Fullscreen HTML5Med Res (??MB)HTML4

By Nicholas Vinen The Currawong Stereo 10W Valve Amplifier, Pt.3 In the last two instalments, we introduced the Currawong valve amplifier, described its circuit and gave the PCB assembly and wiring details. This final article describes the optional remote volume control, the acrylic cover and the setting-up procedure. Y OU DON’T HAVE to build the remote volume control board but we think most constructors will want to. It’s just so convenient when it comes to setting the volume and is far easier than having to wander over to to wind the volume control up or down. If you intend building the remote control into the Currawong, you should have already fitted the motorised pot to the main board. The 50 x 50mm remote board hangs from the front-right corner of the main PCB via a tapped spacer and is connected via a 4-pin header. There is also a connection from the remote control board to the pot motor. 64  Silicon Chip If you aren’t fitting the remote control option to your Currawong amplifier, skip down to the “Initial power up & testing” cross-heading below. IR remote control circuit The remote control circuit is shown in Fig.12. It’s based on the low-noise remote-controlled preamplifier used in the Ultra-LD Mk.3 Stereo Amplifier described in the November 2011 issue. Basically, we took the remote control parts used in that project and put them on a separate PCB, without the preamplifier circuitry (which is already present on the Currawong’s main board). It works as follows. The remote control is set to generate Philips RC5 protocol codes which are picked up by infrared receiver module IRD1. Its output goes to pin RB0/INT on PIC16F88 microcontroller IC2. IC2 decodes the remote commands and if it detects a relevant code (volume up/ down/mute), it then uses its RB1-RB4 output pins to drive transistors Q10Q13 which are arranged in an H-bridge configuration, to drive the pot motor in the appropriate direction. A 1µF capacitor is connected across the motor terminals on the PCB to reduce hash from the motor brushes siliconchip.com.au CON11 22Ω + 1 5V REG 2 7805 LED1 GND 100 µF 25V LED2 +5V OUT IN 10k 1 0 0 µF 100nF GND 4 100Ω 1 µF MMC IRD1 3 1 λ 3 x 10k 2 3 4 5 6 7 8 9 10 (INPUT BOARD NOT USED) 17 3 LK7 LK7 5V: MUTE RETURN 0V: NO MUTE RETURN CON13 1 +5V 1 0 0 µF 2 14 MCLR Vdd RA0 RB3 RA4 RB4 1k 9 '1' 12 '2' 13 '3' 11 IC2 PIC16F88-I/P RB6 RB1 RB7 RA1 RB5 RB2 X1 4MHz 22pF 22pF 16 OSC2 AN3 OSC1 RA2 K MOTOR – 1k 7 1k 8 Q11 BC337 10k B C E Vss 5 Q14 BC547 1 µF MMC C B 2 Q13 BC337 E ENDSTOP ADJUST VR3 1k 18k B C E 10Ω 100nF A SC CON12 K 1N4148 20 1 5 E C 1 µF MMC D8 1N4148 18 1 B C +5V 15 Q12 BC327 E A D7 1N4148 RB0 B 1k 10 A 6 Q10 BC327 + FROM MAIN PCB (CON10) K CURRAWONG REMOTE VOLUME CONTROL B E 1 C 7805 IRD1 BC327, BC337, BC 54 7 2 3 GND IN GND OUT Fig.12: the circuit for the add-on remote volume control is based on the one used in the Ultra-LD Mk.3 Stereo Amplifier (November 2011). The infrared signal is received by infrared receiver IRD1 and passes to microcontroller IC2 which decodes it and uses Q10-Q13 to drive the pot motor in the required direction. Power comes from the main board. while there is also a capacitor soldered directly across the motor terminals, at the other end of the figure-8 wire from CON12. IC2 monitors the motor current across a 10Ω shunt resistor. The feedback voltage is adjusted using pot VR3 and goes through a low-pass RC filter (18kΩ/100nF) before being fed to analog input AN3 on IC2. IC2 can thus detect the increase in current when the pot hits one of its end-stops. This feedback is used for the mute function. When mute is pressed, the motor is driven anti-clockwise until the pot hits its minimum end-stop. IC2 detects the increase in current and shuts the motor off once minimum volume has been reached. If mute is then pressed again and LK7 is in the high position, the motor is driven clockwise for the same time as it took to reach the end-stop, thus returning the pot to the original volume level. For this to work, VR3 must be adjustsiliconchip.com.au ed correctly. If it’s set too high, the motor may stop prematurely while if set too low, the motor may not stop once minimum volume has been reached. In the November 2011 design, IC2 flashed an acknowledge LED to indicate when a valid remote control command was received. We have used the same output (RA2) to drive NPN transistor Q14 which pulls the cathodes of small signal diodes D7 & D8 low in acknowledgement. These go to either end of red/green LED1 on the main board via pin header CON11. As a result, when a command is received, LED1 is shorted out and so it flashes off briefly. This avoids the need for an extra LED to be fitted for the remote control function. The only change in the microcontroller software compared to the UltraLD Mk.3 remote preamp is to increase the time that pin RA2 is driven high upon receipt of a valid remote command. That’s done in order to make the LED flashing more visible. PIC microcontroller IC2 uses 4MHz crystal X1 for time-keeping. This is required as the remote control commands are sent at a particular frequency and the micro needs to be able to “lock on” to these commands to properly decode them. Multiple input option We’ve kept the original design’s Transformer Bolt Earthing – Warning! Note that the mounting bolts for mains transformers T1 & T2 must not be separately earthed (ie, via earth leads) if the amplifier is mounted in a metal chassis.That’s because running earth leads to them would result in a shorted turn on each transformer and this would immediately blow the fuse in the IEC socket. January 2015  65 sistors here, since they fit more easily. Solder the IC socket in place next, with its notched end to the left, followed by REG2. Prepare the regulator by first bending its leads down through 90° about 6mm from the tab, then attach the tab to the PCB using an M3 x 6mm machine screw and nut. Make