the unusual feature of no significant threshold knee.
This is probably a silly question, but are you sure you had zero volts on Vgs (shorted), not just left floating, and didn't get the pins mixed up?

~950mA is certainly beyond the acceptable limits. The people to contact would be the part manufacturer. Future Electronics generally deals in parts from active manufacturers, not surplus or no-name, but I wouldn't talk to Future, you'd just get a sales person with no ability to investigate your problem. The mfg should have a representative who should investigate bad parts.

I could not find 2sk1058 on Future Electronics website at all.

Oznog,

I believe you are confusing Future Electronics with the company Futurlec, which is completely different. Futurlec seems to carry some of the more hard to find and obsolete parts, and I use them because they are cheap, but I guess I got what I paid for! Anyhow, here is the link to Futurlec:

Futurlec, The Electronic Components and Semiconductor Superstore

2SK1058: Transistors - Mosfet - 2SK1058

2SJ162: Transistors - Mosfet - 2SJ162

Audioguru: Digikey also has data sheets for both devices. Thanks!

 

[audioguru]

The datasheet does not have spec's for the max current when the Vgs is zero. Every other manufacturer does.

Two graphs show a typical current of zero when the Vgs is zero.

 

[Speakerguy]

2SK1058 has the unusual feature of no significant threshold knee.

I believe it is a lateral mosfet, not a vertical mosfet. Verticals are much better at switching, laterals are more linear and are used more often for audio amplifiers.

 

[Oznog]

I believe you are confusing Future Electronics with the company Futurlec, which is completely different.

Ah, you believe right.

 

[Analog]

Thanks to all that replied. I am going to buy some of the 1058s from other sources, and I'll let you know how they pan out.

 

[Oznog]

The datasheet does not have spec's for the max current when the Vgs is zero. Every other manufacturer does.

Two graphs show a typical current of zero when the Vgs is zero.

Naw, but it DOES specify a "Vgs Cutoff Voltage" of 0.15V-0.45V, conditions Ids=100mA Vds=10V. So below that Vgs, current must be below 100mA. Which it is definitely not.

 

[audioguru]

The Mosfets are probably "seconds" that do not meet spec's.

 

[Analog]

The Mosfets are probably "seconds" that do not meet spec's.

Methinks you are correct...

 

[mneary]

Are the parts labeled? Is it possible that you received P-channel devices instead of N-channel?