Thanks for stopping by to let us know...
Just as a caution - if you haven't already read enough about it...
I'm finding a lot of MOSFET conversion from the older 2078' Transistor to the Newer MOSFET - there are some "set ways" and some that aren't so "Set" ...
The above posts are meant as a Guide line for you, (I know) - but to each their own.
As to the "Why" of this follow up, I've have been repairing others' works. At least in replacing the final - they use specific types and I've seen just about everything.
The ones to caution you on, are...
The 13N10's and the IRF520 - they aren't the problem per-se, but it depends on what one is driving the other...
Some older Galaxy dual finals radios have use the IRF520 with no problems except they produce crappy amounts of backwards swing - a lot of that has to do with the EN parts used and their drive levels (as well as their more dedicated rework to interface their boards to the new line of thinking regarding BJT's to MOSFET from Current to Voltage). It's a little more complex than that, but the CRUX of the problem is the Bipolar (Predriver section) to MOSFET Final/Driver - so if you decide to go on your own - the "TX voltage feed Bias" method works ok, but it's trigger happy.
Keep your Gate voltages to around 3 to 4 volts per part
- - try not to "Share" the same feeder resistor
- - give each MOSFET their own power feed - don't tie it's Gate bias with The other.
- Use a resistor divider from the power feed buss line - your' 8 Volt TX line, for EACH MOSFET.
- Give each MOSFET their own source feed resistor
- - 8.2K or 6.8K or even 7.5K from the 8V buss is fine
- - just remember to use a means to take off voltage - so that means you'll need a bleed Down resistor from the Gate too.
- Value of the resistor value of the Tap off from the divider? Well, a simple buffer resistor just to keep things from going catastrophic on you is fine - no big values - even 1K is a bit much but then anything above 100 ohms keeps RF to the part (MOSFET), not feeding back into your divider network - helps. Think of it as, What Happens In Vegas, Stays in Vegas...
The EN parts like the 1230 or 369's are less problematic but then have LIMITED levels of power they can process. They can't handle a lot of swing from the Driver, for a purpose of using say a "swing mod" or you're redesigning the Final to drive a High drive amp to drive it with swing. EN's aren't made for that - they'll "pinch up" your audio and aren't the best as making MOSFET's work linearly - produced like what Swing Mods are. You'll find the power feeder resistor divider works better for this.
The 369 part designs, so you know, were designed for BIAS feeds, so they needed a voltage - albeit a trickle one. Read this as a voltage presence with hardly any power behind it for drive current.
The 1230 part was made for AM and FM Class C and Class D stuff, not really meant for SSB.
The EN DR-FN pair as well as the 1230 - were designed for PRESET levels of RF they can use; detect and process for Gate voltage - with limited swing range in power. Look at it this way, you can put a restrictor plate on a car, and hire a driver to drive in circles with it all day, but they are meant for specific levels of speed. It takes a little more effort to obtain maximum power - that delay as well as the conditions of the turns in the track - make the effort to maintain a set speed harder or more difficult to do than to use a vehicle with no restrictor plate to limit the "air flow" of the normally aspirated ones.
Translate RF CARRIER to speed - RF being sent to the the Gate is that "speed" - so AM carrier and FM are a type of envelope and emission to provide the Diode in these little dittys' a means to rectify and produce a DC voltage to apply to the Gate and keep it on for the time you transmit.
But they can't throttle back power if you use a lot of high levels of modulation - they simply keep rectifying the power present and applying more and more to the Gate.
The IRF520's in a typical Galaxy use 5.1Volt Zeners for this reason - to limit power delivery and "Clamp" if it's exceeded.
Which is what brings me to the 13N10's debacle...
Found several radios with the 13N10s' blown but checking their Bias setting, found them to be pretty close to normal - a little high - but necessary for the Audio not to sound pinched or scrunched up from not enough Bias drive voltage. Fine when you run FM or limited Modulation (READ: LIMITER INTACT AND SET TO 70% MOD)but their Bias can be set too high for swinging their drive levels - they'll trigger on, with RF present, BUT REFUSE TO TURN OFF until they blow up (Latched ON). This problem is why I suggest the "buffer" and Bleed down resistors - because once the MOSFET gets warm, it gets a little too sensitive in the Gate region and tends to want to turn on and stay on until it cools down.
EVEN WHEN GATE TX VOLTAGE IS REMOVED.
Can't stress that enough.
The 13N10's are really good, but are SLOWER, read this as a different type of response curve.
The 13N10's Rise and Fall times give you the "slope" of 27MHz's RF wave, so as they Rise (Crest) and Fall (trough) they look and appear wonderfully as a simple clean Sine wave. Don't worry, that is a good thing - it's lets people get comfortable with the idea that your part, and the Maximum Useable Frequency - are nearly identical.
Ok, with me so far?
The problem is with Drive levels - the preceding stages leading into the Gate of a 13N10 may contain parasitic spurries and harmonics - which are bandwidth problems. (Like Heavy Modulation) The 13N10 can't handle these very well and actually can fail based upon a frequency response problem of exceeding their ability to handle RF in the spectrums above their maximum useable Frequency.
- Remember this: The useable frequency and the extra spurious frequencies and their drive levels produced - are rectified by the diodes' singular function in the EN devices to provide a DC Gate turn on. IT's can be quite higher in voltage than the Gate can properly use. It is why they use a resistor in series in line and as well as a paralleled Bleeder Resistor designed to take some of the charge away at a given rate of discharge.
- So when you use Heavy Modulation; that drive level of audio is RECTIFIED along with the RF carrier present - so if you sum the two together, you can get quite a large amount of drive level - enough that it can exceed the Gates Input voltage and or Current - which is seen as capacitance - you can spike the Gate with more power and damage it.
Remember they are slow(er) than other MOSFET's. So think of these extra, higher frequencies as a "ringing effect" and looks like a Gate Capacitance problem - like it's seeing excessive Gate capacitance and so they "latch" and refuse to "Fall" and blow up in the process.
This is why I posted so many messages in this thread, I want you to have THE BEST success in getting this to work with far fewer failures than from what I've had. You're seeing all my notes and what I've learned from these so far.
