durability: IRF520 vs HG2sc1969

[Uncle Ronnie 336]

in your experience which have you popped more of? honestly i dont have a good track record with the HG's. they seem to be more sensitive than mosfets

 

[nomadradio]

The IRF 520 famously does not tolerate high SWR. We have had good luck with the HG1969, but it's not as tough as the original Mitsubishi.

Keeping the input circuit of an amplifier close to 50 ohms is your only hope using either of those parts. A worn-out relay is a common radio killer. And if you're barefoot, the antenna system can't have any intermittent 'twitch' issues or steadily high SWR.

What the world needs is a radio final that puts out monster power and tolerates ten-to-one SWR with no more than a shrug.

Sounds like that old politician "A car in every garage and a chicken in every pot". And about as likely.

73

 

[brandon7861]

Just my opinion here...

520's are a dollar a piece at mouser. The 1969's are 4x that price (at ICA) and legit parts are rapidly disappearing from the market.. I never had issues with mosfets blowing up or getting hot, but then again, I'm not asking them for 30w either..

Its not that the parts "cant handle" a little extra power, 520's are rated for 60w or 7A, the problem is that they are mismatched when you do "turn them up". For example, on AM, the collector (or drain) is sitting at 6v. The impedance of the final is approximated by E^2/(2P)=Z so if you have 6v at the drain and ask it for the 4w the radio is designed for, the impedance of the transistor is about 4.5Ω.

What happens when you ask it for 10w on AM? The transistor impedance drops to 1.8Ω! Unfortunately, the matching network between the transistor and the output filter is still designed to match 4.5Ω to 50Ω, not 1.8Ω. 6v/1.8Ω is still under the 7A rating, but some of the power is immediately reflected back into the part when it sees the 4.5Ω matching network and the probability of magic smoke goes up.

Every time the power changes in the transistor, the transistor impedance changes. Now I'm not accusing you of doing that, I am simply pointing out that if people stop turning them up, they will stop blowing up!

Edit: I used to have a smith chart plot of a common radio final section that showed how a mismatch at the antenna jack is transformed through the filter and matching network into a mismatch at the transistor (for comparison to the effects of power change), I am trying to find that....

 

[TM86]

What the world needs is a radio final that puts out monster power and tolerates ten-to-one SWR with no more than a shrug.

Aren't there some LDMOS devices that fit that description? I mean, fit it if you happen to have 50V and gobs of current to drive them.

 

[Shadetree Mechanic]

Just my opinion here...

520's are a dollar a piece at mouser. The 1969's are 4x that price (at ICA) and legit parts are rapidly disappearing from the market.. I never had issues with mosfets blowing up or getting hot, but then again, I'm not asking them for 30w either..

Its not that the parts "cant handle" a little extra power, 520's are rated for 60w or 7A, the problem is that they are mismatched when you do "turn them up". For example, on AM, the collector (or drain) is sitting at 6v. The impedance of the final is approximated by E^2/(2P)=Z so if you have 6v at the drain and ask it for the 4w the radio is designed for, the impedance of the transistor is about 4.5Ω.

What happens when you ask it for 10w on AM? The transistor impedance drops to 1.8Ω! Unfortunately, the matching network between the transistor and the output filter is still designed to match 4.5Ω to 50Ω, not 1.8Ω. 6v/1.8Ω is still under the 7A rating, but some of the power is immediately reflected back into the part when it sees the 4.5Ω matching network and the probability of magic smoke goes up.

Every time the power changes in the transistor, the transistor impedance changes. Now I'm not accusing you of doing that, I am simply pointing out that if people stop turning them up, they will stop blowing up!

Edit: I used to have a smith chart plot of a common radio final section that showed how a mismatch at the antenna jack is transformed through the filter and matching network into a mismatch at the transistor (for comparison to the effects of power change), I am trying to find that....

Now you are making me think. I have often wondered what the impedance is between output stages like predriver to driver to finals. And how to figure out what is needed. I find RF power transmission fascinating.

 

[9C1Driver]

I would take a dual 1969 or 2312 radio any day over this mosfet stuff. Never blown any of the old school finals in a radio.

 

[brandon7861]

Now you are making me think. I have often wondered what the impedance is between output stages like predriver to driver to finals. And how to figure out what is needed. I find RF power transmission fascinating.

If I can't find my smith chart plot, I will do it again (and put it in its own thread so as to not hijack this one). I do recall from everyones favorite book that the most common strategy is to design the low-pass filter with 50Ω input and output impedances (to simplify the math) and using an L-network to match the low impedance of the transistor to the 50Ω at the input of the filter. But from what I have seen, its not always that straight forward (but a great place to start). I even have a short pdf discussing this very topic, I will add that to the smith chart stuff when I get to that (hoping tonight).