https://www.nxp.com/docs/en/data-sheet/MRF101AN.pdf

rigmo wrote: ↑Mon Dec 26, 2022 6:53 pm

XXL wrote: ↑Mon Dec 26, 2022 3:52 pm Single. Theres no need for 600w. 300 is fine.

Doesn’t work well and no filter.

XXL wrote: ↑Mon Dec 26, 2022 7:55 pm Doesn’t work well and no filter.

the only one good is this one

No gerba files anywhere, can’t buy it from anywhere, and can’t copy it because it’s got fancy coils on it with no parts list.

The coils are made from copper tape, which is really expensive, but handles a lot of power and radiates heat well.

Well if anyone makes one up I’ll happily try it. Even if it only did 200w I’d be happy.

XXL wrote: ↑Wed Dec 28, 2022 3:47 pm Well if anyone makes one up I’ll happily try it. Even if it only did 200w I’d be happy.

I dunno if this is much help XXL?

I am hesitant on commenting on high power amplifiers?

Anyways, you can purchase those PCBs from NXP but they are very pricey! The full parts list is also listed on the NXP datasheet!

The dimensions are also included, so with some careful measurements a design could be replicated at a fraction of the cost!

You do not need to use fancy coils either! Just keep in mind the operating ratings!

A low pass filter can also easily be added! Again, keep an eye on the ratings!

jvok wrote: ↑Mon Dec 26, 2022 5:19 pm

Do you heave schematic

i have had the run 120w all day running cool also

rigmo wrote: ↑Tue Oct 28, 2025 8:00 pm

jvok wrote: ↑Mon Dec 26, 2022 5:19 pm

Do you heave schematic

https://community.nxp.com/t5/Other-NXP- ... anguage=en

I knew I seen it somewhere. That's the phillips 88-108 and the files are .dxf. Enigma's version is available here:
https://enigma-shop.com/products/mrf101 ... 4287011106

I really don't have any time for those useless, expensive 3-terminal fuses!

The fundamental thermal design of the devices is appalling. Motorola are banking on high efficiency, big heatsinks and ideal loads. Unfortunately, in the Real World™, these conditions are seldom met.

My current PA for this power level uses the old MRF317, driven by a 2SC1971 and needs around 500mW drive for around 110W output. A real workhorse design. The reasons for this line-up are:

1. Cheap - we have large stocks to use up of both devices.
2. No tweaks needed - it's a substantially broadband circuit.
3. It's stable - it doesn't care much about the load it "sees", and doesn't do erratic things when poorly matched
4. It's utterly reliable, clean, and quick to construct.
5. It's reasonably efficient in the configuration we use. It doesn't have the wild gain figures that you can get from some modern FETs, but it also doesn't require neutralisation like many FETs do (bringing back memories of valves!). We're using two stages to get ~23.5dB of gain - about the same as you could get from one of the newer FETs, but without the unreliability.

We use temperature-moderated biasing, so if the load gets really out of hand, the PA output power diminishes! The minor drawback is that we need two supply voltages - 28V for the '317 and 15V for the '1971. Fortunately, we also need 15V for the exciter, so it's not really an issue. One further benefit is that 28V PSUs tend to be cheaper than high current 13.8V types, so we're cutting costs all 'round.

When I first saw the MRF101 and its specifications, I thought "that's ideal". When I tried out the first prototypes (and examined them closely with an infra-red microscope) I saw the flaws in the concept, and the hot-spotting that causes premature device death!

If you really want 100W and a 13.8V supply, use a pair of BLW60s or even MRF238s. You'll make lots of heat, but the devices are really rugged!

Please come on Albert......
You have always the best designs in the world but you not showing any of your Transmitters.
The MRF101 is a realy good device and in the "enigma"design it runs cool and stable.
For what i think you never tested this design or a MRF101 mosfet.

Sadly, you're wrong. The MRF101 (and - worse - the MRF300) are stupid designs. Do the basic mathematics: the size of the pad behind the '101 is too small to get rid of the 30 Watts of heat it's going to produce. It's marginal at best. The '300 is even worse, and can NEVER get rid of the heat it makes. That's why no real commercial manufacturers have taken them up.

Why do you think that serious manufacturers use FETs like the BLF177 or MRF151? Those packages are designed to dissipate their heat easily, and I've had rigs running 24 / 7 / 365 for over 25 years using those "old" parts. We've had (as I recall) three failures - all due to nearby lightning. (I just went to South America to replace a failed MRF151G rig with a valved unit with a pair of 4CX250 output bottles, since valved PAs are much less susceptible to static damage.)

Talking to the rep from NXP recently, he told me that the MRF300 will be discontinued in its current package format since users "have complained about their unreliability". When the new format parts become available, I might try them again, but still have yet to be convinced that low voltage devices are a good idea, since the current draw is ridiculous.

Incidentally, Head - I tried MRF101s before you even knew they existed! The "Enigma" board is a pretty good design - within the limitations of the device - but his designs for better FETs are far superior. I tried those early samples, and blew a couple up. I completed a couple of PA boards, and we ran them into perfect 50Ω loads for a few hours. I examined them with an IR camera, and could clearly see the hotspots.... not good.

Je weet echt niet met wie je praat!

Hah, I have exploded about 10-15 MRF300 fets. Its nice fet, but case is just too small as Albert said, I had simple rig running 125W carrier, around 500Wpep on 48m band, fets bolted to the huge heatsink machined straigh in mill, aluminium L-profiles top of the fet to keep front cooler, and these fuckers still run 70-90°C at case temperature, heatsink just barely 40°C.

4hours fully modulated to the dummyload and thay die/explode.

Suono from Italy use MRF300 AN and BN mosfets in there transmitters.
They run a single MRF300 on 250W and a double on 500W and it's runs cool for 24/7 dat in and out.
I have some MRF300 transmitters running also 24/7 and no problem
But overdrive or mismatch on the Gate will be a problem and will blow your MRF (never overdrive more then 1dB)

for sale

Is the original NXP board better than enigma board ?

Not really - they both suffer from the use of a really poorly designed FET. Forget those ridiculous FETs - they're dreadful. If you want that power level, use a 28V or 50V device in a proper case that can get rid of the heat!

pukje62 wrote: ↑Tue Dec 23, 2025 8:37 pm Is the original NXP board better than enigma board ?

I'm assuming you're talking about the MRF101 boards as I'm not aware of Enigma doing an MRF300 board - please someone correct me if I'm wrong. Generally, the test circuits in the device datasheets are just that - intended for inspiration or performance measurement. As far as I can see, some of the NXP test circuits (there are multiple for different frequencies) don't have filtering, for example. Personally I'd go with the enigma board, it's designed for a specific purpose and again correct me if I'm wrong but I'm sure I read that it did well in NXP's own design challenge competition.

When I first saw this FET and its big brothers I was a bit taken aback by the packages but tried the enigma design with the '101 and yes it does perform well and overall, at a decent price. I've never considered the '300 as it just rubs my fur up the wrong way - can't see how you'd get a real world device to reliably run 24/7 at 300W out in that package. I think you'd be operating close to a few "red lines" and that any such design wouldn't be very forgiving. Also I know so many people with fragments of these that I'd be more inclined to use a more "traditionally" packaged FET for this power level. I expect it would be cheaper in the long run, especially if you're designing from scratch and tweaking.