IC-7600 3 & 6khz roofing filters - how effective are they really?

I am currently operating a ProIII and am very happy with its overall performance despite being somewhat dated. I'm considering a 7600 and have been researching its performance for some time.

I've read Sherwoods stuff on the narrow roofing filters and it makes sense. Stating the up conversion 64.455 mhz 1st IF is why the 3 and 6 khz filters are not what they claim to be.

Has anyone actually done any on the air testing with these two filters in a crowded contest? What was your impression?

 

See here my good fellow;
http://www.ab4oj.com/icom/ic7600/main.html

 

I just got a new 7600 so I'm still testing it. I have been nothing but impressed with the radio so far and I've made of point of trying out every feature.

I can tell immediately that the receiver has better selectivity than my IC-7200. I have been using the 7200 on an 80M net for a couple of years and I've always had a noise problem. The 7600 seems to be almost completely immune to it. I can't believe how much better it is. I tried turning the roofing filters on and off and selecting different ones, to see if that why the receiver sounded so much better, but there was very little difference on or off. I'm a roofing filter skeptic. I believe that a properly designed receiver will have more selectivity and better adjacent channel rejection, and therefore little need for roofing filters. I think that's what I'm seeing.

 

The first thing to learn about roofing filters is why they're needed. Here's one place to start: http://www.w8ji.com/receivers.htm . I used a Pro III for a few years and did a lot of CW contesting with it. Where it ran into problems was on 160 and 20 meters where the band was crowded with very strong signals, thus generating IMD that was evident from the increased background noise level (sounded a bit like pouring broken glass pieces into the trash). The simple way to reduce the problem was to back off on the RF gain (or switch in attenuation) to the point where the noise disappeared. Later, I installed the Inrad roofing filter in the Pro III and even with the 3 KHz or so bandwidth, that was enough improvement to eliminate the IMD problem in CW mode in my particular situation. It likely would not have had the same benefit if I lived in an area where many very strong stations were close-by.

The roofing filters are not there to improve the skirts of the filters in later IF stages. And, even a good receiver like that in the K3 will not do well in a CW contest using a 3 or 6 KHz filter in the first IF. It's the signals that fall outside of the range of the filter in the later IF stages that cause the IMD problems. A "properly designed receiver" will allow installation of a roofing filter with a bandwidth commensurate with the mode in use so as to minimize the IMD problem by preventing the signals that fall outside the bandwidth of the later filter from generating IMD.

 

N0SYA: I've read Adams stuff extensively and found it very informative but I was looking for actual on the air impressions.

AJ4WC: The 7600's NB is excellent I'm told. I agree, a poorly designed front end will make a roofing filter useless.

K8AC: I use the Sharp shape setting, which I believe adds to that "broken glass" sound. I can hear weak signals ever so slightly better on
Sharp. It also gives me ringing on CW with about 5% of the stations using a 250hz or narrower setting. I don't think the 7600's front
end is designed poorly. If Sherwood is correct, the roofing filters are less effective due to the high 1st IF frequency. Roofing filters
simply do not work properly above 20-30Mhz. And yes, I use attenuation quite often. I found my self in the 160M contest with AGC off and 12db ATT to hear the Czech Republic. I would have never heard him otherwise.

 

Can't comment on the 7600 settings. To say that roofing filters do not work properly above 20-30 MHz just isn't correct. Most of these Japanese up-converting designs use a first IF in the range of 60-70 MHz and indeed you aren't going to find crystal roofing filters at those frequencies in the narrow ranges you'd want for CW (500 Hz or less). But, that doesn't mean that they aren't effective. In the case of the Pro III, adding the 3-5 KHz bandwidth roofing filter made a very significant improvement.

 

Floyd

I don't think Rob Sherwood is claiming roofing filters don't work at all. If I'm understanding his comments correctly, the 7600's 3khz roofing filter tested at 5+ khz and the 6khz was higher than 6. The FT2000's test was even worse. I believe IMD with the 3khz was higher than the 6. That's not good. Rob explains this is due to the high frequency 1st IF. These filters do not work very well above HF. This is why roofing filters seem to be far more effective in down conversion receivers.

Dave

 

To KY3W - should have made it clear who my response was directed to. I was commenting on the AJ4WC statement: "I believe that a properly designed receiver will have more selectivity and better adjacent channel rejection, and therefore little need for roofing filters." One more comment before I disappear: The need for roofing filters has NOTHING to do with adjacent channel rejection or selectivity. The signals that cause the IMD problems are not within the passband of the filters that provide the selectivity you're speaking of, but are outside of the passband. The purpose of the roofing filter is NOT to improve the selectivity, shape factor, etc. of the later IF filters, but rather to reduce the number and amplitude of the signals that fall outside of the filters you depend on for selectivity. Those are the signals that mix in various combinations to cause IMD problems.

 

N0SYA: I've read Adams stuff extensively and found it very informative but I was looking for actual on the air impressions.

If I had a 76 I would do some testing using strong am and cw sigs, tuning close to the sig and then selecting the various roofs and seeing how they do in real world instances.

 

K8AC: Point well taken. The unwanted signals are outside of the passband.

N0SYA: The only on air demo I found was on YouTube where the op switched back and forth between the filters. It did make a difference in that demonstration.