Mrf9180 Keeps blowing!!

[XXL]

Can anyone help me with these shit devices. They just die so easy. 1st one decided to die just 3 days later of being on, 2nd one was an NXP type, blew for some reason when I put the audio cable into the RCA socket and a third fake which was running very nice doing 150w warmish blew when I switched it off. WHAT THE ACTUAL FUCK. the amp is a nice design, the devices are shit. I run the bias around 3-3.5v and input 500mw for up to 170w out, but usually turn em down to 150.

[Bton-FM]

Maybe try the Freescale ones from enigma shop I’ve heard these are rock solid.

[XXL]

yep tried that, still popped just by switching on and off.

[Electronically]

Set bias at 2.6. Make sure power supply is around 21v 8amps at least

Sent from my AMN-LX9 using Tapatalk

[Electronically]

Don't run them on 26v

Sent from my AMN-LX9 using Tapatalk

[teckniqs]

Yeah it sounds like the Bias and input voltage is being over-driven, as far as I'm aware you shouldn't be able to get 170w out of them with just 500mW drive, so if you are turning the Bias up to achive this then it's turning it up too high above the limit and probably the reason why it's blowing all the time.

...Always start low and very slowly work your way up when it comes to FETs.

[Albert H]

I run them from a 300VA 15V secondary transformer, which when rectified and smoothed gives about 21V DC. The bias supply is taken from the 21V rail through a well-filtered 78L05, with three silicon diodes in series with the output to give about 3.1V at the top of the bias presets. I use multi-turn presets. This makes it easy to get the bias dead right (usually about 2.45V - 2.55V). Always start low!

Also I have a protection circuit across the bias trimmer - just a small 2N7000 or BS170 FET, with a C/R network in its gate, so that at power-on, the bias voltage is virtually zero, and over the next 10 seconds (roughly), the FET goes slowly out of conduction, and the bias voltage comes up smoothly to the value previously set on the preset. My "high SWR detected" circuit also turns on that FET to kill the bias to the PA and (hopefully) save the expensive output device before it suffers!

[XXL]

Yeh I always set the bias voltage low before bringing up the voltage. It does run really nice at 26v and 3.5v but just keeps blowing. I thought this was a 26v device. I'll try and use more drive with lower voltages. Thanks

[teckniqs]

I wouldn't say for one moment that it runs "really nice" if it keeps on blowing, that is very contractionary.

.....No, it's a 24v device.

[Albert H]

I found that even at 24V they could be fragile. Running at a little less than that means that they need a little more drive, and the bias is more critical, but they're much less likely to spontaneously die.

[BriansBrain]

Well,
I have been running my Chinese board with a replaced Genuine Motorola MRF9180 in it for the last 4 months @ 16/½ hrs a day

@ 26V @ 8A = No Problem

150W Linear 26V @ 8A-S.jpg

RF Rack-S.jpg

Motorola.pdf - MRF9180.JPG

[Electronically]

Brian it says 26v but it's not recommended to run it at 26v you might say you've been running it for 16 odd hours a day. To me I wouldn't even dream of running it at 26v a day. I'll stick with with 21volts that way it will last a lot longer than your 26v.

Sent from my AMN-LX9 using Tapatalk

[BriansBrain]

As you can see from the date on the PS...

It was put on Brian's RF Rack 30/11/2019

That's exactly 18 weeks ago or 126 days = to 2,079 hours ago and it's still going strong

[Electronically]

And you've never popped that chip?
I'm guessing that you have your bias set around 4?

Sent from my AMN-LX9 using Tapatalk

[thewisepranker]

You can't just quote the 26 V on the datasheet without looking at the numbers directly next to them, those being "at 880 MHz".

When taking a device well out of its specified range, you've got to bear in mind that there is no data to help you design the thing. There are two ways to go about it - either get lucky or be lucky. You can't get lucky, there are too many variables, but you can be lucky enough to have both the equipment and knowledge to be able to plot the data for yourself, which is what the people that came up with these FM band designs had to do.

[BriansBrain]

And you've never popped that chip?

No it's still going strong

I'm guessing that you have your bias set around 4?

I don't set bias using voltage measurement
I set the bias using the current draw of the board.
With no RF drive I set the bias at +/- 250ma when the amp is cold +/- 21°C
Which creeps up to +/- 425 when the unit gets hotter (no RF drive)

[BriansBrain]

You can't just quote the 26 V on the datasheet without looking at the numbers directly next to them, those being "at 880 MHz".

When taking a device well out of its specified range, you've got to bear in mind that there is no data to help you design the thing. There are two ways to go about it - either get lucky or be lucky. You can't get lucky, there are too many variables, but you can be lucky enough to have both the equipment and knowledge to be able to plot the data for yourself, which is what the people that came up with these FM band designs had to do.

Well...
I seen 26V on the datasheet... so I wen't with 26V because the device must be able to cope with 26V - and it does

I also talked to a couple of RF engineers I know from the past (1980's)...
They told me 26V will be fine these devices are tough when it comes to voltage and current draw
They also said... it's the gates you have to be careful with.

Which is true...
The only times I have killed any MRF9180's is some form of unwanted surge to the gates

So in my personal tests...
With a Genuine Motorola -or- Freescale MRF9180 26V @ 8A is OK if you keep it cool

MRF9180 - Genuine & Fake.jpg

[ronald001]

How many windings are your coils in the output filter?
I have the same kit - and measured my filter ( 2WDG per coil ) and it looks very bad.

I had to change the coils to get a acceptable impendance / SWR

[BriansBrain]

How many windings are your coils in the output filter?
I have the same kit - and measured my filter ( 2WDG per coil ) and it looks very bad.

I had to change the coils to get a acceptable impendance / SWR

Mine was not a kit it was a ready built board...
As you can see my LPF has 4 coils + each coill has 4 windings

IMGA0576.JPG

[teckniqs]

Can you see what the values are for the filter caps?