Batboard

I recently pulled one out of the rack to do a retune for a different transmitter frequency. Having not had access to a network analyzer before, I decide to check it out and was surprised to see the insertion loss was around -2.0dB to -2.5dB

Put it on a tracking generator and followed the standard alignment with the test probe from J18 to J16 and still got the same -2.0dB IL.

Can anybody else confirm or deny they have similar loss through one of these MSF5000 Tx filters? If you've had better performance, are there any tricks besides or in addition to the factory alignment procedure? There seemed to be too much interaction between each cavity to do it by "eye" on the NA

Thanks.

I'm assuming you are refering to the UHF internal pre/post filter. That is about normal. It is not a real effecient setup. Some will remove the internal filter setup and go with an external Bp/Br set of cans with a little less insertion loss. YMMV.

Gary

That's about what I got also. I did find that using a spectrum analyzer and tracking generator would not give me the optimal setting; I had to use a sig gen and RF voltmeter and go "by the book" to get that real sharp peak. Perhaps other equipment would do it better.

The pre-filter loss is probably around 1.5dB, but that's overcome by the more-than-plenty IPA output.

The internal filter/duplexer, along with the circulator, tends to drop the output power by 3dB, so if you start with a 110w PA, the station is rated for 55w to the antenna.

Bob M.

I agree. By the book gives the best results, if only using a SA and a tracking generator. I acquired a return loss bridge (spendy item!), some time back, and have had the best results with it. Just need to make sure you are using a quality 50 ohm termination at the output side of the filter(s).

Appreciate the feedback and info. Yes this is for a UHF MSF5000... sorry I forgot to mention that. I edited the Subject field to now say UHF.

Guess I'll either have to live with the insertion loss or go with a true duplexer or set of Bp cans.

Thanks.

On base stations, they use an N-female barrel connector between the IPA and the PA, and a low-pass filter between the PA and the antenna relay. The filter is a 10 inch long tube with female N connectors on each end that has a nice cutoff around 620 MHz.

I used a base as a repeater by bypassing the antenna relay and feeding the output of the LPF to a duplexer mounted on the back of the cabinet. I'm feeding 80 watts into the duplexer and getting 55 watts out. So the loss isn't too much better than the internal filter/duplexer would give. But the external one is a 6-cavity Bp/Br Celwave base duplexer that has exceptional performance, and leaves very little for the front end to complain about.

I have a return loss bridge but didn't bother using it since the four sections of the post-filter interact with each other so much. You definitely can NOT adjust each filter more than once; if you go back, you find different tuning positions and end up chasing your own tail. If there was a way to decouple the stages from each other, the RLB would be the way to go. But when done carefully according to the book, I don't see why the results and performance should not be just as good.

Bob M.

Well, for sake of an argument, I may have been just lucky with the
RLB. I have only tried it once as an experiment. I started at the front of the filter and went forward (ONE, AND ONLY ONE PASS!!). I ended up with a couple of tenths of a DB better than I arrived at using the probe and SG.
I used a Bird termination load. I preset the filter slugs by the book as if I was doing a large frequency move in both tests. This was using a spare Pre/Post filter that I have. I did exchange the filter in the UHF machine I have running here in my shack. I run some on air tests (RLB tuned version), and could not tell the difference with the one that was in before. The original was a "by the book" tuned P&P filter. Again, YMMV!

Gary

My two bits worth:
Everything has a trade off.
In this case the MSF is synthesized with 15 channels.
This would imply that the bandwidth of all filters must pass all frequiencies within a 1 meg window.
The transmit filters are windows to allow a large spacing between channels with out loss of performance.

This would not be easy with a conventional duplexer.
One member used a few cavities on the transmitter with a "T"made with 1/2 wl cables to the receiver and antenna port.
workes well with no desense.

The duplex option costs 3DB in transmit power.
The basic MSF5000 UHF R-ll is rated at 110W with the single circulator.
The factory repeater has a tripple circulator as well as the filter assembly.
The output is rated at 55W.
It is possible to force more power, excessive power causes the solder joints in the circulator to eventually fail.


Proper termination of ports during alignment is essential for proper performance. always measure desense.

I have a bunch of these filter kits.