Low Pass Filter Placement

[Captain Kilowatt]

Why would it be providing a near 50 ohm impedance at the harmonic frequencies? RF is reflected when the load impedance is different than the driver output impedance, which is what I think you are looking at however the filter effectively bypasses the harmonics to ground much the same as bypass capacitors remove RF from audio lines. The filter simply "selects" which frequencies to bypass to ground and ignores everything else. I would think that the overall impedance to harmonics would be close to zero ohms.


OK I just paused to do something after typing that last line. I just connected my MFJ-269 to my Barker&Williamson FL-10\1500 low pass filter and the output of the filter into my 50 ohm dummy load. The filter has a steep roll-off around 35-36 MHz. Below that cut-off freq. I saw 50 ohms on the MFJ-269. As I approached trhe roll-off freq the impedance dropped sharply to zero and stayed there throughout the range of the MFJ-269 above the cut-off frequency. Since the impedance was zero ohms the high frequencies above the cut-off freq are basically shunted to ground.

 

[jazzsinger]

If a low-pass filter is absorbing harmonic power instead of reflect it, then it would provide a near 50 ohm impedance at the harmonic frequencies.

i think thats where your getting mixed up, the low pass filter doesn't absorb harmonic power, it shunts/directs it too ground whilst allowing frequencies below the cut off to pass through relatively unhindered, its the ground that absorbs harmonics not the filter, the filter only routes them.

 

[Kamikaze]

Since the impedance was zero ohms the high frequencies above the cut-off freq are basically shunted to ground.

If that were true what happens when the coax to the filter is an electrical odd multiple of a quarter wave at the harmonic's frequency?

 

[dudmuck]

i think thats where your getting mixed up, the low pass filter doesn't absorb harmonic power, it shunts/directs it too ground whilst allowing frequencies below the cut off to pass through relatively unhindered, its the ground that absorbs harmonics not the filter, the filter only routes them.

Ground absorbing RF power is a concept worthy of thought.
Probably the best example of RF power on ground is with an antenna that doesnt have a counterpoise, or is inadequate. In that case there would be RF power on the coax shield, and the shield is radiating.
So if there were harmonic power on the chassis of a low pass filter, i would think it would be radiating from the chassis and coax shields connected to it.

 

[Captain Kilowatt]

Ground absorbing RF power is a concept worthy of thought.
Probably the best example of RF power on ground is with an antenna that doesnt have a counterpoise, or is inadequate. In that case there would be RF power on the coax shield, and the shield is radiating.
So if there were harmonic power on the chassis of a low pass filter, i would think it would be radiating from the chassis and coax shields connected to it.

There is always RF current on the coax shield but it is normally contained on the INSIDE of the shield. Common mode currents flow on the OUTSIDE of the shield. Since the RF is flowing on the centre conductor AND THE INSIDE the harmonics are still passed to the filter regardless of any common mode currents on the outside shield however harmonics carried on the outside of the shield will be radiated thus diminishing the effectiveness of the filter. The problem in this case is not the filter but rather the common mode currents present.

 

[C2]

I was considering a clean CW signal.

Take your 30 MHz LPF with a pure resistive 50 ohm load on the output.

Feed 60 MHz CW. There will be no reflection?

What if you open or short the output, still no reflection?

 

[C2]

i think thats where your getting mixed up, the low pass filter doesn't absorb harmonic power, it shunts/directs it too ground whilst allowing frequencies below the cut off to pass through relatively unhindered, its the ground that absorbs harmonics not the filter, the filter only routes them.

Well, what do the experts think?

 

[W6HSA]

The grounding or rerouting of the filtered signals is interesting. I have a Bencher YA-1 that does not have a means of grounding as it is pop riveted together. The ICE filter, on the other hand, has two grounding posts. They advise grounding the filter close to the point of ground; whereas, others recommend connecting their filters directly to the transmitter or amp.

The Bencher should be grounded by virtue of the shield of the coax which is grounded via the radio and antenna ground. Filters directly behind a transmitter/amp are grounded through the radio/amp/antenna. It appears to me that in all of these options for placement ground is achieved. How it is achieved is a matter of installation preference or requirement. I guess the higher the power, the more stringent the grounding requirement, and since the ICE is a high-power handling filter, they recommend placing it as close to the grounding source as possible.

On a separate note, the Bencher was given to me because it was 'soaking up output'. I adjusted the screws on the bottom to match the transmitter's output without the filter and it appears to be working fine. My colleague has a spectrum analyzer and I can't wait to see how this 'tune up' actually works. I also noted that I can get more watts out of it in turning the adjusting screws a bit passed the equalization point, but I assume it is changing the impedance. I matched it as closely as I could using the LP-100a. The SA will tell the real story.

As an update, I had a few locals 'listen' to both filters and they unanimously selected the ICE. Appears my so-called tune up may be narrowing 11 meters. I haven't tried it on the lower bands yet but I found it interesting. It will go on the SA tomorrow. I'll post back up.

Interesting subject indeed!

 

[W6HSA]

I used a MFJ Antenna Analyzer and a dummy load to check the Bencher. I first measured the ICE filter to get a baseline. I then measured the Bencher and it was significantly off. The SWR was just under 1.9 and the impedance was substantially off as well. I then adjusted the filter to match the ICE. I was able to adjust it to 1.2 at 50 ohms. The ICE had a bit of capacitance and inductance just before the meter zeros out while the Bencher showed zero. The difference was minuscule, so the spectrum analyzer will be the tale of the tape. My colleague with the spectrum analyzer invited me up to his shop so we can sweep them and verify measurements against the performance of the filter's cutoff. I'm curious to know.

