@secret squirrel - depends on what you would consider "more output"...
Is That correct that the MOSFET conversion done correctly will increase output, or was he just doing work to increase his bill. This was 10 years back and the MOSFET conversion was the big new thing.
I would like to think that due to the loss of the Bipolar that you'd normally put in there. I would look into why you lost the Bipolar (The catastrophic failure or upgrade?) before you think about "more power" the man is telling you you'll get.
That is where many of us need to stop and rethink the situation...
We lost the Transistor to age, Obsolescence and a popular Mod using MOSFET - but particular types of MOSFETs' LD-MOSFET Lateral Diffusion
We can't just drop in any MOSFET in there and expect them to work like the Bipolar.
So your reference to the "Bill" is correct. They have no choice but to upgrade, But in some instances, the Upgrade is more of a problem than it is worth. It may have been easier to drop in the Bipolar even though it costs more, the conversion and the issues around the modifications needed to obtain the performance add to the Bill.
You already have a radio set up for AB bias from the predecessor Bipolar- so you're fine, many others though, are not and run the "self biasing" method that Bi-Polar's used in Cobra 29/25,s 19's Unidents' (PUN) PC-66-78 series, nearly all AM style used a method of biasing on the Final, that if you drive it harder, it will just move from Class C~D into Class C on up to nearly AB2 due to the nature of the audio drive and what's present on the Base lead to re-rectify thru that Biasing Resistor using the PN junction the Base provided as the Diode in which to do so.
- When it comes to the above statement about the Moving from one Class to another - refer to the "10-ohm Smoke" moniker - you force the transistor to work harder by "lifting" the Base lead above ground - the more impedance (Reactance ohmic) you use, the harder the device pushes itself into a different class - simply by a circuit it makes from the Blocking Cap, the Base junction and the Resistor placed across the junction even with Biasing present - let's get back to the moment though. The harder we drive RF into a Transistor - the more it will work to a point where it's saturated, and the distortion begins, but look at the results; the Transistor goes into - nearly Class AB or even A from that drive - even when clipped. It's now acting like a switch.
- This is where it gets harder to describe but the device when saturated, is not producing the pure waveform your want, but it is acting on the input to turn on like a DC switch if it's allowed to. You just don't get the faithful reproduction of input to output. How? By Drive and the problems with processing the power present at the Base - so you have made - caused - the Transistor operate more like a switch than a power RF device. You "swamp" the input with LOW-reactance low- impedance - path of offset the inherited rise (rectification) and keep the amp in a particular Class of conduction for a longer period of time in conduction angle as well as a necessary effort to remain faithful to th input and not act like a switch.
It's those EN devices that get us into the same trouble - the same type of biasing used for Bipolar is the mess that makes this upgrade to MOSFET not exactly a pure drop in performer. You have the input and output requirements - more on the input than any other effect, that being - Gate capacitance. So much energy is being stored in the capacitive reactance, it effects the device (MOSFET). The drive present and the rectification the EN devices add, into a working voltage lifts itself (MOSFET) into a level of conduction the same way the Bipolar can - and skew the waveform - hence the Rise and Fall rates are a necessary calculation factor to keep the stage operating like an RF amp, not a DC switch. You also have to keep the Gate capacitance low enough so you don't run into the troubles with skewing and Non-linear operation - LDMOS is not foolproof - they are not Bipolars.
- The bad thing about Bipolars are the very thing that makes them more endearable - they don't provide "tons" of power, but they can shift their Class easier than a MOSFET - as witnessed by the level of problems generated by it's switching versus linear tracking Bipolars are notorious for.. (Swinging at the Knees)
- Shifts to a higher class was easier to obtain in the Bipolar than in the MOSFET - we simply tweaked the window of drive (Admittance) and the amount of current needed to turn on at the Base and the level of RF can drive the Bipolar itself into another class when overdriven. Even self biased, a Signal present and applied, the Bipolar would simply push itself into a higher class, MOSFET's don't - their linear region is far smaller and act more non-linear.
A good example of this is some of the issues around MOSFET conversion to drive AMPS versus the Bipolar - and using MOSFET output to drive Bipolar AMPs. The windows they operate in are not the same and when overdriven with too much power - even the 10-ohm resistor is not enough to swamp out the saturation event and the Bipolar is driven into a switch mode more than RF signal amp..
It is easier (Safer really - the SOA) to use a Bipolar to drive an MOSFET stage or amp than to have an MOSFET drive a Bipolar stage or amp. Simply due to the dynamics of power level performance related to the Safe Operation Area - SOA.
So, I'm sorry to see you having to use a upgrade on a chassis that was otherwise idealized for the Bipolar generation. Many will say the upgrade was well worth it, while others may say that you should just get a newer radio with them already in there. But if you look at how the idealization of, or how the optimization was done - the MOSFET may be efficient - but the power levels in both Drive and Output between Bipolar and MOSFET are different and require different sets of support for them - they are not exactly drop ins.
