DPBOX HTTP Interface: R%20GERAETE%201DPBOX HTTP Interface GERAET @EU de:DA2JP 29.10.96 02:26 0 13409 Bytes Heath SB-104A mods *** Bulletin-ID: 02A423DB0MWS *** Read: 100 Path: DB0SIF!DB0SIF!DK0MTV!DB0HOM!DB0FRB!DB0CZ!DB0KCP!DB0MWS Sent: 941002/1005z @DB0MWS.#BAY.DEU.EU [BBS Muenchen, JN58TE, DL2RBI/DG6MAY] de DA2JP @ DB0MWS Improved Performance for the Heathkit SB-104A Note* the following is a reprint from a now out of print mœ饹4PСՅɕѥ乑ɽݹȥͭ444difications for better selectivity, sensitivity, and overload capability. The Heath SB-104A is a good transceiver. It can be made even better by incorporating the simple modifications described in this article. The modifications, if made according to the directions given, will provide significant improvements in: 1. Better sensitivity, especially in the 10 and 15 meter bands. 2. Receiver selectivity in the SSB mode. 3. Receiver strong-signal-handling capability. These modifications, as well as a few others, have been developed over a two-year period with great care and attention to detail. Before snipping any wires, I strongly recommend that you fully understand what is being accomplished by each and every circuit change. In addition, the modified circuits should be studied and compared with the original Heath circuits. Receiver Sensitivity Improvements In my opinion, the SB104A suffers from inadequate sensitivity, especially on 10 and 15 meters. The six bandpass filters for the 80 through 10 meter Amateur bands, located on circuit board "G" are diode switched. That is, when the radio is on a particular band, diodes on circuit board "G" associated with bandpass filter in use are forward-biased to provide a low loss RF path for that band. The diodes do have some loss, however. These losses can be reduced by replacing diodes D701 through D704 (Heathkit part designations); D707 through D710; D713,D714; and D717, D718 with Motorola MPN3401 PIN diodes, which are intended for RF-switching use. To make the mods, first remove the original diodes. Install the new PIN diodes in place of the original diodes. Pay attention to the polarity of the MPN3401s. These devices are in a square epoxy package; the end with the ridge, or high spot, is the cathode. The leads on the MPN3401 are very short, so they must be mounted on the foil side of the board. Mixer Improvements The next step is to replace the receive mixers. The original first and second mixers on board "G" can be improved by substituting Minicircuit Labs SBL-1 broadband mixers. These new mixers provide better isolation between ports and have less conversion loss than the original mixers. They also have good strong-signal handling capabilities. They are commonly used in high performance UHF receiving systems. To make the mixer modification, first remove the Heath first mixer, consisting of T701, T702, diodes D719, D721, D722, and D723. Also remove capacitors C741 and C742. Apply some epoxy to the top side of one of the SBL-1 mixers and cement it to the component side of the board. The pins should now be facing upward. Wire the mixer according to the pinouts given in the manufacturer's literature. Also couple the input and output with 0.01 uf capaciotrs. Use T702 to Z-match the ouput. Next, remove the Heath second mixer by removing T703, T704, D724, D725, D726, D727, and epoxy the new SBL-1 mixer to the board. Wire accordingly, using T703 and T704 to Z-match the input and ouput. No additional caps are needed, as they already exist on the board. Finally, locate transistor Q702. This transistor is a 2N5109, which is an epitaxial planar low-noise device. It is used as a post amplifier between the firts and second mixers. Remove R721, the 1k-ohm collector resistor, and replace it with a 1mH choke. Next, remove R722, the 560-ohm emitter resistor, and replace it with a 100-ohm resistor. Finally, replace C745 emitter bypass capacitor with a 0.01uF disc capacitor. The above modification serves two purposes. First, it increases collector current to about 100mA, which greatly reduces the chances for the stage to clip on strong signals. (the 2N5109 is rated for an Ic of 400mA). It also increases the stage gain, which is needed to overcome the losses of the second crystal filter. At this point, reinstall board "G". Turn on the SB-104A, and check out the receiver