I was extraordinarily lucky on my way into work last week. While passing through an older neighborhood close to the office, I happened to notice a mass of metal on the curb for garbage pickup. Lo and behold, it was the classic CM-4228 antenna!

Fortunately, the homeowner was outside tending to his garden so I pulled up for a chat. He explained that he has recently moved into the house and thought the antenna was an obsolete satellite receiver of some type and that his wife thought it was an eyesore. So he pulled it down and placed it curbside. I downplayed the value of the antenna and casually mentioned that I'd be glad to take it off his hands. He said I could take as much trash as I wanted. Nonchalantly, I put it into the car and happily drove away.

The antenna was in very good condition and the original ChannelMaster balun was still attached. This weekend was my chance to give it a try. With the trusty Sencore SLM-1456, I did an A/B comparison against my AD 91XG. Both of these were performed in the attic and within 15 minutes of each other. Much like Ken Nist's shootouts, my prime interest is in raw gain measurements since they do show the relative reception traits in that regard.

Neither antenna was altered or hacked in any way. I was curious to see how each performed in "as shipped" form.

We have a nearly full allotment of facilities in the UHF band from 14 up to 51, with a good dispersal in between. Channel 38 has been off the air for a month, so it's omitted in spite of it being among those listed in the TV Fool for my home:

http://www.tvfool.com/?option=com_wrapper&Itemid=29&q=id%3d8d17206520ebf4

CM-4228

+----+-----+-----+------+----+
! CH.!LEVEL! MER ! BER !DIG.!
! !POWER! CCN ! !QLTY!
! !dBmV ! dB ! ! !
+----+-----+-----+------+----+
! 14 !- .7! 32.6!2x10-8!PASS!
! 18 !-21.2! 26.4!2x10-9!PASS!
! 19 ! 1.4!>36.0! <10-9!PASS!
! 20 !-18.0! 35.8! <10-9!PASS!
! 23 ! 3.0!>36.0! <10-9!PASS!
! 25 !-19.5!>36.0! <10-9!PASS!
! 27 !-20.6! 35.9! <10-9!PASS!
! 29 !- .1! 35.8! <10-9!PASS!
! 30 ! 2.6!>36.0! <10-9!PASS!
! 31 !-21.4!>36.0!1x10-8!PASS!
! 32 !- 1.2!>36.0! <10-9!PASS!
! 34 !-19.3!>36.0! <10-9!PASS!
! 35 !- 2.4!>36.0! <10-9!PASS!
! 36 !- 1.2!>36.0! <10-9!PASS!
! 39 ! 3.3! 33.0! <10-9!PASS!
! 40 ! 1.1! 32.7! <10-9!PASS!
! 41 ! 1.1!>36.0! <10-9!PASS!
! 42 !- 1.3!>36.0! <10-9!PASS!
! 43 !- 5.6!>36.0! <10-9!PASS!
! 44 !-14.7! 35.5! <10-9!PASS!
! 45 !- 5.6!>36.0! <10-9!PASS!
! 46 !- 8.5!>36.0! <10-9!PASS!
! 48 !- 2.1!>36.0! <10-9!PASS!
! 50 !-23.8!>36.0! <10-9!PASS!
! 51 !-26.9! 34.3! <10-9!PASS!
+----+-----+-----+------+----+


91XG

+----+-----+-----+------+----+
! CH.!LEVEL! MER ! BER !DIG.!
! !POWER! CCN ! !QLTY!
! !dBmV ! dB ! ! !
+----+-----+-----+------+----+
! 14 ! .3! 28.4!1x10-6!PASS!
! 18 !-20.1! 27.7!3x10-9!PASS!
! 19 ! 2.6!>36.0! <10-9!PASS!
! 20 !-17.8! 35.6! <10-9!PASS!
! 23 ! .6!>36.0! <10-9!PASS!
! 25 !-20.1!>36.0! <10-9!PASS!
! 27 !-21.5! 35.9! <10-9!PASS!
! 29 ! 2.1! 35.9! <10-9!PASS!
! 30 ! 3.1!>36.0! <10-9!PASS!
! 31 !-21.7!>36.0!3x10-8!PASS!
! 32 ! .7!>36.0! <10-9!PASS!
! 34 !-18.7!>36.0! <10-9!PASS!
! 35 !- 3.6!>36.0! <10-9!PASS!
! 36 !- .2!>36.0! <10-9!PASS!
! 39 ! 4.5! 35.1! <10-9!PASS!
! 40 ! 3.8! 35.0! <10-9!PASS!
! 41 ! 3.3!>36.0! <10-9!PASS!
! 42 ! .9!>36.0! <10-9!PASS!
! 43 !- 4.0!>36.0! <10-9!PASS!
! 44 !-12.5!>36.0! <10-9!PASS!
! 45 !- 2.7!>36.0! <10-9!PASS!
! 46 !- 6.7!>36.0! <10-9!PASS!
! 48 ! .2!>36.0! <10-9!PASS!
! 50 !-22.9!>36.0! <10-9!PASS!
! 51 !-23.6! 34.9! <10-9!PASS!
+----+-----+-----+------+----+

