Can someone tell me why this transistor's Ft is only 5MHz!?

[keantoken]

Cob is 2.2pF (at 100V, but it's about 6.2pF at 1V).

Maybe it's one of those weird "gotcha" parts where Ccb is really low but Cbe is near 1nF?

http://www.google.com/url?sa=t&sour...1QLM3JjC0Y87PkPLA&sig2=mtoBfcQ31ZAtfI0PC4Dxgw

I was looking through parts on a disemboweled monitor and found it. Looked remarkably similar to the 2SC3423 at first, then I saw Ft and took a double take.

It still mystifies me why some parts have extremely low Cob and yet have low Ft.

Is this transistor suitable for use in a VAS?

- keantoken

 

[AndrewT]

I think the answer, for this one, lies in the compromises forced on the designer to achieve high voltage capability.

 

[VHF man]

One should not confuse ft with speed.
They do not necessarily go together and speed will be dependant on the overal cct design.

 

[keantoken]

I agree Andrew.

VHF, design is important, but transistor parameters will set the upper limit for high-bandwidth designs.

One might think that, as long as you have enough drive to overcome the transistor's parasitics, BW is not a problem. But if this was true, I don't think Ft would be specified like it is.

- keantoken

 

[VHF man]

I agree Andrew.

VHF, design is important, but transistor parameters will set the upper limit for high-bandwidth designs.

- keantoken

I was thinking that the low transition frequency might be due (in part) to the low beta.

IMO - The hFE - IC curve is not ideal with that device.

 

[keantoken]

You're right, it's not very linear, it would make an awful VAS. However a lower Hfe should help the Ft, no? A transistor with higher gain would have lower Ib, which would be overcome more easily by the same size Ccb.

- keantoken

 

[VHF man]

You're right, it's not very linear, it would make an awful VAS. However a lower Hfe should help the Ft, no? A transistor with higher gain would have lower Ib, which would be overcome more easily by the same size Ccb.

- keantoken

You correctly identify the importance of source impedance to bandwidth but..
By definition Ft is the point where current gain = 1. So assuming all other parameters are equal the transistor with higher hfe will also have a higher Ft.
I think many designers confuse device Ft with the device 'speed' and hence my original statement.
With any amplifier design you need to start with a defined output current and bandwidth and then work from PA stage back to the input stage - making sure you can actually source the designed current.

 

[keantoken]

So assuming all other parameters are equal the transistor with higher hfe will also have a higher Ft.

No, it will have a lower Ft because the smaller Ib will be overcome by the parasitic capacitor current.

- keantoken

 

[VHF man]

No, it will have a lower Ft because the smaller Ib will be overcome by the parasitic capacitor current.

- keantoken

except that Ft calcs asssume that the signal source has zero impedance.

 

[keantoken]

Source impedance doesn't matter if we are measuring current gain.

- keantoken

 

[wahab]

Source impedance doesn't matter if we are measuring current gain.

- keantoken

It matters since you are measuring an AC current gain...

 

[keantoken]

Granted, but how does this refute my argument below?

it will have a lower Ft because the smaller Ib will be overcome by the parasitic capacitor current.

- keantoken

 

[wahab]

Granted, but how does this refute my argument below?



- keantoken

It doesn t refute this argument which is himself contradictory with
the one i refuted...

 

[keantoken]

Don't confuse me...!

Who's, right, who's wrong, and why?

- keantoken

 

[VHF man]

Source impedance doesn't matter if we are measuring current gain.

- keantoken

OK I agree my last point was confusing. I was thinking of a measurement situation where ic is the controlling parameter and ib is the controlled parameter...
There's no doubt that parasitic capacitance will determine Ft but my point remains that Ft will be higher - assuming all other parameters are constant.
It's loglical that Ft must reduce as hfe reduces to 0, it is therefore logical that Ft will increase if hfe increases.

How can there be any other answer?

 

[wahab]

The (low) Ft just says that at 1 mhz , current gain is in the order
of 5 for the referenced collector current.
As pointed by Andrew, this device is aimed at switching high voltages,
and it s surely optimized in this particular duty..
It won t be used where linear operation is required..

 

[VHF man]

The (low) Ft just says that at 1 mhz , current gain is in the order
of 5 for the referenced collector current.
As pointed by Andrew, this device is aimed at switching high voltages,
and it s surely optimized in this particular duty..
It won t be used where linear operation is required..

At least there is one thing we can all agree on - that this device is no good for most audio applications.

 

[keantoken]

Okay, you're right, I goofed. My argument supports yours, it doesn't refute it.

Still, this transistor might make a good tube amp CCS? The low gain disqualifies it for a lot of things...

- keantoken

 

[Elvee]

Cob is 2.2pF (at 100V, but it's about 6.2pF at 1V).

Maybe it's one of those weird "gotcha" parts where Ccb is really low but Cbe is near 1nF?

http://www.google.com/url?sa=t&sour...1QLM3JjC0Y87PkPLA&sig2=mtoBfcQ31ZAtfI0PC4Dxgw

I was looking through parts on a disemboweled monitor and found it. Looked remarkably similar to the 2SC3423 at first, then I saw Ft and took a double take.

It still mystifies me why some parts have extremely low Cob and yet have low Ft.

Is this transistor suitable for use in a VAS?

- keantoken

It is normal for a high voltage transistor: they have thick junctions, hence low beta and low speed.
You can compare this with PIN diodes: they have a low capacitance, yet they are extremely slow.

 

[keantoken]

Despite the low beta, Cob is still very low which means low miller effect, so it can still be useful for HF applications. In fact, the low beta itself helps decrease miller effect. Even so, it looks like Early affect might be pretty bad for this transistor.

- keantoken