gain linearity

[jitter]

When looking at the datasheet of the MJL1302/MJL3281 complementary transistors, I noticed it says "Gain linearity from 100 mA to 7 A".

Does this mean that quiescent current should be set to at least 100 mA ?
What would happen if I were to use these in an amp that is normally set to 38.9 mA "idling current" as the service manual calls it?

[sgrossklass]

Originally Posted by jitter

Does this mean that quiescent current should be set to at least 100 mA ?

No.

Originally Posted by jitter

What would happen if I were to use these in an amp that is normally set to 38.9 mA "idling current" as the service manual calls it?

That depends on what it had before. If the old finals were similarly spec'd in terms of fT, beta, current, wattage and SOA, the MJLs should work fine.

[CBS240]

Gain typically drops for outputs at small currents, partly because transconductance drops a bit. However, when you bias a complementary EF output stage, for operation within the bias setting, +/-40mA, the output conductance is the sum of the two device conductance (+ Re/2). Only when one device conducts twice the bias current does the other shut off. The idea is to achieve a linear transfer through the current crossover and the bias is set so that the low transconductance region of each device is 'trumped' by the conduction of the other device thus presenting a more constant output impedance to the load, and so doing produces less distortion.

[jitter]

Originally Posted by sgrossklass

No.

That depends on what it had before. If the old finals were similarly spec'd in terms of fT, beta, current, wattage and SOA, the MJLs should work fine.

I intend to use these in a Marantz PM40SE that used to have the Toshiba 2SA1265N/2SC3182N (O-version) complementary pair. I took them out to repair another amp that uses the same. The amp was working fine, so I know it's just a matter of putting new ones in and adjusting the quiescent current.
As far as I understand from the datasheets, the MJL1302/MJL3281 should work:
fT: the same, 30 MHz;
beta: pretty similar;
current: better
watttage: better
SOA: better.

The only thing I'm not yet sure about is the collector output capacitance (Cob). The MJLs are nearly three times that of the Toshibas (600 pF vs 220 pF). Would that affect performance/sound quality or is it not very important?

[jitter]

Anyone on the collector output capacitance?

[AndrewT]

I am beginning to see comment that the ONsemi spec for Cob is a max figure rather than a typ that is usually presented in datasheets.

I have no way of confirming that.

[homemodder]

I dont think so Andrew, there is a price to pay for better SOA, larger die = larger transistor parasitic capacitances.

[AndrewT]

You don't think I have seen those comments on this Forum?
Or
You don't think I have interpreted those comments correctly?
Or
You don't think those comments were ever posted on this Forum?
Or
You don't think I have no way of confirming my statement?

[jitter]

Originally Posted by AndrewT

I am beginning to see comment that the ONsemi spec for Cob is a max figure rather than a typ that is usually presented in datasheets.

I have no way of confirming that.

Actually, you do, it's right there in the datasheets. Your answer gave me the idea to look at some of the MJL1302 again.

Image1 is of a Motorola MJL1302 datasheet from 1995, note three colums (min - typ - max).
Image2 is of an ONSemi MJL1302 datasheet from 2005, now only min - max colums are given.

In both cases Cob is listed in the max column, and no other value is given. So, it is 600 pF max indeed.
The 220 pF for the Toshibas is a typical value, however, no min or max are listed.

I'm going for the MJL1302/3281 because an amp without its transistors is useless anyway. I'll report if I hear anything out of the ordinary when they're in the amp.

[AndrewT]

ONsemi's later datasheet may well explain why I am beginning to see Forum comment that their quoted Cob can be different from what other manufacturers show.