While passing through an older neighborhood close to the office, I happened to notice a mass of metal on the curb for garbage pickup. Lo and behold, it was the classic CM-4228 antenna!

Concerning the "rescued" CM-4228 that I luckily obtained a few weeks ago, I had a chance to look it over more closely tonight. It appears that some of the bowtie elements may have become bent. Some of them don't appear to have the same width at the widest part of the 'V'.

Since they are so easily spread or squeezed back into place, how far apart should each bowtie 'V' be? I found Ken Nist's models but these appear to be with his modificiations. I'm just looking for a simple bowtie measurement for the "as is" classic CM-4228 to shape them back to factory specs.

how far apart should each bowtie 'V' be?
On the original CM4228a, they should be 5.5 inches apart at the ends of the V.

If the whiskers are +.5/-2 inches of that, its not going to make a measurable difference in gain. Consistency of the spread of all the whiskers helps I think, but I may have to try modeling something like having the inside whiskers spread narrower than the outside ones.

What I would look for is to make sure that the whiskers are all flat in a plane, and the phasing lines arent bent up. The phasing lines should be parallel except at the crossovers. The connecting phasing lines between the two bays should be symmetrical, including the bends. There is a taper in the connecting phasing lines.

Also one of the first things to check is whether your CM4228 has had the factory upshift trick applied - if you see slight scoring on the whiskers about an inch from the tips then you have a stock CM4228, but if there are none then check the length of the whiskers against Ken Nist's dimensions. If your antenna's whiskers have been cut at those factory scoring points it was to shift the antenna's peak reception upwards to better suit the old UHF channels from the high 50s to 83. The only way to fix that is to use new whiskers or solder/braze extensions to the tips.

Ah, good point. I remember some other uhf antenna from long ago that had the same upshift trick scoring built into its elements.

The length of the whiskers should be 7 7/8" measured from the rivet apex to the whisker tip.

Also one of the first things to check is whether your CM4228 has had the factory upshift trick applied - if you see slight scoring on the whiskers about an inch from the tips then you have a stock CM4228, but if there are none then check the length of the whiskers against Ken Nist's dimensions.

On the original CM4228a, they should be 5.5 inches apart at the ends of the V. The length of the whiskers should be 7 7/8" measured from the rivet apex to the whisker tip.

Thanks to both of you for the detailed and clear information. The CM-4228 that I rescued from curbside trash collection DOES have the scoring an inch from the end of each whisker and they measure at 7-7/8" from the rivet to the tip.

The two bottom bowties were indeed somewhat bent, probably from how it was placed near the curb for the sanitation truck. I got them straight and broadended them out from about 4-3/4" inches to the 5-1/2" gap at the outmost 'V'.

Just a couple more things (a question and an observation):

One the raw gain chart (http://www.hdtvprimer.com/ANTENNAS/comparing.html) are the AA numbers for the CM-4228 with the factory upshift or not? The collision of red colors at the high end of the UHF band makes this hard to discern.

Both the net and raw gains on those charts show a discernble notch right around channel 23 for the CM-4228. That is contrary to my measurements (above) and continues to be case after making the small whisker adjustment. If anything, it seems that the antenna has a discernible peak on 23.

One the raw gain chart (http://www.hdtvprimer.com/ANTENNAS/comparing.html) are the AA numbers for the CM-4228 with the factory upshift or not? The collision of red colors at the high end of the UHF band makes this hard to discern.

