Are these amps any good?

[LeeCavanagh]

Are these amps any good?

9D242737-8455-4F4C-B04B-E0B06016BD43.jpeg

Soon to be available on Dutch rf shop

[Bton-FM]

Are these amps any good?

9D242737-8455-4F4C-B04B-E0B06016BD43.jpeg

Soon to be available on Dutch rf shop

It looks very similar to the NXP design with 2 MRF300.

[LeeCavanagh]

Should be good then, but may get hot?

[Albert H]

Two of our guys - LB & CD - have been working on MRF300 PAs. They report that they work really well, but getting rid of the heat is a huge problem. They've used the biggest heatsinks we have, and used blowers. They used an infrared camera to plot the heating across the metalwork, and found that the transistor package isn't good for heat dispersal. They found severe hot-spotting at the transistor pad, and lots of heat transmitted down the device legs, in one case causing the solder to melt!

[Bton-FM]

Two of our guys - LB & CD - have been working on MRF300 PAs. They report that they work really well, but getting rid of the heat is a huge problem. They've used the biggest heatsinks we have, and used blowers. They used an infrared camera to plot the heating across the metalwork, and found that the transistor package isn't good for heat dispersal. They found severe hot-spotting at the transistor pad, and lots of heat transmitted down the device legs, in one case causing the solder to melt!

Here someone used a copper heat spreader solve the problem:

https://qrpblog.com/2019/10/a-600w-broa ... s-devices/

[Albert H]

They tell me that they tried a copper heat spreader (which cost almost as much as the damn transistors!). It helped, but there was still hot-spotting in the devices.
Incidentally, if you reduce the bias, so that they're working closer to Class C, the dissipation goes down markedly, but the purity suffers - once again, we get to a compromise!
With reduced bias and an extended lowpass filter, they find that they can run a pair of these devices on a reasonably sized heatsink at 500W out. That's going to be good enough for most purposes - the field difference between 500 and 600 Watts is very small - 600W just looks better on the power meter! At the reduced power level, the hot-spotting is much reduced.

[LeeCavanagh]

Interesting, thanks.

[LeeCavanagh]

Do we know how many AMPS this amp needs? Just wondering if my existing 10amp 50v power supply will be enough?

[Bton-FM]

Do we know how many AMPS this amp needs? Just wondering if my existing 10amp 50v power supply will be enough?

It’s best to try and get a power supply that can deliver twice the output of your amp - a 1kW supply.

1000W/50v = 20 amp power supply. No amp is ever 100% efficient so you have to supply it with more power than it outputs.

[LeeCavanagh]

Do we know how many AMPS this amp needs? Just wondering if my existing 10amp 50v power supply will be enough?

It’s best to try and get a power supply that can deliver twice the output of your amp - a 1kW supply.

1000W/50v = 20 amp power supply. No amp is ever 100% efficient so you have to supply it with more power than it outputs.

Ah ok, thanks , i think its a 750w power supply , but these these are often miss rated i hear, so two in parallel would do just fine then. Cheers.

[rigmo]

Two of our guys - LB & CD - have been working on MRF300 PAs. They report that they work really well, but getting rid of the heat is a huge problem. They've used the biggest heatsinks we have, and used blowers. They used an infrared camera to plot the heating across the metalwork, and found that the transistor package isn't good for heat dispersal. They found severe hot-spotting at the transistor pad, and lots of heat transmitted down the device legs, in one case causing the solder to melt!

MRF300
Sure problem with temperature dissipation stay. But good match on input and output get high efficiency, less heating
yes right, If not know whit you dealing with... drop output power...