Back to the DIY QRP corner
Back to the home page

Visited times.

SM6LKM's OPTIMIST 80 - Building Instructions (DOS text version / HTML) IF you are a QRP enthusiast, IF you have heard about a strange tool called "soldering iron", IF you are getting tired of CW-only operation, IF you think commercial QRP kits are too expensive, IF you think small is beautiful... This may be the rig you are looking for! Optimist 80 is a small single board DSB transceiver designed for the 80 meter band. Almost all components are available from the Swedish supplier ELFA AB. There are no esoteric components in this rig. The only exception I can think of is the small variable capacitor, CV1, but it can probably be found in numerous junk boxes. Short description: - Optimist 80 is a QRPp transceiver for 80 meter DSB. - Direct conversion receiver. - Variable input attenuator. - Narrow preselector. - Drives a loudspeaker directly. - Varicap tuned VFO, 3600-3800 kHz (3500-3800 kHz is possible, but not recommended). - Low LO leakage radiation means less hum problems when powered from a mains supply. - Linear class A power amplifier. - 1 watt P.E.P. output power, QRPp class. - Overload indicator with LED = poor mans ALC... - A joy to operate. Functional description, receive: Preselection is handled by the resonant circuit formed by CV1, C1, C2, C3 and L1. The signal from the antenna, attenuated by RV1 to a proper level, is capacitively coupled into the bottom of the resonant circuit. The J-FET Q1 exhibits a very high input impedance. Together with the relatively loose antenna coupling, this results in a high Q. The circuit is so narrow that even small excursions with the VFO will necessitate re-adjustment of the preselector. Besides beeing an impedance converter, Q1 also acts as a phase inverter providing a symmetrical input signal to the mixer, IC1. Q1 has no voltage gain at all in this circuit, rather a few dB loss, but, the high Q of the preselector circuit results in a considerable gain anyway. The mixer circuit, IC1, contains both a doubly balanced mixer and a local oscillator. The oscillator tank consist of C11-C16 and L2. Tuning is realized with the varicap diodes D2 and D3. The 10-turn potentiometer, RV2, is the tuning control. RT2 and RT3 are used to adjust the band edges. The symmetrical audio signal coming from IC1 pins 4 and 5, goes straight through T1, passes through the audio filter R21, R22 and C21-C26 and finally gets amplified to loudspeaker level by IC2. Functional description, transmit: During transmit, the relay K1 routes supply voltage to the +12TX line. A small current flows through the diode D1 which loads the preselector tank, thereby shifting its resonant frequency and lowering its Q. This is very important because even the slightest coupling between the LO and the preselector tank would otherwise excite the preselector tank to prohibitive levels. If too much LO finds its way into the mixer input, it will upset the mixer balance resulting in degraded carrier suppression. The +12TX line is also supplying power to IC3, the microphone amplifier. The amplified speech signal from IC3 pin 6, is routed via L1 to Q1. At audio frequencies, L1 can be considered a dead short. Q1 acts as a phase inverter during transmit also. The mixer balance is adjusted with RT1. It should be adjusted for minimum residual carrier. When the speech signal is mixed with the LO in IC1, a double sideband signal will result. This DSB signal, coming from IC1 pins 4 and 5, goes to the broadly tuned transformer T1. At radio frequencies, the center taps of T1 are practically shorted together by the capacitors C21 and C22 in the audio filter. T1 and C18-C20 forms a broad resonant cicuit which picks out the mixing products that belong to the 80 meter band. The secondary winding of T1 has an output impedance very close to 50 ohms. The power level can reach about -10 dBm (0.1 mW) P.E.P here without excessive distortion. The DSB signal is routed from T1 to the linear amplifier consisting of Q2-Q4 and their surrounding components. All three stages are operating in class A. The available gain in this amplifier chain is around 55 dB. Because of this, A -10 dB attenuator, R33-R35, has been insterted before the amplifier in order to minimize residual carrier and noise from the DSB generator. The amplifier chain is designed so that the final stage has the lowest power margin i.e. the final stage will be the first to suffer from flat-topping when the amplifier is overdriven. For obvious reasons... The amplifier output goes through the T/R