From owner-qrp-l@Lehigh.EDU Wed Nov 26 14:06:20 1997 Received: from fidoii.CC.lehigh.EDU (fidoii.CC.lehigh.EDU [128.180.1.4]) by oucsace.cs.ohiou.edu (8.8.5/8.8.5) with ESMTP id OAA19320 for ; Wed, 26 Nov 1997 14:06:19 -0500 (EST) Received: from Lehigh.EDU ([127.0.0.1]) by fidoii.cc.Lehigh.EDU with SMTP id <35015-53616>; Wed, 26 Nov 1997 14:06:03 -0500 Received: from nss4.cc.Lehigh.EDU ([128.180.1.13]) by fidoii.cc.Lehigh.EDU with ESMTP id <34871-20076>; Wed, 26 Nov 1997 14:05:07 -0500 Received: from m7.boston.juno.com (m7.boston.juno.com [205.231.100.196]) by nss4.cc.Lehigh.EDU (8.8.8/8.8.5) with ESMTP id OAA15124 for ; Wed, 26 Nov 1997 14:04:59 -0500 Received: (from gsurrency@juno.com) by m7.boston.juno.com (queuemail) id MmM07822; Wed, 26 Nov 1997 12:38:04 EST Message-Id: <19971126.223719.3430.1.gsurrency@juno.com> Date: Wed, 26 Nov 1997 22:05:41 -0700 Reply-To: gsurrency@juno.com Sender: owner-qrp-l@Lehigh.EDU Precedence: bulk From: gsurrency@juno.com (Gary Surrency) To: "Low Power Amateur Radio Discussion" Subject: HW-9 - Some additional ideas (long) X-Mailer: Juno 1.38 X-Juno-Line-Breaks: 1-4,11-12,14-15,23-24,33-34,39-42,49-50,57-58,65-66, 79-82,90-91,100-101,112-113,124-125,132-135,146-149,151-162 X-Listprocessor-Version: 8.1 beta -- ListProcessor(tm) by CREN Status: RO I'm still tinkering with my HW-9 and have a few more things to share. Hit delete if not interested. Thanks. RELATIVE POWER METER MODS With the earlier mods I made to the transmitter stages, increased BFO injection, and the increased VFO from the HW-9 mod article in QEX / QRP Power, I have more than enough transmitter output on all bands. In fact, on 80m - 12m the output is enough to pin the relative power meter easily and that makes it hard to reduce the CW level to the 5 watt level based on the relative power reading without using some care. So I made a few changes in the relative power sensor that feeds the meter. Since Heath only spec'd the rig for 3 - 4 watts output, a few minor changes are necessary after increasing the the output to ~ 5 watts. I replaced 10k resistor , R431 with a 12k resistor. This reduces the voltage to the meter circuit and makes the meter sensitivity more in line with the true maximum output of the rig, which on mine is as great as 9 watts on the lower bands. To keep the meter from pinning and slamming against the full scale limits, I placed a IN4242 12 volt zener diode (R/S pn. 276-563, 2 for $0.99) from the cathode of D408 to ground. A smaller wattage zener diode would be fine, I just had the 1 watt unit in the junk box. There wasn't a good point on the top of the PCB to place this diode, so I put it on the underside of the PCB. This zener limits the relative power sensor to an output of 12 volts and sets the meter's maximum range to the 60db over S9 limit. 5 watts shows up at about 40db over S9 on all bands now, and I find it much easier to judge the power output from the meter reading even if I don't switch the WM-1 on. The scale seems to be more meaningful with the maximum reading limited by the zener, and the needle is no longer slamming against the pin as it did before. Try it - you'll like it. :-) Note that you should do the mod mentioned in the QEX article to the Mute line of U301 to prevent the transmitted signal from affecting the AGC / S-meter readings caused by the IF amplifier "hearing" the output signal. This can cause misleading relative power measurements. It's very easy to do this mod. ACTIVE AUDIO FILTER MODS The active audio filter was peaking at about 950 hz, while the IF f_C (center frequency) is about 550 hz. So I took a look at the filter's components. Heath used .001 uf ceramic discs (C339, C341, C344, C345) for the filter, which have a tolerance of 20%. All of the caps in mine were low in value, in the 850 - 980 pf range. I found some nice polyester caps (the slim green kind) and matched four with my AADE L/C meter at almost exactly 1000 pf (.001uf) to replace the ceramic discs. The f_C is now at 710 hz, at least more in line with what was intended. A possibly better value might be .0012 uf if you can find some, or get them from Digikey as I plan to do. The polyester caps are much more stable and can be found with as good as 2% tolerances. The higher capacitance should bring the f_C down to around 600 - 650 hz to more closely match the f_C of the IF filter. I'm going to try and adjust the BFO frequency slightly to see if I can get the IF and AF f_C to be more closely matched. All of my other rigs use green polyester caps in the AF filters, and they perform as designed. While you're at it, you might want to replace C338 on U304B, the low-pass fitler. I have found that many of the glass axial monolythics Heath used in the HW-9 are waaaaay off value, and a good .01 uf cap for C338 should bring the filter's low-pass characteristics back in line. Some of the glass caps were as much as 0.194 uf in value! Use a .01 uf polyester cap instead of the glass axial unit. While you're at it, check the associated resistors in the low-pass and band-pass