to make sure it receives on all bands. Next, pull out the board, install the extender board along with board "G" and retune the first and second mixer trimmers according to the Heath operation manual. If a scope and sweep generator are available, the board may be sweep aligned. Taming The Noise Blanker The next step is to rewire the noise blanker. Using a piece of RG-174 or other shielded cable, Add a DPDT switch to allow the noise blanker to be bypassed when not desired. I found that the noise blanker caused cross modulation, even when turned off, by virtue of its being in the signal path at all times as originally wired. Improving SSB Selectivity To improve skirt selectivity on SSB, remove the original crystal filter from circuit board "E" and install a Fox Tango Corportation 33h2.1 filter in its place. The input and output of the new filter are coupled with 15.5uH coils in series and by two 150pF capacitors parallel to ground on the input and output. Further Improvements More modifications were made to the SB-104A to acheive the following goals: 1. To further improve the strong-signal-handling capability. 2. To improve the active audio filter. 3. To reduce receiver hiss. 4. To provide a slower AGC release time. Make the following changes to board "F". Proceed as follows: 1. R513 Remove the 2400-ohm resistor and replace it with a 1 meg resistor. 2. R502 Remove the 820-ohm resistor and replace it with a 1500-ohm resistor. 3. R514 Remove the 620-ohm resistor and replace it with a 100-ohm resistor shunted by a 0.01uF capacitor. 4. R517 Remove the 2400-ohm resistor and replace it with a jumper. 5. R511 Remove and discard (or save it for your junkbox). 6. R512 Remove the 4700-ohm resistor and replace it with a 1mH choke. 7. R516 Remove and discard. 8. R518 Remove the 10K resistor and replace it with a 1 meg resistor. 9. R572 Remove the 4700-ohm resistor and replace it with a 2200-ohm resistor. 10. R541 Remove the 1500-ohm resistor and replace it with a 1K resistor. 11. R545 Remove the 820K resistor and replace it with a 2.2 meg resistor. 12. R546 Remove the 5.6 meg resistor and replace it with a 33 meg resistor. 13. C535 Remove the 2.2uF tantalum capcitor and replace it with a 5uF 15 VDC electrolytic capacitor. 14. Q502, Q503 Remopve and replace with 2N3819 JFET transistors. On the right hand upper corner of board "F", from the component side of the board, locate the foil going to Q517 base and carefully drill a 1/16 inch (1.6mm) or smaller hole through the base foil and the ground foil. Scrape off the green or blue coating around the holes and install a 0.1uF Mylar capacitor in the two holes. Locate coil L501 and remove the associated 100-pF mica capacitor. Replace it with a 130pF mica capacitor. Solder a 10-75 pF trimmer across the pins of L501. Piggyback this trimmer on top of L501 by soldering the trimmer directly to the top of the pins on L501. Install board "F" in the extender board in the SB-104A. Either peak the 10- 75pF trimmer for maximum noise or, if a signal generator is available, put the rig on 80 meters and inject a signal into the antenna jack. Only use enough signal to get an S-5 or so meter reading. Peak the 10-75pF trimmer for maximum S-5 meter reading. Use care not to saturate the IF. Use only as high a signal as necessary. Next, remove board "D" and change capacitor C441 (33pF) to 100 pF. This change increases the HFO injection and reduces receiver overload. Remove transmit audio regulator board "B" and make the following changes: 1. Change R217 from 4700 to 2200 ohms. 2. Remove Q207 and replace it with a Radio Shack 276-2026 transistor. The reason for these changes on board "B" is as follows. Q207 is the PTT switching transistor. When Q207 conducts, the relay in the SB-104A closes, and the unit is in the TX mode. Before I changed the transistor, I'd had 2 failures of the original Q207. For that reason, the Radio Shack device was installed; it's a tab-type transistor and is more capable of supplying the necessary collector current without premature failure. If this change is made, you must reduce the value of R217 from 4700 ohms to 2200 ohms. If you don't plan to change Q207, then leave R217 alone. When installing the new Q207, bend the leads of the transistor at a right angle and allow the transistor to lie over