Both the net and raw gains on that chart show a discernble notch right around channel 23 for the CM-4228. That is contrary to my measurements (above) and continues to be case. If anything, it seems that the antenna has a peak on 23.
Those graphs posted there now look to be for the new CM4228HD, except he didnt rename the interim Raw Gain chart. IIRC, his old graphs for the CM4228 looked to be much different, smoother with higher gain. I think we probably have the old graphs on this forum somewhere, at least in comparison to the DBGH.


I have his nec file for the old CM4228, so I guess I could rerun them. That will take some time though.

I've never seen a graph of the CM4228 with the upshift trick applied but from anecdotal evidence over the years it was apparently a very simple and clear shift up the band and left the channels below 30 somewhat weaker and below 20 noticeably weaker.

That will take some time though.
Yeah, about 10 hours of computing time, heh. Heres what the old CM4228a graphs showed, no upshift trick :
(frequency between each point is 6 mhz, ie one channel)

Raw Gain :
http://img545.imageshack.us/img545/7720/cm4228arawgain.jpg

SWR:
http://img689.imageshack.us/img689/4508/cm4228aswr.jpg

Channel 23 Net Gain is about 13.75 dBi. Peak gain occurs at around channel 47.

A quick way to look at the SWR figures to subtract from Raw Gain to get Net Gain :

SWR Mismatch Loss dBi
----- -----------------
1.0 0
1.5 .18
2.0 .51
2.5 .88
3.0 1.25
3.5 1.60
4.0 1.94
4.5 2.25
5.0 2.55
6.0 3.08





You can see how crappy the new and improved CM4228HD is, heh.

I've never seen a graph of the CM4228 with the upshift trick applied
I could run that. Exactly how much is the upshift trick ? 1" off the whiskers exactly ?

300Ohm, FYI here's a link that gives a nice overlay of the new and old 4228s (along with some theoretical gain with modifications to the HD).

http://www.hdtvprimer.com/ANTENNAS/TemporaryPage.html

It would be interesting to see how the upshift model compares to these.

Yeah, theres the old CM4228a Net Gain graph, scaled out further than mine above. I knew it was still around somewhere, heh.

If re_nelson can confirm the upshift trick distance, Ill get the computation on that started.

All this talk about the up-shift and looking at the plots above has now got me thinking about a "downshift" now. Specifically, a way to move that 30 channel higher gain area (between ~ ch 28-58) back 10 channels to cover the ch 18-48 range. I'm thinking of clipping on an extra inch or so to the whiskers of my 4228a... if only it were so simple... :)

Specifically, a way to move that 30 channel higher gain area (between ~ ch 28-58) back 10 channels to cover the ch 18-48 range. I'm thinking of clipping on an extra inch or so to the whiskers of my 4228a...
That would mean increasing the size of the antenna, not clipping it down.

Bad choice of words on my part - I meant extending the whiskers by "clipping" on an additional length of wire to the end :).

if only it were so simple...
It might be. Adding 1 1/8" to the whisker length to make it an even 9", looks physically good. I wouldnt go much further than that though.

Im running the numbers on that while I wait for a response from re_nelson.

You're pretty much at an M4 then, heh. We haven't talked about that design lately.

If re_nelson can confirm the upshift trick distance, Ill get the computation on that started.

I can confirm that the distance from the end of each bowtie whisker's to the uptick groove is precisely one inch and not a micron more or less. :-)

We all know that the CM-4228HD is the newer version made in China. So what about the 'a' suffix applied to the classic CM-4228. Is that indicative of the factory altered version to favor the lower UHF spectrum?

So what about the 'a' suffix applied to the classic CM-4228. Is that indicative of the factory altered version to favor the lower UHF spectrum?
I dont think so, the 'a' version has been around for a long long time. If it is, its because of the loss of channels 70 - 83. The CM4228 has been around since those times. Personally, I think it was just a small manufacturing change.

You're pretty much at an M4 then, heh.
Well, a horizontally stacked M8 with pretty good phase lines. :p

Here are the CM4228a graphs with 1 1/8 inches added to the whiskers, leaving everything else the same:
(frequency between each point is 6 mhz, ie one channel)


Raw Gain:
http://img4.imageshack.us/img4/6843/cm4228plus1125inchesraw.jpg

SWR:
http://img818.imageshack.us/img818/6789/cm4228plus1125inchesswr.jpg