filters to be sure they are all within a few percent of the schematic value. For some reason, Heath used very high value resistors in these filter networks, and that just contributes to more noise. My other rigs' audio filters use caps that are a decade larger, i.e. .012 uf, so the resistors are much smaller and the developed noise is less. The QEX article reprinted in QRP Power used a completely different set of RC values for the active filter, but some of the values may be hard to find. If you can find the parts in that article, it may be a better design. I just used what I had, and will probably try to collect the parts in the article to try at a later date. The improvement I found in the filter's behavior was good, but it's still not the equal of the Tac-1 and ARK30 filters. SOLUTIONS TO LOW POWER OUTPUT While corresponding with another list member, I have found some areas to check if your transmitted output is low. First, be sure one of the final PA's is not blown. Get some larger heat sinks for Q405 & Q406. I used NTE401's. They just fit under the plastic band switch, and you can move one of the old heat sinks to Q404. This is necessary if you want to preserve the MRF237's at power levels near 5 watts. They are pretty tough, but the larger heat sinks are welcome insurance to your peace of mind. :-) Be sure diode D407 isn't shunting RF output from the PA's output. While transmitting on 80m into a dummy load at full output, it should *not* get warm. If it does, change it out with another 1N4149 and check C443. If C443 is much over .01 uf, it should be replaced with a smaller value. It is probably OK to use .005 uf or even .001 uf, but the sensitivity might suffer a little on the lower bands. I found I didn't have to use less than a .01 uf if I used a good 1N4149 diode at D407. Insure the bias network for the diode switch is working OK, and check the components for failure or out-of-spec values. Forget trying hot-carrier diodes or small PIN diodes for D407 without completely redesigning the bias network Heath uses, as they will quickly fail without the proper bias currents. The IN4149's have greater breakdown voltage ratings, and are more rugged than the PIN or hot-carrier types without the required bias changes. The QEX has complete details on using PIN diodes, but it's a difficult mod to do and the space is very limited around the T/R switch area. The Heath design will work OK, but some work is necessary to find and use decent components. Increasing the RF output to 5 watts or more exposes the shortcomings of Heath's design. A pair of back to back diodes would probably be a *much* better idea. Low output can be a result of poor BFO drive, and / or VFO drive. I covered this in an earlier posting, but it should be mentioned again. Use the RF detector Heath provided on the Oscillator board, and verify the BFO, VFO, HFO, and Premix outputs are adequate. Good BFO injection is essential to the SBL-1 dbm. If all of that checks out, then try changing Q401 to a MPS2222a, and Q402 to a MPS3904. My HW-9's output on 80m would fall off rapidly from 9 watts to as little as 3 watts in a few seconds until I replaced Q404. The output was fine on all other bands, so this was a little odd to observe it on only 80m! But a new Motorola 2N3866 did the trick, and increased the output on 10m too. Check my other HW-9 post for additional details on transmitter mods. If anyone has done the QEX article mod to the PA stage that "flattens out" the gain-vs-frequency characteristics. I would like to know how it worked, and what you thought. E-mail me direct. I'm still contemplating that mod, and may try it if the results were good. The rig needs some form of "ALC" to control the maximum power output on the lower bands. I'm looking at a way to use the relative power output to accomplish this, but the QEX mod might be the better way to go. SIDETONE THOUGHTS I am disappointed with the poor sidetone waveform, so I leave it turned off. If you feel the same way about the sidetone, but reducing the sidetone adjustment to minimum doesn't completely eliminate it, change R367 from 150k to 220k. This should do the trick. If anyone has ever seen a decent circuit to make a sidetone oscillator using op-amps, I would like to know about it. All of them I have seen produce lousy waveforms. Why everyone doesn't just use the "twin-T" oscillator is beyond me. I simply can't stand anything but a clean sine wave sidetone, so I usually use another receiver for monitoring my signal. Perhaps a RC phase shift network is the solution, but the "twin-T" is simpler and produces a sine wave with a single transistor! Oh well. ******************************************************************************************** Thanks for the bandwidth, and let me know what you think if these mods are useful to some of you who have the HW-9. references: QEX, October 1990, pgs. 3-9 Reprinted in: QRP Power, copyright 1996 by the ARRL, pages 3-26 to 3-32 72, AB7MY Gary Surrency Chandler, AZ (Near Phoenix), QRP-L #571, AZ ScQRPions, ARRL VE