RF Transistors, is there a 'minimum' frequency where they work?

[cmb]

TL;DR

Some RF transistors are not 'characterised' for lower frequencies, can they still be used?

Some RF transistors are not 'characterised' for lower frequencies, can they still be used?

I get that a lower operating frequency (HF/1.8MHz) may not be the commercial target for an UHF transistor (>136MHz) so no effort spent on characterising them.

Likewise HF transistors >1.8MHz not characterised for kHz.

But is there any design/material limitations that prevent them from working at 'any' lower frequencies? I thought hFE was a maximum, but I am seeing 'minimum frequencies' for hFE in some RF transistor data sheets.

 

[Baluncore]

Some RF transistors are not 'characterised' for lower frequencies, can they still be used?

Yes.
But they are significantly more expensive and fragile than available lower frequency devices.
Ask first what frequency range will it need to cover, and how impedance will be matched over that range.

 

[DaveE]

Summary:: Some RF transistors are not 'characterised' for lower frequencies, can they still be used?

Some RF transistors are not 'characterised' for lower frequencies, can they still be used?

I get that a lower operating frequency (HF/1.8MHz) may not be the commercial target for an UHF transistor (>136MHz) so no effort spent on characterising them.

Likewise HF transistors >1.8MHz not characterised for kHz.

But is there any design/material limitations that prevent them from working at 'any' lower frequencies? I thought hFE was a maximum, but I am seeing 'minimum frequencies' for hFE in some RF transistor data sheets.

They should work OK.

Remember that the data sheets are also a prescription for the manufacturer's testing programs. Suppose you made a device that was intended for and sold to a particular application, like UHF. You save test time and improve yield (money) if you don't test for parameters that aren't important, like audio. This applies to initial characterization, process control, and manufacturing tests.

Knowledgeable designers with normal requirements will buy a different part from you for those other applications. They really don't care too much about people that complain about a UHF device in an audio application, and vice-versa; they'll just suggest you specify a different part.

It also isn't unusual for practicing EEs to use some parts in unusual ways, or to care more than the manufacturer about certain parameters. You may need to do you own verification of suitability before a part is approved.

 

[tech99]

If used in low frequency circuts you might find they are prone to parasitic oscillation at high frequencies (UHF etc) due to the connecting leads.

 

[DaveE]

If used in low frequency circuts you might find they are prone to parasitic oscillation at high frequencies (UHF etc) due to the connecting leads.

Yes. I was always scared of devices that had lots of gain way beyond the frequencies I was working with. You may not care about UHF design and layout for an HF circuit, but you may have to worry about all of that stuff anyway. You'll want to kill off that gain right at the device with good layout and bypassing.

 

[cmb]

Good points well made. As I also thought, probably mainly down to marketing and development engineer effort.

To those two technical points; resonant frequency. Yes I will take that on board it is a very good point. It's an observation that's slightly less helpful in choosing between two particular parts my choice has come down to (as they are very good value, cost wise). We have the AFT05MP075N characterised for 136 to 520MHz, versus the MRFE6VP5150N characterised for 1 to 600MHz. I want to operate 40m. Which is the 'faster' to be more concerned with UHF oscillation gain?

Matching; both are rated to >65:1 VSWR. I am struggling to imagine something they won't match into! ;) These latest LDMOS are amazing if they deliver on all of that.

 

[alan123hk]

If used in low frequency circuts you might find they are prone to parasitic oscillation at high frequencies (UHF etc) due to the connecting leads.

This is a good and professional commentary.
I think many electronic engineers may have similar experience.

A company I worked for once had a master who forced me to use an operational amplifier with a bandwidth of 10MHz to design an audio instrumentation amplifier. The reason is just because that operational amplifier can provide a higher signal to noise ratio. I have told him that this is a dangerous attempt, and there is no need to pursue a signal-to-noise ratio that exceeds the actual need. But he didn't listen to my advice and insisted on doing so. As a result, the finished product of that instrumentation amplifier has a parasitic oscillation that cannot be determined where it is generated and cannot be eliminated. This is an embarrassing and unfortunate situation.