Looking for feedback | Lime SDR + PA rig for FM band

[vu2tve]

[x-post from QRZ forum]

Hi. This is my first serious R.F. project, so with very little or no background experience in building high power R.F. stuff, I'm looking for some feedback on this effort.

In the attached photo, there are the following components
1. A limeSDR USB in its own enclosure
2. It's TX output is connected to a 1W linear PA that I purchased from AliExpress a long time ago, and it has good performance
3. The small amplifier is connected to a MRF101AN based amplifier, and a tiny bias circuit (which has a diode thermally coupled to the MOSFET). This is capable of putting out 100W, and will be biased to run in class-A (hence the ginormous heat sink)
4. (Missing) The output of this big amplifier will go to a Bandpass filter
5. The big amplifier needs a 50V DC power supply, which I am using an off-the-shelf SMPS for
6. The small amplifier needs a 12V DC power supply, which I am using an off-the-shelf SMPS for
7. (Missing) I guess fuses in various places, and some power switches, and perhaps some LCD display modules for temperatures, voltages, etc.

I would like some feedback on
1. The R.F. part of the circuit.
2. The power supplies being used
3. Optimum design for an enclosure. Right now, I just used a wooden board as base since it was easy to screw things down to the board in case they need to be moved later. Once things become more final, then I can build a metal box for this.
4. Anything else glaringly obvious that I missed, and any other tips.

Thanks, and have a wonderful new year 2022

73 // VU2TVE

[vu2tve]

- Thanks for the pointer about USB3 isolation/hub. USB3 powered hubs are a lot cheaper than isolators. Would they work?
- The alligator clips will go and will route power wires away from the SMA connectors as you say.
- I was making a BPF for the output stage, but will make another BPF for the stages between the two stages.
- Add an inductor on the 12V power supply input; You mean something like a choke? Any pointers to what kind?
- The heat sink for the big amplifier is a huge (& heavy) aluminium heat sink. If it doesn't seem to do the job, i'll add a fan before moving to copper, which is difficult to find.
- I'll shield the amps.

Thank you. These are very helpful tips!

[radiolistener]

2. It's TX output is connected to a 1W linear PA that I purchased from AliExpress a long time ago, and it has good performance

I have this "2W" Chinese "amplifier". It is NOT LINEAR and have a lot of distortion when you trying to use it at 1 W power. Here is my measurement.

It is acceptable if you trying to use it for CW carrier and have good low pass filter on it's output. So, there is your big mistake, you're forgot to put low pass filter after this "amplifier".

If you want to get linear result from that "amplifier", you're needs to put max -30 dBm on the input and the output will be about 0.031 W. If you exceed -20 dBm, your output will be very distorted. The output power at -20 dBm on the input will be about 0.44 W.


Regarding to your installation, I think it can be RF jamming device to make interference for neighbors electronics. Because it will have a lot of distortion, high out of band emission and harmonics. Is that your goal? Note that this is illegal to use RF jammers and government easily can locate your position and confiscates your equipment. And you will be fined. If your jammer make RF interference for aircraft communications, this is criminal and you will be arrested.

Therefore, I advise you to not carry out such experiments if you are not sure that the amplifier gives a clean signal. Just don't try to attach so powerful output to antenna if you didn't tested if for harmonics level and out of band emissions.

[vu2tve]

As from your graphs, I've observed something similar with this particular amplifier. At low input power, the signal is much cleaner than at high input power, and my intention is to use it with low input power.

From your and other previous comments, it does seem like putting a LPF would be a good idea, so I will do that. I have not sent radio waves into the air with the big amplifier connected to the chain, but sent it to a dummy load (not shown in part of the setup).

I understand the illegality of sending high power and distortion "radiation" into the air, and am a licensed ham operator so am aware of laws in my region but still, thanks for pointing those out.

Perhaps I can delve into the application a little more. The sdr will produce multiple FM carriers (between 6-10), and my required power output is ~1W per carrier, but the laws of physics dictate that I have an amplifier "capable" or producing 100W output even though the avg total power in case of 10 carriers will be 10W. This is a specific project application in a specific location that I am building for, which I can clarify in a private post if you desire.

[vu2tve]

I would eventually hopefully be looking to replace the chinese amplifier with a custom built one, and integrate into the same PCB as the MRF101 amp and include the biasing circuit as part of a single pcb. but that's for a future iteration.

[radiolistener]

From your and other previous comments, it does seem like putting a LPF would be a good idea

This is not idea. This is mandatory must be.

Without filter all harmonics will be amplified with second stage amplifier.

[vu2tve]

Another thing I was thinking was to atleast have a regulated linear power supply for the 1st stage. I feel fairly confident to build that.

But a linear supply for the 50V stage, I feel less confident for, so the SMPS will have to do for now.

[Theboel]

Just my 2 cent,

1. LPF is a must
2. Do not drive that PA more than 700mW

now my curiosity the cost of lime SDR is not cheap compare to "well" made PLL board with 5w output and lime has weakness about durability (my own and somebody else experience) so You must plan very special thing. did You ?
mind to share ? 

[9aplus]

When Lime SDR USB is properly cooled, than can be reliable.
See the picture with Lime full right.
Oversized metal box with whole lime board contacted with blue heat transfer compound
and with passive coolers on Lime chip and FPGA. That keeps the temperature within 45 C range.

My recommendation is to first generate needed signals with Lime than to see that on spectral analyzer.
When happy to add BPF before any next amplification stage.

Here own 2 x Lime SDR USB and 1 x Lime NET micro

[Theboel]

yes yes yes I can not agree more heat is Your worst enemy with lime