Question on RF or higher transistors as AF amplifiers.

[Thecomedian]

Is there any distortion, unwanted negative effects, or other business associated with using a transistor that has high frequency range like 80-120 Mhz in AF application? Does the intended function of the transistor come with any AF frequency drawbacks that make them less suitable for that application, or can any transistor of the AF band or higher function as AF amplifier without issue? I haven't found any technical information about limitations or overlaps in functions from searching, thinking about buying some decent gain Ge transistors to compliment some lower gain ones I have that are waiting for a high gain partner.

[R.G.]

Is there any distortion, unwanted negative effects, or other business associated with using a transistor that has high frequency range like 80-120 Mhz in AF application? Does the intended function of the transistor come with any AF frequency drawbacks that make them less suitable for that application, or can any transistor of the AF band or higher function as AF amplifier without issue? I haven't found any technical information about limitations or overlaps in functions from searching, thinking about buying some decent gain Ge transistors to compliment some lower gain ones I have that are waiting for a high gain partner.

As a practical matter, it's difficult to find silicon transistors with decent gains at 50MHz or higher.

"AF" transistors are just transistors that are not *guaranteed* by the maker to do well at some higher RF frequency.

The current gain and frequency response are not related in a clear way. So when you say "looking for high gain partners" that's not directly related to frequency response.

In general, for good design practice, you would like to use devices with a highest practical response (look for "Ft" in the datasheet) that is at least 10x the highest frequency you want to mess with, so that the high frequency quirks of the device are well beyond the needs of your circuit. You can then apply external limitations to the bandwidth of your circuit without worrying about interactions with the device quirks.

There are some issues with very high bandwidth devices if you do not plan to externally limit the circuit response. I have a 20MHz scope I use for most things and a 75MHz scope I get out when I need to do something athletic. But if your circuit is oscillating at 100MHz,  a 20-50MHz scope can't even SEE it. All you'd be able to tell is if it interfered with FM radio or aircraft NAV/com equipment. This would be rare, but is an issue.

As a practical matter, you can't escape this last. The small signal MOSFETs we use like the BS170 and 2N7000 have responses well into multiple hundred MHz.

Don't sweat what you can't control.

[Thecomedian]

Thanks again for your help. I have been aware of f(t), and I've been doing some brute force tests in spice to validate the theory of capacitance between b-c junction as a function of reverse bias voltage ratio . It's really interesting, a few volts difference in a fuzz face, and at around the 100k-10mhz range for a specific set of transistors can see a much more "preserved" gain curve if the collector voltage is raised so the Vcb ratio is higher, although it's not really relevant for AF unless Im using 1mhz or less f(t) transistors.

From what I understand, then, there's no "reverse f(t)", where a transistor has poorer performance if its "built for RF" and can handle higher frequencies, and any 500khz+(x)hz/khz/mhz transistor will work just fine for the application, then?

[R.G.]

Thanks again for your help. I have been aware of f(t), and I've been doing some brute force tests in spice to validate the theory of capacitance between b-c junction as a function of reverse bias voltage ratio . It's really interesting, a few volts difference in a fuzz face, and at around the 100k-10mhz range for a specific set of transistors can see a much more "preserved" gain curve if the collector voltage is raised so the Vcb ratio is higher, although it's not really relevant for AF unless Im using 1mhz or less f(t) transistors.

Yes, transistor high frequency response is dependent on collector current. It was once upon a time, in a decade about 40 years ago, common practice to worry about the idle current in transistor collectors having an effect on the high freuqency response. Transistors are so much better today that it's been almost forgotten. And yes, you have to use pretty bad transistors for this to have any effect at audio.

From what I understand, then, there's no "reverse f(t)", where a transistor has poorer performance if its "built for RF" and can handle higher frequencies, and any 500khz+(x)hz/khz/mhz transistor will work just fine for the application, then?

I would say it a different way. Transistors that are built for RF have RF responses and DC responses, which may be slightly different from the RF responses. These devices react to frequencies below 20kHz as though it's DC. In many cases, the issues with RF devices have to do with packaging in ways that do not introduce inductance and capacitance on the internal leads from the package leads to the chip itself. At high RF, the inductance of a loop of gold wire (which was a common way to connect chips to package leads back in the TO-5 days) can seriously damage RF response.

So - don't sweat the RF-ness. It'll work at least that well at DC - which the RF devices think audio is.

[psychedelicfish]

From my experience, AF115s sound pretty good if you were thnking of getting any of them. They work nicely in a rangemaster, but they do get a bit distorted in the stock circuit.