K1259/J200 pairs - report on testing

[keland_uk]

ok I have sourced some K1259 and J200 MOSFETS from cricklewood in the uk (https://www.cricklewoodelectronics.com/)

its for a repair of a Sony TA-3000ES

I have a Peak DCA pro
J200 gives
P-Ch Enhancement mode MOSFET Red-D Green-S Blue-G
Vgs(on)=1.421V at Id=4.99mA and Ig=0µA Vgs(off)=0.941V at Id=5.1µA
gm=45.1mA/V at Id=3.0mA to 5.0mA Rds(on)<1.0Ω at Id=5.0mA and Vgs=8.0V with body diode
K1529 gives
N-Ch Enhancement mode MOSFET Red-D Green-S Blue-G
Vgs(on)=2.423V at Id=5.04mA and Ig=0µA Vgs(off)=1.775V at Id=4.8µA
gm=38.5mA/V at Id=3.0mA to 5.0mA Rds(on)<1.0Ω at Id=5.0mA and Vgs=8.0V with body diode

Vgs off for the 4 off J200's are 0.941, 0.914,, 0.878 and 0.910V
Vgs off for the 4 off K1529's are 1.775, 1.698, 1.791 & 1.862V

So the K1259's are within spec of the Y variant (Vgs off range O: 0.8- 1.6 Y: 1.4 to 2.8 from datasheet)
So the J200's are not! They look like O variant (Vgs off range O: -0.8 to -1.6 Y: -1.4 to -2.8 from datasheet)
(mind you the DCA75 says +ve voltages, spec sheet says -ve.....

so the question is
1) Does it matter as long as N-N and P-P reasonably matched?
2) I have read you match N to N and P to P on amplifiers. What is the max difference in mV
3) RTFM, (i've just read it) DCA75 can drive the gate between 0V and 8V, evidently no -ve voltages, and implies MOSFET's don't care if +ve or -ve on gate!) Is this true, and would the +ve Vgs(off) be different to the -ve Vgs(off) values?

Ctt attached, also photo of transistors

[Kleinstein]

The amplifier seems to use only 1 P and 1 N channel FET per channel. In this case there is hardly any need for matching. Matching would mainly be needed if more FETs are used in parallel. Then it is matching N to N obvious as the N's would be in parallel. Rather good matching would help - maybe within some 20 mV. It is still problematic and needs extra care.

There is no need to match the Vth of the N and P side. The relevant number there would be the sum (without the sign). This would effect the threshold and will usually need adjustment anyway if the FETs are changed. So even very different (e.g. 3 V) thresholds for left / rigt channel would not matter.
Ideally one could like a similar Ggm at a higher current (like 100 mA or 1A), but this is usually wishfull thinking as P channel fets usually have a lower gm for a similar size. So the circuit should be OK with a difference there.

[keland_uk]

Many thanks, I had seen the bit about parallel MOSFETS and matching them, it's good to know that in this instance not super relevant.

[Audiorepair]

I have to say that if I were Cricklewood Electronics, I would be mightily miffed at being accused of possibly selling fake transistors.

Did you not notice what you did?

[magic]

Reputation of the vendor is pretty much everything without much more extensive testing. If you are unlucky, a fake part could appear similar and then blow up under load.

Note that Vgs(on) is specified for 0.1A current in the datasheet and the gm plot begins at 30mA, but you are testing at a few mA.

[keland_uk]

well said, no intention of a slur. Sometimes even vendors are hood winked :-(, I've changed the title.

mea culpa