relay, K1, it is cleaned in a 5-pole lowpass filter and finally it reaches the antenna terminal. The choke RFC4 shorts any 50/60 Hz hum present at the antenna terminal and, also, it provides a DC path to ground that prevents static build-up on the antenna. The 5-pole lowpass filter may seem pedantic. Well, it is. You can omit C58 and C59 and replace L3 with a piece of wire. I have not tested this transceiver with a spectrum analyzer, but, even if the harmonics are only 20 dB below the fundamental frequency, that's quite acceptable when related to the 1 Watt power level... As the final amplifier is working in class A, it maintains a nearly constant collector current under normal conditions. When the final amplifier is beeing overdriven, the required RF peak current becomes greater than the available standing DC bias. Consequently, flat topping will occur. At overdrive, the DC current through the transistor will be modulated by the speech. The overload indicator uses this fact to an advantage. The voltage drop across the emitter resistors R50 and R51 is fed to a lowpass filter formed by R32 and C54. This filter is necessary to get rid of the RF that is present on the emitter. The filtered signal is then amplified by the OP-amp, IC4, to a level suitable for driving the overload LED, D5. The modulation level is just about perfect when D5 barely starts to flicker on voice peaks. The forward voltage drop of D4 and D5 results in a desirable, and distinct, threshold. Construction: The PCB is made of double-sided PCB material. The copper foil on the component side is used as a ground plane and should not be etched at all. Print the PCB layout file and make a PCB using your favourite methods. Drill all holes with a 0.8 mm drill. Then, drill the holes for Q3 and Q4 with 1.0 mm. Drill the large square terminal pads, X1-X19, with 1.2 mm, and finally, the four mounting holes and the Q4 mounting hole with 3.2 mm. Look at the ground plane view. On the ground plane, on the component side, countersink all holes that are filled with black using a 2.5 mm drill. Be careful to avoid drilling through the board. The holes marked with a thin ring on the ground plane view are ground connections to be soldered on the component side and must NOT be countersunk. The grounded component pins have octagonal solder pads. Polish the board and give it a protective layer of solder-through lacquer. Start with all the components that have at least one pin grounded. Solder the grounded pins on both sides of the board. However, some components do not need to have their ground pins soldered on the component side. They get their ground from a nearby pin that is easy to solder on the component side. Plan the placement sequence carefully. If components needing ground plane connection is "built" in between other components, it could be very difficult to reach them with the soldering iron. Try to mount horizontal components, such as resistors, about a millimeter above the ground plane. Look out for shorts between ungrounded pins and the ground plane. C3, C4, C5, C21, C22, C25, C26, C28, C29, C32, C33, C34, C35, C36, C37, C38, C40, C45, C60, C61, C64, C65, RFC4, RT1 are examples of components that do not need direct connection to the ground plane, but it doesn't hurt to connect them. The electrolytic capacitors have square pads for the positive terminal. The electrolytic's don't have to be soldered on the component side. The toroid inductors should be carefully close-wound. Wind T1 with 23 turns to begin with, leave a few centimeters in a loop and continue winding 23 more turns in the same direction. Cut the loop at the middle and you have 46 turns with a center break. The number of turns is equal to the number of times the wire passes through the hole in the toroid. It is easy to mix up the wires of the small transformer T2. Look at the phasing dots in the schematic diagram. These dots are in the same end of the twisted pair. The T50-2 toroids should be mounted horizontally a few millimeters above the board. As spacers, use pieces of perfboard without copper, or any other plastic material of suitable thickness. You may need to drill a few holes for the wires, especially those wires coming down from inside the toroids. The small toroid T2 is mounted vertically, standing on its own legs. When the rig has been tested, put some lacquer on the coils to enhance frequency stability and to prevent microphonic effects. The final transistor, Q4, has to be insulated from the heat sink using a mica washer, or equivalent, and some silicon