Is this Genuine part?

[CrazyTiger]

Hi, I am trying to repair a RF Amplifer with defective Output Stage. The original FET installed is SML D1010UK.

Top FET is the original Piece, Middle One is from Ebay (Chinese Supplier), Bottom is from Ebay (UK Supplier).



The Middle one is definitely fake, the size is smaller, the tabs are shorter and the biggest giveaway seller immediately refunded, the instant I mentioned the difference is the physical form.

The Bottom one I am not very sure , I bought 2 of these, it came in a nice ESD box. They look like original. But inspecting closer, I am thinking this is fake as well.




The markings wears off when wiped with IPA, the markings are not straight, I have been told, when components are marked by manufacturer, very strict quality control is done on marking, so the markings will be always repeatable and straight.



Another red flag is between the two FET Die Encapsulation, its not sitting well, its sitting in a bit of an angle and raised up.


Whats everyone's opinion on the bottom FET?

Reason why I did not buy from reputable seller---> The MOQ was 50 Pieces  :'(

[Gyro]

I really can't give a conclusive opinion on whether they are genuine or not. It would be an unusual package to fake, more likely it would be less valuable part in the same package re-marked, either that or a salvaged part, probably removed using a gas flame on the rear of the PCB and no ESD precautions. Possibly a combination of both.

I just wanted to add a warning that some of the parts have a BeO (Beryllium Oxide) marking. You should avoid any attempt to open them to examine the dies or or in any way crack or abrade the package.

[Psi]

I just wanted to add a warning that some of the parts have a BeO (Beryllium Oxide) marking. You should avoid any attempt to open them to examine the dies or or in any way crack or abrade the package.

yep, was about to post that myself.
Don't mess around with Beryllium Oxide, handle it gently and respect it.

[Harry_22]

Hi!
It's a big problem to find an original SML transistor. All the new ones turn out to be fakes and do not produce even a third of the power. At best they blink like light bulbs and burn out eliminating further problems.

Look for used ones from disassembly they are definitely original.
It can also be determined by the inscription that the original is painted in black paint while the fake is laser engraved.

The housings must not be sanded or sawed; If you want to see the crystals heat it up and lift the lid. Then you can carefully close it and use again.

I’m also replacing faulty SML transistors in a 2.7 kW amplifier but apparently I’ll have to give up on this.

PS
Bend the legs upward when soldering (please see the picture).

[wraper]

The markings wears off when wiped with IPA, the markings are not straight, I have been told, when components are marked by manufacturer, very strict quality control is done on marking, so the markings will be always repeatable and straight.

Nope, especially not on pre 2000s components. Marking became really straight only when laser marking became popular, still sometimes not straight. Heck, I have a funny batch of specialty NOS genuine ICs where manufacturer fucked up laser marking by making a typo, crudely blacktopped them (much worse than Chinese counterfeiters) and put a proper marking again.

[wraper]

Whats everyone's opinion on the bottom FET?

Hard to say, it's either an old stock genuine part (some differences over the years of production are quire common) or some other part relabeled. Only measuring it's performance will really tell. The easiest that would reveal fakes would be measuring input capacitance to see if it's not wildly different from a known genuine part.

[CrazyTiger]

I tried the FET. It did not short like first one, however the output cant go above 40W @ 24V Operation Voltage with a Gate Bias Voltage of 7.2V. The rated output for the amplifier is 80W.

For crude testing I wacked in two random IRF450 with a bad heatsink, at least it went above 60W at bias voltage of 4.2V. Surprisingly no smoke.

[CrazyTiger]

Update:
I manage to find the problem, It was not the FET final stage FET (D1010UK), the gain on the first stage ERA-4SM slumped. Replacing it solved the gain issue. Now I get a linear gain about 50dB +/- 1 dB. So I can fairly assume the part was genuine. I can get about 100w at 0dBm input. The rated spec is 80W.

But I am not very happy with the Gate Voltage. The original output stage amplifier had a Gate Voltage of 4.2V. Its set via a simple voltage divider and a 10V zener.

With 4.2V the new D1010uk was not conducting at all. I had to bring it to somewhere around 6V to have a nice undistorted output. The D1010UK datasheet does state the Vth is about 1V-7V, seems very wide if we compare with modern Ampleon RF Fets. Prior installing my peak atlas did indicate about 4.8V.