Dealing with mixing products

AM Radio discussion. Directional arrays are FUN!
Post Reply
kcbooboo
Posts: 495
Joined: Sat Dec 10, 2011 4:53 pm
Location: East of the Mississippi

Dealing with mixing products

Post by kcbooboo » Mon Oct 12, 2015 6:07 pm

I've got two stations about a mile apart and 260 kHz apart. I'm getting a healthy signal on the sum, but nothing on the difference. I can hear station A's audio (5kw) on the mix and the FIM points to station B (1kw). They're both running the same brand of solid-state transmitter. Unfortunately this mixing product is not quite 60dB down. The FCC limit calls for 73dB down on all emissions more than 75 kHz away from the carrier, for a 1kw station, and this one fails the mask. This was discovered at the annual EPM spectrum and harmonic session.

Is a rather simple LC filter tuned to station A going to give me sufficient (20dB to be safe) notch/rejection if it's installed at the TX output on station B, or do I need more than one stage or something like a pass/reject filter? This is something that's best contracted out; I don't have the parts nor expertise to make something that would work the first time.

Suggestions welcome.

Bob M.

Kelly
Posts: 490
Joined: Thu Aug 21, 2008 5:32 pm
Location: Washington D.C. Area

Re: Dealing with mixing products

Post by Kelly » Mon Oct 12, 2015 6:35 pm

Mark could speak to this better than I, but my guess would be the need to install a trap for each station, trapping the other would be the easiest and least expensive.
Skype:kellyalford Twitter: @KellyAlford

User avatar
Deep Thought
Posts: 3364
Joined: Thu Mar 20, 2008 9:23 am
Location: La Grange, IL
Contact:

Re: Dealing with mixing products

Post by Deep Thought » Mon Oct 12, 2015 9:19 pm

Yeah, solid state transmitters are very susceptible to this. It's a pretty easy fix though and if you can figure out which station's transmitter is doing the mixing a simple three-component trap will get rid of it. If you send me the station call letters (via PM if you want) I'll work out the designs for you. 260 KHz is not that difficult to get more than 45 dB rejection.

Depending on the ATU design, it may be possible to integrate the trap into the existing network.
Mark Mueller • Mueller Broadcast Design • La Grange, IL • http://www.muellerbroadcastdesign.com

kcbooboo
Posts: 495
Joined: Sat Dec 10, 2011 4:53 pm
Location: East of the Mississippi

Re: Dealing with mixing products

Post by kcbooboo » Tue Oct 13, 2015 6:51 am

Thanks. I wasn't sure if it could be a simple two-component LC or if it had to be the three-component version. I was just inquiring at this point. The stations still have to decide what to do and how they want to do it.

Bob M.

User avatar
KPJL FM
Posts: 587
Joined: Fri Nov 16, 2007 9:28 am
Location: planet Earth, 3rd rock from sun

Re: Dealing with mixing products

Post by KPJL FM » Tue Oct 13, 2015 7:46 am

Since Murphy still works in radio, it might be a good idea to have both stations trap each other. There's nothing like spending time/money on one station, then immediately having to explain how once the first station was cured, the second station also has the problem (just at a lower level).
Trim to fit, paint to match, tune for minimum smoke.

User avatar
Deep Thought
Posts: 3364
Joined: Thu Mar 20, 2008 9:23 am
Location: La Grange, IL
Contact:

Re: Dealing with mixing products

Post by Deep Thought » Tue Oct 13, 2015 8:27 am

kcbooboo wrote:Thanks. I wasn't sure if it could be a simple two-component LC or if it had to be the three-component version.
You can do it with just two components if you have a way to compensate for the added reactance at the pass frequency. That's why I mentioned the possibility of incorporating it in the ATU or phasor common point network.

The good news is that since they aren't sharing the tower the reject side isn't working too hard and the voltages aren't terribly high, so relatively inexpensive parts can usually be used (10 amp coils, 10 KV caps) for lower-power stations (<=5KW).
Mark Mueller • Mueller Broadcast Design • La Grange, IL • http://www.muellerbroadcastdesign.com

kcbooboo
Posts: 495
Joined: Sat Dec 10, 2011 4:53 pm
Location: East of the Mississippi

Re: Dealing with mixing products

Post by kcbooboo » Sat Mar 24, 2018 8:09 am

I didn't do the EPM spectrums for one station in 2016 but I was asked to do it in 2017. Both are running B-E transmitters, the one producing the mixing is only 1kw directional full-time.