grease. When R24, R50, R51 and D6 is soldered in place, Q4 and its heat sink can be bolted to the board. After the VFO has been tested, and covers the expected frequency range, solder a screen around the VFO tank (C10-C17, L2, D2, D3). Thin PCB material or brass will do. Make it about 20 mm high. Do not do this before the VFO has been tested. See below. Don't forget the wire link that connects one end of C39 with IC4 pin 7. Solder terminal pins to the large square pads, X1-X19. Don't forget to solder the ground pins on the component side. TESTING THE RIG: Inspect the board CAREFULLY. Look out for reversed components and shorts. Don't be afraid to use an ohm-meter. Connect the tuning potentiometer RV2 to X1, X2 and X3 according do the schematic diagram. The clockwise end (CW) to X2, the counter-clockwise end (CCW) to X3. Connect the variable capacitor CV1 with short lengths of wire to X4 and X6. The body of CV1 should go to X4, which is ground. Connect the attenuator potentiometer RV1, according to the schematic diagram, with short lengths of wire to X4 and X5. Solder a LED temporarily to X11 and X12. X11 is the positve terminal. Connect a pushbutton switch from X13 to ground. This is the PTT. Connect a loudspeaker to X14 and X15. Connect a length of screened cable from RV1 to X16 and X17 according to the schematic diagram. X17 is ground/screen. Solder a pair of supply wires to X18 (PLUS) and X19 (MINUS). DO NOT connect the supply with reversed polarity. D8 and F1 are not yet in the circuit. Make sure the PTT is in the OFF state. Set RT1 near mid-position. Connect 12-14V supply, preferably current limited to 200 mA or so. A pleasant (?) hiss should be heard from the loudspeaker. About 9-12 mA should be drawn from the supply. Measure the following voltages with respect to the ground: The output voltages from the regulators IC5 and IC6 can easily be measuered on IC1 pin 8 and IC2 pin 6 respectively. Nominal voltage is 6 volts. 5.8V to 6.2V is satisfactory. The voltage at IC2 pin 5 should be about 3 volts. The voltage across R4 should be in the range 0.5 to 1 volt. Turn on your QRO rig and listen for the VFO. If your QRO rig has two VFOs, or memories, store the frequencies 3600.0 and 3800.0 kHz to simplify adjustments of the band edges. Turn RV2 max. clockwise. Adjust RT2 to zero beat on 3800 kHz. Then, turn RV2 max. counter-clockwise. Adjust RT3 for zero beat on 3600 kHz. RT2 and RT3 are interacting a little so you have to repeat this sequence a few times at both band edges to get it right. If the VFO refuses to cover the intended part of the band, you'll have to experiment with the values of C11-C16. In rare cases, you will need to adjust the number of turns on L2. You may wish to experiment with capacitors with different temperature coefficients to make the VFO extremely stable. When the VFO is running as it should, preferably with some adjustment margin on RT2 and RT3, it is time to put some lacquer on L2 and solder the screen walls in place. When this is done, fine-tweak the band edges again as described above. Check if the preselector is able to cover both band ends by turning CV1 with the tuning control, RV2, at its two extremes. A sharp noise peak should be heard when CV1 is tuned through resonance. Connect a real antenna to X9 and X10. X9 is ground/screen. Adjust RV1 for a pleasant listening volume. Listen around on the band. Don't forget to follow with CV1. There is no AGC, so be careful if you are using headphones. The receiver is finished. Disconnect the power supply. Replace the antenna with a 50 ohm dummy load on X9 and X10. Two 100 ohm 0.5W resistors in parallel (NOT wire wound!) makes an excellent dummy load for this rig. Short the microphone input by strapping X7 and X8. Connect an oscilloscope to the dummy load, the ground clip to X9. Increase the supply current limit to 1 amp. and connect the power supply again. The moment of truth is here. Close the PTT switch from X13 to ground. BANG!? If everything is as it should, the entire rig should consume about 400 mA from the supply. If it draws more than 500 mA, something is probably wrong. The final transistor dissipates about 3 watts, with or without modulation, so it gets quite hot. This is the drawback of class A, but it's completely normal. De-balance the mixer by turning RT1 until a clear sine wave is displayed on the oscilloscope. Turn the tuning control, RV2, between the band edges and check with the 'scope if there is a broad resonance with a peak at around 3700 kHz. If the resonance is not centered around 3700 kHz, C18-C20 