[LONG] Re: [Kenwood] Replacing filter in TS 870

NR1DX [email protected]
Tue, 15 Oct 2002 23:45:12 +0000


Duncan

This is a problem I struggled with but never fully resolved, let me share 
with you my observations and experience to date.

So far I have found the following to be true on two TS-870 with serial 
numbers 705xxx and 706xxx respectively.

The problem:
In crowded high signal (40 to 60db over S9) conditions typical of a large 
metropolitan area such as Boston, New York, Washington during a major 
contest. Strange things begin to happen to the received audio which are for 
lack of a better description the inability of the DSP to respond to 
multiple strong signals which it thinks are really in the pass band. This 
appears to be a result of the fact that the AGC is apparently very much a 
part of the active feedback loop for the DSP algorithms which is why under 
"normal" conditions you can really get a lot out of the DSP through 
critical adjustment of the AGC depending on the type of noise or 
interfernce you are trying to filter out.

Kenwood's Solution:
The response by Kenwood on this problem that they at first tried to deny 
existed was to reduce the over all gain in the loop by redistributing the 
gain as outlined in one of their technical or service  bulletins ( I can 
find the number if you really need this). They also suggest that you crank 
in the attenuation , AIP and lower RF gain all of which basically serves to 
get the out of passband signals down to the level where the dynamic range 
of the DSP can handle it, but doesn't really get at (I believe) the 
fundamental problem, see the next paragraph.

My observations and proposed solution:
I experimented with trying different and narrower filters to no real avail 
which really confused me for awhile, until late one night after being at 
this too long I accidently left the filter completely out of the radio 
(essentially open circuit)and powered it back up with the signal generator 
still hooked up. The generator had been set to produce an "in passband" S 
meter reading of 60db over. Much to my surprise with the filter 
completely  out of the circuit (technically infinite pass band rejection) 
the S meter still registered S-9! A series of other experiments were under 
taken with the help of Motorola RF engineer K9MK. At first we thought that 
the there might be bleed through from the diodes used to switch the filter 
but that was eliminated by removing the diodes as well .... no change. The 
conclusion we came to that the basic board layout was allowing signals to 
couple "around" the filter! The thought at this point is to cascade a 2nd 
filter "off board". Technically this should be relatively easy to do by 
designing a proper matching network/transformer so that the cascaded 
filters don't mismatch each other. Physically this should also be easy to 
do by simply inserting the 2nd filter at the point where the coax leaves 
the the RF board at CN10 that goes to the 455KHZ mixer on the CONNECTION 
board at CN412. The lack of a proper equipment to measure the impedances at 
CN10 and CN412 plus the impedances of the filters and the time to do it is 
where the project has ground to a halt for the moment.

Suggestions:
Before you go further if your serious about this quest to make an otherwise 
good radio into a "super" radio, get a copy of the service manual, get 
access to some professional test equipment, study schematic and board 
layouts from the 3 KHZ filter to the 455 IF mixer. Do a full radio 
alignment once just to familiarize yourself with all the levels and 
adjustments, the service menus, especially the ones that center the 
passband. Make sure you can solder delicate SMT components with out 
butchering the circuit board, find a scrap computer card to practice on 
before you take an iron to your expensive radio.

Epilogue:
For those of you who live in the country don't worry about the problem. 
When I first bought my 870 I lived in "Silly Corn Valley" (rural 
Wisconsin)and I couldn't figure out what people were complaining about. 
Then I moved just outside Boston where the radio became literally unusable 
during CQWW and ARRL DX contests on CW..... When iI get the time the search 
for the grail will continue, perhaps this "maap" will help another knight 
blaze the trail before me

Good luck
Dave
NR1DX


At 05:23 PM 10/15/02 +0000, Duncan Cameron wrote:
>Does anyone out there have any experience with replacing the YK88S2 (8.83) 
>filter in the Kenwood TS-870 with another filter. In looking at the 
>service manual, I feel it would not too badly affect the DSP if the 
>replacement filter were only slightly narrower or slightly wider. Kenwood 
>parts lists the YK88S2 as a 3K bandwidth filter. I am thinking of 
>replacing it with a 4K and a 2.1K filter on a small board with relays and 
>a switch.  This would allow me to open up the 870 (more like the 850 which 
>the 870 takes much of its design from) and narrow it down to reduce the 
>adjacent channel interference that has been a bit of a problem in using 
>the 870 for contesting. My hope is that the DSP filter selection for xmit 
>and recv will operate as effectively with after a change in the filters.
>Any thoughts or tips among list subscribers?
>Thanks
>Scott C.
>
>
>
>_________________________________________________________________
>Send and receive Hotmail on your mobile device: http://mobile.msn.com
>
>- - -
>
>Your moderator for this list is:
>Larry Wilson KE1HZ [email protected]
>
>_______________________________________________
>Kenwood mailing list
>[email protected]
>http://mailman.qth.net/mailman/listinfo/kenwood

Dave
NR1DX
[email protected]

"A man who picks a cat up by the tail learns a lesson he can learn no other 
way" .... Mark Twain