From cbfsb!att!pacbell.com!mips!mips!munnari.oz.au!uunet!bonnie.concordia.ca!ccu.umanitoba.ca!bison!draco!sys6626!inqmind!bills Mon Mar 16 17:23:09 EST 1992 Article: 14382 of rec.radio.shortwave Path: cbfsb!att!pacbell.com!mips!mips!munnari.oz.au!uunet!bonnie.concordia.ca!ccu.umanitoba.ca!bison!draco!sys6626!inqmind!bills From: bills@inqmind.bison.mb.ca (Bill Shymanski) Newsgroups: rec.radio.shortwave Subject: DX440 whine removed Message-ID: Date: 15 Mar 92 18:15:10 GMT Organization: The Inquiring Mind 1 204 488-1607 Lines: 104 It is possible to remove the whine that the DX440 emits by shielding internal wiring, however, having done it, I don't think everyone will find it worth the trouble. The problem has to do mainly with lead dress in the area of the volume control. The tuning knob apparently is scanned by the microprocessor just as the keyboard is scanned; this results in square-wave-like signals being fed to the knob. As the knob is rotated, one of 8 radial spokes inside the knob makes contact with three fingers in the knob assembly; I surmise the microprocessor watches these three, notices which one was contacted first, and then deduces which way the user is turning the knob. The whine is audible when one of the spokes is touching one of the fixed contacts. Immediately adjacent to the wire assembly that connects the knob to the CPU board is a gray shielded two conductor cable that connects the volume control on the left side of the DX440 with the audio power amp on the right side. This wire was dressed right against the CPU/display board on my radio. As well, there is a green wire (common ?) connecting the audio amp with the CPU board in the area of the volume control. Due to the close proximity of these wires and the mix of high- level digital and low-level analog signals, the whine is coupled into the audio signal at a point downstream of the volume control; thus, the whine is most noticable at low volume and masked when you turn up the level. I tried bypassing the contacts in the tuning knob with small capacitors; this was totally ineffective since any value large enough to reduce noise stopped the tuning knob from working. I came to the conclusion that the shielding of the audio signal in the area of the CPU/display board was inadequate, and decided to double-shield this wire. I used two pieces of braid stripped off co-axial cable. A piece from RG 8, about 200 mm long, was slipped over the tuning knob wire assembly ( slightly flatten the braid so that you can work the connector through the braid), then insulated with tape ( heat-shrink tubing would be much tidier - don't start this project at 9 PM on a Saturday night). I ran the gray shielded pair through a piece of braid stripped from RG58 - this requires un-soldering the gray pair from the audio amp. On each braid, leave enough length to make a pigtail you can solder to a convenient nearby ground. Tape or use heat-shrink to make sure the braid doesn't short against the underside of the tuner/IF board. After doing all the above, the noise was almost totally eliminated. Slight manipulation of the green wire from the volume control to the audio amp reduced and then eliminated the noise - I taped the green wire down over the little piece of copper shielding found in the area. After doing all this, the radio has absolutely no whine no matter what position the tuning knob is in, both on the internal speaker and on stereo headphones. It took a long time to do this; I advise anyone attempting to duplicate this result to make careful note of the screws used and where they came from. I usually suffer from the Brazil-nut syndrome when reassembling consumer electronic gear, however someone not all thumbs would have less trouble than I. Observe the usual precautions when attacking a prized possession with a soldering iron. And, of course, you blow the warranty away the moment you open the back of the case. Someone was kind enough to post the observation that installing an extra diode on the CPU/display board would extend the FM tuning range down to 76 MHZ. Well, it does...but FM sensitivity goes into the pit. Two local 360,000 watt EIRP stations gave me no LEDs at all, and lesser FM stations were inaudible. I had to dismantle the radio again and cut the diode I'd put in. I could verify the loss of sensitivity on the bench, powering up with the diode in and out and showing the loss. This seems very odd to me; I don't know why the CPU is telling the radio to go deaf, maybe there's so many strong FM stations where they use 76 MHZ that you must decrease sensitivity. I also verified that the first of these three ( D404) blocks tuning above 26.1 MHZ. Still don't know what the third diode was for; didn't give me aircraft band or lower LW or faster scanning, so I left it out. My DX 440 works fine on the CB and 10 metre ham band ( although I can't get 10 metre FM, of course). I tried the filter also posted earlier here; and it works wonderfully well. The lower SW bands are actually useable now with an external antenna when before they were wiped out by local top-40. Doesn't help MW or LW reception of course; but the ARRL handbook lists other filters that look useful for low-pass, too. As always, I take sole responsibility for my actions and for the opinions expressed above, and I expect users of this information to behave similarly. Bill VE4STW bills@inqmind.bison.mb.ca The Inquiring Mind BBS, Winnipeg, Manitoba 204 488-1607