the top of IC202. This will allow the PC board to slip into its compartment in the chassis. Additonal changes to board "B", which are optional, are as follows. R214, the collector resistor of audio transistor Q201, may be reduced from 33k to 15K. This change will eliminate asymmetrical clipping - which may cause slight audio distortion during transmit in some units - in Q201. Capacitor C204, the 0.01uF coupling capacitor on Q201's base, may be increased to a 0.1uF Mylar. This will change increase the low frequency response of the transmit audio. This is a psersonal preference. You may like the transmitter audio better one way than the other, so get some on-the air checks from a few local stations and try the two different capacitor values. Finally, one change suggested by Heathkit is as follows. Remove the ALC/ Filter board and change capacitor C887 on Q802's emitter to a 0.68uF Tantalum. If your rig is of late vintage, the 0.68uf cap may already be installed. Test Results After Modification Three other active Amateurs are located within a half mile of me. After modifications were made, I made on the air checks with two of these stations. I was able to tune my SB-104A 18-20 KHz away from the other station's 60dB over S-9 signals and only slight desensitization was noted. Stations as weak as S-3 30 KHz away from the local 60dB over S-9 local SSB signals were solid copy, and only a slight hiss was noted while the local station was transmitting. These tests were made with the noise blanker off. The Heathkit SB-104A was also tested side-by-side with a top of the line (1981) Japanese transceiver. Both rigs were connected to a common antenna. The two unit ran neck and neck as far as sensitivity was concerned. All bands, 80-10 meters, were tested. When the two rigs were tuned to the same station, the SB-104A had much less receiver hiss than the Japanese rig, which made the SB-104A much more pleasant to listen to. The modified AGC action was very pleasant to listen to. No pumping was present after the modification. A comparative check of selectivity was also made on both units. Tuning the same station on upper sideband on the Japanese rig, and moving off frequency produced a high-pitched "donald duck" response that could be heard up to 3.5 KHz away from the center frequency. However, on the SB-104A, tuning more than 2.8 KHz away from the center frequency prooduced a sharp cutoff of the signal and the same signal that was heard 3.5 KHz away on the Japanese rig was undetecable on the SB-104A. The IF control on the Japanese rig was purposley left in the center position and not used during the checks. Turning the IF shift knob did not help the Japanese rig; however, the cascaded filters in the SB-104A were definatley doing their job. Before modification, the SB-104A exhibited lockup of the AGC, which would manifest itself with the S meter hanging up at S-6 accross large portions of the band when a strong signal was on. After modification, this problem totally disappeared, and was just a barely perceptible increase in hiss was noted. Some Afterthought Concerning the SB-104A As mentioned, the tests were made with the SB-104A noise blanker off. Turning on the blanker still causes cross modulation. This is because the noise blanker keys on signal as well noise. Because of the broadband nature of the rig, the blanker is subjected to 500 KHz or more of crowded spectrum when turned on, and it just can't handle that much signal. One answer to the problem is to put a monolithic crystal filter about 6 or 8 KHz wide ahead of the noise blanker. During the modifications, I placed the 2.1 KHz filter ahead of the noise blanker, and the cross modulation totally disappeared. However, the blanker became totally ineffective on noise spikes. Propagation through the filter caused the pulses to be rounded off, rendering the blanker ineffective. For this reason, only a modest filter should be placed in front of the blanker so that the pulses will not be rounded off, and the spread range of the signals presented to the blanker at any one time will be reduced. Well, gang, that's it. I would be more than happy to mail a photocopy of this article (which includes many diagrams) to anyone who wants it. Just drop me a note at DA2JP @DB0MWS.DEU.EU. Vy 73 es Good DX to all.