High Reflected Power Puzzle

FM does it with frequency!
User avatar
jthorusen
Posts: 85
Joined: Sun Sep 02, 2018 4:35 am
Contact:

Re: High Reflected Power Puzzle

Post by jthorusen »

Well, the MT300 is back at the station. I have not tried it into the antenna yet (I have to work out some audio splitter and rack mounting issues before I can make a switch back and forth easily). I did create a +j reactive load to test with and the Broadcast Concepts pallet performed flawlessly. No self oscillations detected. It works like a champ. I have finally finished the proof measurements for spurs and harmonics and the transmitter is very clean. The third harmonic is actually the worst and there is a 3.6 dB margin even there. I cannot recommend Broadcast Concepts more highly.

I spoke with them concerning type acceptance and their position was that even if you have a type accepted transmitter, if you are radiating where you are not supposed to, that's bad from the FCC's point of view. On the other hand, if you are clean, then that's what they are looking for. I would add my own attitude that what parts you had to replace to get there are not significant, so long as you got there.

For DaveSt: The FM300-108 pallet is now a problematic item at best. The replacement SD2932 that I got from Mouser could not be biased into conduction with the on-board bias network.... component values would have had to be changed in order to get it to draw the specified idling current... and even then, there is no guarantee that this would stop the self-oscillation. My guess would be that it would not; there is a shutdown circuit in the transmitter that removes the gate bias when reflected power becomes excessive. This circuit was fully operational, but since the FET was already biased off, it had no effect on the self-oscillation. SD2932 transistors are no longer being manufactured, according to what I have been able to learn, and it would appear that the ones that are still out there are very different than the ones for which this pallet was designed. Therefore, since a replacement SD2932 from Mouser is about $140 and probably will not work in the circuit, while a complete SD2942 pallet from Broadcast Concepts is about $308, the proper repair technique would seem to me to be pretty obvious. Of course, this is with 20-20 hindsight; I didn't know this at the time. Hopefully, I will spare someone else some grief by posting the results here.

Regards,
James K. (Jim) Thorusen
KB6GM
Central Coast Electronics
www.centcoast.com
NW Oregon Consulting Bdcst Eng.

User avatar
Deep Thought
Posts: 3536
Joined: Thu Mar 20, 2008 9:23 am
Location: La Grange, IL
Contact:

Re: High Reflected Power Puzzle

Post by Deep Thought »

jthorusen wrote:
Sun Feb 03, 2019 4:47 am
I spoke with them concerning type acceptance and their position was that even if you have a type accepted transmitter, if you are radiating where you are not supposed to, that's bad from the FCC's point of view. On the other hand, if you are clean, then that's what they are looking for. I would add my own attitude that what parts you had to replace to get there are not significant, so long as you got there.
The FCC allows you to modify type accepted transmitters as long as you do the measurements to ensure they still meet the spurious and harmonic emission rule limits. If you modify the frequency-generating bits you also need to ensure that will remain inside the +/- 2 KHz carrier frequency limits over the expected operating temperature and supply voltage range. Keep the measurements on site.
Mark Mueller • Mueller Broadcast Design • La Grange, IL • http://www.muellerbroadcastdesign.com

DaveSt
Posts: 56
Joined: Mon Jul 23, 2018 2:21 pm
Location: UK

Re: High Reflected Power Puzzle

Post by DaveSt »

To properly shut down the amplifier with the bias, it is not sufficient to remove the bias voltage. It needs to be driven to a negative voltage of several volts.

For the harmonics, usually the output filter after the pallet will have elliptical sections to knock out the second and 3rd harmonics and a low pass section to sort out the rest. Sometimes small adjustments to the coils can improve performance on one frequency. So you might be able to get the 3rd harmonic more comfortably within the specification.

User avatar
jthorusen
Posts: 85
Joined: Sun Sep 02, 2018 4:35 am
Contact:

Re: High Reflected Power Puzzle

Post by jthorusen »

Sounds reasonable. I did not dig into the shutdown circuit to see what it actually did; all I know is, it wasn't enough.

The low pass filter and directional couplers for forward and reflected power are in a sealed can. Opening it up would be a lot of trouble, so I think that the 3 dB will have to do. I think that some of that is in the spectrum analyzer, anyway. I did not have a notch filter sharp enough to look anywhere near carrier, so I just padded down the power after the directional coupler. In looking far from carrier (up around the 5th and 6th harmonic), the signals were in the noise floor so I reduced the pad value to look there, which changed the ratio between the harmonics and the carrier slightly. Therefore, I think the 3.69 dB actual measured safety margin for the third harmonic represents a worst case situation. It may well be better than that, although probably not much.

I used the FCC formula (43 + 10 log10 (power in Watts) to derive the value of attenuation required. I then calculated the actual permitted power in dBm represented by that level below carrier and it turns out to be -13 dBm. The actual measured values turned out to be:

2nd = -17.23 dBm
3rd = -16.69 dbm
4th = -30.14 dBm
5th = -28.54 dBm
6th = -41.9 dBm

About the variation one might expect from a push-pull amplifier; even harmonics slightly better than odd ones.

I think she'll do!

Regards,
James K. (Jim) Thorusen
KB6GM
Central Coast Electronics
www.centcoast.com
NW Oregon Consulting Bdcst Eng.

Post Reply