I then got a few fellows on the air and did a filter shootout! The Bencher and the ICE 'sounded' identical to them, whereas before, the Bencher sounded 'pinched in', cutting off the audio dynamics. I have two NOS Avanti filters constructed the same way which were off as well. I set them up and they couldn't tell the difference among any of the filters. I discovered the sequence to setting them up: the first one gets you in the ballpark, the second one is a finer adjustment and the third one is very granular and gets you spot-on in tiny increments. In fact, I could introduce a bit of capacitance to match the ICE but I somehow think that tiny difference is inconsequential. I'll post up after my visit with the spectrum analyzer.

I hope this helps others. I've found this exercise very interesting.

 

[Moleculo]

Very interesting stuff, indeed.

 

[Kamikaze]

A filter designed to cut off RF harmonics effected the on channel audio?

 

[W5LZ]

"A filter designed to cut off RF harmonics effected the on channel audio?"

Yes it could, but that would depend a lot on the frequency of use and just how far 'off' that filter's adjustment was, and signal bandwidth too. I'd have to think that if the frequency of use wasn't very far from the 'cut-off' frequency there would be a 'better' chance of that audio difference being heard. If that frequency of use was a ways down from that 'cut-off' frequency, then I seriously doubt if it'd be heard at all, at least not with a 'low pass' filter.
So could it happen? Yes. Did it? Beats me, I wasn't listening...
- 'Doc

 

[Captain Kilowatt]

A filter designed to cut off RF harmonics effected the on channel audio?

My guess is that the increased SWR is what was to blame. I missed what radio the OP was using while testing the filter but most newer HF radios will begin to have some audio issues when the SWR approaches 2.0:1 as a result of ALC action.The original run of FT-857's and 897's had this problem when the SWR was a mere 1.5:1. I had one of the originals. :headbang The newer "D" series has the ALC threshold set higher to around 2:1and the audio is improved under slightly higher SWR conditions. I also have a newer "D" version of the 857 and find it is indeed better.

 

[W6HSA]

When we swept the Bencher with the spectrum analyzer last year, it was limiting the signal above 25mhz and reducing the power output so I shelved the filter. After the adjustments, the roll off didn't start until after 30mhz as designed and the insertion loss was negligible.

With the SWR now down at 1.2 at 50 ohms, it allowed the radio, an ICOM IC-7000, to deliver the signal. We also tested it with a 10/11 meter radio where the degraded audio was most detectable in AM. We concluded it was a combination of reduced power output and the early rolloff that was causing the 'pinched-in' audio.

I have since found every one of the filters I have checked were been off. I have re-tuned several more filters bringing their SWR down to 1.1-1.2, ohms to 50-52 with very little-to-no capacitance/inductance.

It appears I have become the filter tuner.

 

[Radioman56]

The "bottom-line" of "passive" (not "active") low-pass-filters (and to back-up our "chastised" member "DudMuck" here ;-)

All LPF circuits / boxes WILL produce reflected power from any / all harmonics that are being suppressed by the LPF., AND will affect amplifier performance.!


The RF power amplifier (PA) in most all radios, employ a harmonic filter / termination circuit in order to boost its efficiency. It consists of a series of passive resonant circuits and tuned to the desired output frequency.

The harmonics generated by the active output device (the PA) are mostly reflected back to the output of the PA, and causes some amount of parasitic / anti-resonance problems between the harmonic-rejection filter (LPF) and the output termination circuit of the PA, which can degrade the efficiency of the PA itself.

This is why you will typically see, that only the first inductor coil (just after the PA final device(s) output transformer) will have had one of its windings (usually only in its center) just slightly "spread" by the factory.

*Just "google" some internal images of new, NON "cranked & spanked", 10 & 11 meter radios.!!

It's VERY similar to an antenna, in that any frequencies outside (above & below) of its resonance WILL be reflected back to the PA output (though, in an LPF circuit, this only effects the frequencies above its design limit).

For an LPF to "ground-out" any of the harmonics, one would have to design the LPF to not only pass-through all the frequencies below its designed upper cut-off limit, but more importantly., it would also have to be designed to be "series LC resonant" to ground at each of the un-wanted harmonics.!

It's NOT similar to an audio low-pass filter (as in electric guitar tone-controls or amplifier tone-controls) in which is simply a "hi-cut" filter, using a capacitor and just two connections of a potentiometer (making it technically a "rheostat"), which does in fact, just grounds-out, frequencies above a certain design-point.

Such a thing of the grounding of un-wanted RF frequencies (regardless of frequency) is usually done with an RF "circulator", wherein any / all reflected power is diverted / circulated into a "dummy-load" of appropriate wattage size. With some circulators being dual or even triple stage, for even further "dumping of watts to ground" before getting back to the PA.

Oh., and one last thing to "drive it home". Here's a "real-world" experiment., for those who are humble enough., to actually prove-out the Truth to themselves...

Take a standard / quality 30 Mhz LPF (ie: a Drake brand, TV-1000, 3000, 3300) and cable it up as follows:

Using an "HF-rig" (that will operate on 54 Mhz, in the 6M band) or better yet, for even greater "stark" results., use a simple 2M, FM rig (at 145 Mhz) and run its output, FIRST into a standard SWR meter, then into the 30 Mhz LPF, then lastly, into a standard 50 ohm, non-inductive dummy-load, and then attempt to take a standard SWR reading (between your transmitter's output, and the input of the 30 Mhz LPF).

Then be honest with your test results...

Regards,
Eric Klein
Klein Communications