We're using a 50mW FM ATV transmitter. We'd like to amplify that signal to something in the watt range, so this means a power amplifier. Since it's FM, we can go nonlinear and get slightly higher output power, although it'll affect the output spectrum of the signal.
We are using the Mitsubishi M67715 linear hybrid power amplifier, which is still available, but not from Mitsubishi.
We purchased a PCB from Down East Microwave to make it easier to hook up: http://www.downeastmicrowave.com/2303PA.htm
For linear amplification, the M67715 requires a 5 mW input. For nonlinear, 10mW intput. So we want to drop the 50mW ATV transmitter from 50mW to 10mW.
According to Glenn:
An attenuator is needed in front of the power amp to set the input power level to 10 mW. There is an application note AN0015 from RF Micro Devices that has a table of Rp and Rs values for various attenuation levels for a 50 ohm Tee or Pi attenuator.
I measured the output of the ATV exciter and using a slug rated for 800 to 950 MHz, I measure 62 mW (+18 dBm), with a supply voltage of 12.17 VDC. If I drop the Vcc to 10.92 VDC, the power is 60 mW (essentially no change). I checked the picture with and without the preemphasis, and the picture quality is definitely better with the preemphasis, at least at indoor lighting levels. No change in power output related to preemphasis.
If I were to guess how to derate the wattmeter slug for operation at 1253 MHz, I would assume a 3 dB rolloff per octave?
800-950 = 150 MHz
Midband = 875 MHz
DUT is 378 MHz above midband, or less than 1/2 octave (0.43 octave?) above midband.
If it's a first order rolloff then 3 dB / octave, so were about 1.29 dB low?
So we should be seeing about +19.3 dBm which is about 85 mW.
Let's go with 85 mW.
So now our attenuator to bring this down to 10 mW for the power amp needs to be -9.3 dB
Using the chart on the RFMicro AN0015 page, the closest attenuation value is -9.54 dB, which is close enough, and so the R's are;
Rp = 100 ohm
Rs = 66.67 ohm
We want to use one (1) 1206 for each Rp, and two (2) 1206's in parallel for the Rs (because that will lie flush with the 3 mm microstrip).
So for a single attenuator we need two (2) 100 ohm 1206's, and two (2) 133.3 ohm 1206's (130 ohm will be okay)."
Accordin to DEM: "The 2303PA requires well regulated 13.8 VDC at 2A for full power output". Crap! We have a 12.0V out, 1.25A DC-DC converter for the WHOLE ATV system. What do we do now? Do we need to go Vbat through a linear regulator to get that?
The M67715 uses the "fin" as the ground connection. But we're using non-conductive white-heatsink grease for thermal contact? This can't be good. Does anyone have conductive heat sink grease?
What the heck is up with the ground connections on this board? The pads where their linear regulator sits aren't even plated through to the other side... but it's a plated through board. Are we SUPPOSED to connect DC ground through the bolt?!