I looked hard to find the mixing sum but it just wasn't there. The carrier was 320mV/m which is certainly healthy enough. I asked the engineer if anything changed but never got a reply. I saw him a couple of nights ago and asked again. He said, matter-of-factly, they installed a $5,000 Kintronic filter between the transmitter selector relay and the input of the phaser. He said it's in a huge box nearly the size of the transmitter.

For a 1kw power capacity that seems awfully big and awfully expensive, but it definitely did the job, as I couldn't detect ANY signal on the offending frequency. I wonder what the notch depth really is.

I've got another station that has a similar issue with a neighbor less than a mile away but only 40 kHz away. That mixing product is 45dB down, which just meets the FCC 73.44 emission mask. Supposedly both stations have traps but I don't think the local one is doing a good enough job. Both stations are running the same DAX-1 transmitter too.

It's nice to get closure on some of these issues.

Bob M.

User avatar
Shane
Posts: 823
Joined: Fri Feb 01, 2008 12:08 am
Location: Omaha
Contact:

Re: Dealing with mixing products

Post by Shane » Sat Mar 24, 2018 10:54 am

Not too surprised at the size of the filter. I have a 70kHz space between a 5kw and a 1kw here. They are about a mile apart, too. The 5kw fundamental gets into the 1kw and mixes with its 2nd harmonic producing a difference only another 70kHz away from the 1kw in the opposite direction.

This is perhaps a more extreme case since the products are so close to the two fundamentals. (There is another product 70kHz down from the 5kw, too, which is the difference of the 5k’s 2nd harmonic and the 1kw fundamental. It got worse when the 5kw installed a solid-state transmitter.)

The filter is built into a full size seven-foot rack and consists of a trap for the incoming in series with a pass for the outgoing.

Lower freq’s than what you have there making components bigger.

The induction of the 5 into the 1’s tower is strong enough to draw sparks off the tower if it isn’t grounded. Anyway, the filter provides enough filtering to make the mix product be down >75dB.
Mike Shane, CBRE
---Omaha---

dd92251
Posts: 23
Joined: Sat Mar 30, 2013 4:21 pm

Re: Dealing with mixing products

Post by dd92251 » Sat Mar 24, 2018 4:00 pm

The Series-Pass/Trap is a pretty common solution. And... it's used as part of multistation diplexers all the time. Long ago I had to keep a 780Khz station out of a transmitter we had on 820Khz...they were separated by a few miles. Installing this type of trap did the trick quite nicely.

User avatar
Deep Thought
Posts: 3364
Joined: Thu Mar 20, 2008 9:23 am
Location: La Grange, IL
Contact:

Re: Dealing with mixing products

Post by Deep Thought » Sat Mar 24, 2018 4:51 pm

Those filters consist of about $3,000 in parts if you pay full marked-up retail, $1,000 if you shop around. The cabinets tend to be one-size-fits-all affairs which are cheaper for KTL (and others) to make than customs for each job.

If you can put the parts into existing ATU or phasor cabinets it is cheaper and more effective, but that requires work and some expertise tuning the things.

Nothing wrong with buying the filter if you have more money than time.
Mark Mueller • Mueller Broadcast Design • La Grange, IL • http://www.muellerbroadcastdesign.com

kcbooboo
Posts: 495
Joined: Sat Dec 10, 2011 4:53 pm
Location: East of the Mississippi

Re: Dealing with mixing products

Post by kcbooboo » Sun Mar 25, 2018 6:00 am

It just seems that two series LC circuits in parallel, with only a 1kw power capacity, shouldn't take up a full rack cabinet's worth of space. To be fair, the phasor at the 1kw station is quite small; it takes up one rack width and is maybe a few feet tall, about the same size as the AM1A tx.

I'm guessing that the filter is a stock unit, one size fits all, one price pays for everything deal. As it eliminated the product to a point that I couldn't detect it on the FIM-41 means they got their money's-worth. The engineer who was tasked with putting it in isn't an RF guy.

Bob M.

Post Reply