Sony VFET...tragic...

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Echowars:

I know a dumpy pawn shop that has an allegedly-working 4650. I could go get it and pass it along for cost. I'm thinking $50.00. Is it worth for parts? I know VFET's are increasingly rare, is this a good find?

I passed on buying the unit a while back. If I could make someone else's day, that would be cool.

jocko
 
I think these are all toasted too, although I haven't tested them.
 

Attachments

  • sonyvfets.jpg
    sonyvfets.jpg
    53.7 KB · Views: 990
Administrator
Joined 2004
Paid Member
Hi Glenn,
They should have high impedance between the gate and other electrodes. Put an ohmmeter probe (low voltage) between source and gate. Measure the drain - source resistance. Reverse the charge from gate to source and measure drain - source again. One way should be low impedance, the other open or high impedance.

Quick and easy with nothing but an ohmmeter.

-Chris
 
anatech said:
Hi Glenn,
They should have high impedance between the gate and other electrodes. Put an ohmmeter probe (low voltage) between source and gate. Measure the drain - source resistance. Reverse the charge from gate to source and measure drain - source again. One way should be low impedance, the other open or high impedance.
Quick and easy with nothing but an ohmmeter.
-Chris

Chris, this would be true for MOSFETs, these are VFETs.
gate to drain or source will show a diode one way, open the other. Drain to source will show about 1 ohm, perhaps a bit more if gate-drain or gate-source were polarised in the depletion direction by the previous measurements. VFETs have a very high pinch voltage, plus have rather interesting modes of failure, so as simple as these measurements are, they are not conclusive, though.
 
ilimzn said:

.. VFETs have a very high pinch voltage, plus have rather interesting modes of failure, so as simple as these measurements are, they are not conclusive, though.
Right. So I haven't tested them. If somebody can dig up a datasheet I'll have a go at it.
I can contact the owner and see if he wants to give them away. The Sony amp I got them from now has an LM3886 amp in its place. The amp layout was such that it was pretty simple to drop it in place of the original boards. Runs much cooler now.
 
Administrator
Joined 2004
Paid Member
Hi ilimzn,
Chris, this would be true for MOSFETs, these are VFETs.
Ahhhh. I guess I haven't run into any V fets then. Thank you for pointing that out. In that case (not knowing) I probably would have put them in my Mosfet transistor testing jig. They would have passed that test. Looking up a spec sheet is always the best plan with an unfamiliar device too.

Correct me if I'm wrong, but many failure modes would involve the gate, wouldn't they? Either way, the inability to control the drain - source current would indicate a bad device. Or do these get really odd failure modes (like lower breakdown voltage, but fine below that)?

-Chris
 
anatech said:
Hi ilimzn,
Ahhhh. I guess I haven't run into any V fets then. Thank you for pointing that out. In that case (not knowing) I probably would have put them in my Mosfet transistor testing jig. They would have passed that test.

Unlikely, they would show as a short in various interesting ways. A VFET is essentially a giant JFET and is a depletion mode device. Actually, they are very triode-y (is there such a word :) ) except for the 150V and 5A maximum rating :) and high gm. It takes about 20-25V to turn them off completely, and Idss is about as much as their max current rating. Rds at Vds=0 is about 1-1.5 ohms, IIRC.
These devices are practically legendary, and I wish they (or the very few others like them that were made) were still made today!


Looking up a spec sheet is always the best plan with an unfamiliar device too.

Always sound practise, but in this case, given the devices are as 'common' as hen's teeth, specs are practically impossible to find. I do have a GIF of the Id vs Vds @ Vgs graph somewhere, for the P channel part.


Correct me if I'm wrong, but many failure modes would involve the gate, wouldn't they? Either way, the inability to control the drain - source current would indicate a bad device. Or do these get really odd failure modes (like lower breakdown voltage, but fine below that)?

Of course, failure of the gate to control the drain curent will be the most obvious indicator. I had the dubious distinction of having been able to see some of these modes. VFETs are one of the rare devices I have seen fail open, but preserving the GD or GS diode on measurement. To be honest I have never seen the lowered breakdown voltage mode, since I have not had the pleasure of working on the biggest amps, so the rail voltages were still within the +-50V region. I did, however, see leaky gates. This normally does not manifest itself in obvious manners, one tends to get a clue indirectly, from unusual imbalances in driver currents or even input stages. If you measure ehat I mentioned in my previous post, you have a high likelyhood the devices are fine, but as I said, it's not conclusive.
 
Originally posted by darkfenriz
Why is it so?
edit:
I checked several mosfets (IRF's) and the only diode they show is well known 'body' diode from source to drain :confused:

Hehe how many times must I say this? :)
Because these are not MOSFETs!!! but power JFETs :) and as any other JFET, they modulate the width of their channel using a reverse polarized PN junction between the gate and the channel.
 
Administrator
Joined 2004
Paid Member
Hi ilimzn,
If you measure ehat I mentioned in my previous post, you have a high likelyhood the devices are fine, but as I said, it's not conclusive.
Quick tests generally aren't. That's why I perform a leakage test on all bipolars. I've found many leaky parts that measure okay for gain and diode test. Your warning is well heeded.
Unlikely, they would show as a short in various interesting ways. A VFET is essentially a giant JFET and is a depletion mode device.
Never thought about that. Thanks for the comparison. I thought that a V fet and J fet were two difference devices.
It takes about 20-25V to turn them off completely, and Idss is about as much as their max current rating.
Given that they are depletion mode devices (that much I knew), this is not a surprise at all.
Actually, they are very triode-y
That's a good way to describe them.

Thanks ilimzn,
-Chris
 
I might actually have a line on a 2SK60 (the 2SJ18 complement appears to be OK). I'll know in a day or two.

The other VFET models are common drain IIRC, but this 4650 is common source. Also, the manual says nothing about gain rankings, wheras the other manuals make it clear that rankings must match (or be within one number of each other). I'm hoping that the 2SK60 that I'm hunting down will be close to the same ranking and I just don't have to worry about it.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.