may need changing to resonate with T1 at the band center. If the output is affected by the setting of the preselector, CV1, there is something wrong with R1, D1 or C5. Adjust RT1 for maximum carrier suppression, that is, as little output power as possible. Increase the sensitivity of the 'scope for the final tweaking. Remove the shorting strap on the microphone input, X7 and X8. Connect a tone generator set to about 1000 Hz. The input level should be adjustable in the range 0 to a couple of millivolts. Connect the trigger input of the oscilloscope directly to the tone generator, if possible, in order to get a stable display. Set the scope's sensitivity to 5V/division. With the 'scope still connected to the dummy load, turn on the supply again and activate the PTT. Increase the level of the modulating tone and adjust the 'scope for a nice picture of "DSB bubbles". Increase the tone level to the point where the "bubbles" barely starts to flatten. With 12.0 V supply, the output amplitude should now be 20 volts peak-to-peak. This corresponds to 1.0 watt P.E.P. into 50 ohms. The overload LED, D5, starts to light up when the final amplifier begins to clip the "bubbles". The sensitivity of the overload indicator can be set by selecting the value of R30. The clipping should be symmetrical. Unsymmetrical clipping may be an indication of an unsuitable final transistor. This will be the case if you try a BD137 instead of the recommended BD131. As the current gain of the BD 137 decreases rapidly at higher collector currents, this transistor is not suitable in a class A amplifier biased to a quarter of an ampere... If you are unable to get modulation, something could be wrong with the microphone amplifier. Check that there is about 6 V (half the supply voltage) on IC3 pin 6 in during transmit. This holds true for the overload indicator as well, about 6V on IC4 pin6. The loudspeaker should be completely quiet during transmit. If it isn't, there is some trouble with R24 and D6. These components are supposed to knock over the audio amplifier, IC2, when the T/R relay puts +12V on the +12TX line. THE DRAWBACKS OF DSB: Unfortunately, DSB transmitters are not compatible with direct conversion receivers. A DSB transmitter sounds like a gargle in a DC receiver. This is due to the fact that the slightest difference in frequency between the transmitter and the receiver will offset the two detected sidebands from each other so they will interfere, add and/or subtract. With Optimist-80 you can only work stations that use SSB. They do. Most of them. Hi. DSB occupies twice the bandwidth compared to SSB. With a single watt of output power, you simply don't care. In addition, since you have a direct conversion receiver, you will be able to hear activity on both sides of your carrier frequency. Since the remote station is listening on your transmission on a single sideband, half of your precious Watt is wasted. But, you can tell the other station that you are running 500 milliwatts. The other half a watt is "on the other side". THE ADVANTAGES OF DSB: One word: Simplicity. THE BOTTOM LINE: The receiver line-up, BF245A - NE612 - LM386, is one of my favourite combinations for simple receivers. I have tried it on more than one band. This line-up was originally inspired by a hand-held fox-hunting receiver designed by SM5CJW. The three stage linear amplifier was designed by Jeff Damm, WA7MLH. I found it in the book Solid State Design, published by ARRL. I just replaced the transistors with suitable european ones. The dual purpose circuit around the NE612 (TX and RX), the receiver front end, the varicap VFO, the overload indicator and the PCB layout are of my own design. You may use the Optimist-80 design in any way you want as long as it is for a strictly non-commercial purpose. DISCLAIMER: There is no performance guarantee at all. All I can say is that I believe the design to be repeatable, and, that the first three Optimist-80 rigs that have been built, is working UFB, as expected, and that they have given their builders lots of QRPp pleasure. Where the component values in the schematic diagram differs from those in the component list, follow the list. Thanks to Hannes/SM6PGP for test-building the first protoype on this PCB. I hope that you'll have as fun with this rig as I'm having. CU on 80. Comments and suggestions are welcome. Vy 72 de SM6LKM Snail mail: Johan Bodin Hjortryd, Ljungkullen S-516 90 DALSJOEFORS SWEDEN Email: sm6lkm.jbeab@swipnet.se Internet home page: http://home4.swipnet.se/~w-41522/ End of file.