sure the screw is done up tightly before soldering and trimming the leads. The ceramic capacitors can go in next; their polarity does not matter. You will be left with a 1µF type to be soldered across the motor terminals later. Follow with the small signal transistors, taking care to avoid getting the three types mixed up. Crank their leads out to fit the PCB pads using small pliers. If you have a low-profile 4MHz crystal, this can be fitted to the top of the board as shown in Fig.13. Otherwise, you will need to cover the metal can with a short length of 10mm diameter heatshrink tubing, shrink it down, bend the leads through 90° and fit it to the underside of the board so that it’s laying horizontally under PIC micro IC2. In this case, solder its leads on the top side of the board. Note that in our photos, X1 is shown bent over to the left but this was found to interfere with the mains power switch when the board was in place, so we later moved it to the underside and bent it in the other direction as described above. The right-angle polarised header for the motor is also mounted on the underside of the board, with its pins facing the righthand edge, for the same reason (again, shown differently in the photo). Solder its pins on the top side. X1* 22pF LK7 SILICON REG2 REG2 CHIP 1µF 7805 Motor 10k CON12* Q11 1k 1k 1k 100Ω 1µF + SEE TEXT Q10 100µF 1µF Q12 Q13 + 44111110+ 01111144 18k 100nF 22Ω 10Ω D8 4148 Q14 4148 D7 VR3 1k * * IRD1 CON11 4MHz 22pF 10k 1k IC 2 PIC16F88-I/P 100µF CON13* 100µF Remote Volume 10k 10k 10k ADD RESISTORS SEE FIG.13 C 2014 MOUNT ON BACK OF PCB SEE TEXT ON BACK OF PCB Fig.13: follow this parts layout diagram to build the remote volume control PCB. This sits just below the main board, so the available component height is limited. As a result, motor header CON12 and crystal X1 (if full height) must be fitted at right angles on the underside of the PCB (not on top as shown in the photo). In addition, the electrolytic capacitors should be pushed all the way down to the board before soldering or else bent over so that they will later clear the main board assembly. 10-pin header CON13, which was used to connect to two other PCBs for input switching. This enables the possibility of fitting multiple inputs to the Currawong and having remotecontrolled switching. This would require the main Currawong board to be built into a larger case with enough room for the extra inputs and the relay board required. In the standard Currawong design, (ie, no input switching), we just connect 10kΩ pull-up resistors from pins 7 & 8 (+5V) to pin pairs 1/2, 3/4 and 5/6 as shown so that the unit will function without the input switching board connected. Power for the remote control unit is derived from the Currawong’s unfiltered low-voltage DC rail of around 15V via pins 1 & 4 of CON11. This supply goes through a low-pass RC filter (22Ω/100µF) before being fed to a standard 5V regulator, REG2. The 5V output from REG2 is used to power the micro and the motor but is further filtered using a 100Ω resistor and 100µF capacitor for infrared receiver IRD1 (plus an extra 1µF ceramic bypass capacitor) in order to prevent motor hash from interfering with infrared command reception. Remote PCB assembly The remote control PCB is coded 01111144 and the parts layout is shown in Fig.13. Start by fitting the two diodes, cathode stripe to the left, then follow with the resistors. You can check their striped bands against the resistor colour code table (Table 3) however it’s also a good idea to measure them with a DMM as the colours can be hard to read clearly. Note that while most of the resistors are laid flat in the traditional manner, the three 10kΩ resistors soldered to the pads for CON13 will need to be fitted vertically, with two leads sharing one of the holes. We used mini 0.25W re-   Table 4: Capacitor Codes Value 1µF 100nF 22pF µF Value 1µF 0.1µF NA IEC Code EIA Code 1u0 105 100n 104 22p 22 Table 3: Resistor Colour Codes   o o o o o o o No.   1   5   4   1   1   1 66  Silicon Chip Value 18kΩ 10kΩ 1kΩ 100Ω 22Ω 10Ω 4-Band Code (1%) brown grey orange brown brown black orange brown brown black red brown brown black brown brown red red black brown brown black black brown 5-Band Code (1%) brown grey black red brown brown black black red brown brown black black brown brown brown black black black brown red red black gold brown brown black black gold brown siliconchip.com.au The 3-pin header for LK7 and 4-pin header socket CON11 are fitted as usual, to the top side of the board. Put the shorting block over LK7 in the position shown for mute return or fit it in the alternative position to disable mute return. Trimpot VR1 is a vertical type, so that it can be accessed once the remote control board has been plugged into the main board. You will need to bend its rear pin out slightly to fit the mounting pads. The three electrolytic capacitors can then go in, with their longer (positive) leads orientated as shown. The infrared receiver is fitted with its leads bent so that the bottom of the receiver is level with the PCB but it is spaced about 6.5mm away from the bottom of the board – see photo. You will need to bend its leads backwards close to the body of the receiver, then crank them up, then bend them back down again about 8mm behind the body of the receiver to fit through the holes on the PCB. The final adjustment to make the infrared receiver “look” through its front panel hole will be done later, when the board is fitted. You can now finish the remote PCB assembly by plugging microcontroller IC2 into its socket, with pin 1 at left. Installing the remote PCB Solder a 4-pin male header to the underside of the main PCB, at bottomright, to match up with the female header socket (CON11) on the remote board. While you’re at it, feed the leads of the remaining 1µF ceramic capacitor through the holes in the two terminals on the back of the pot motor and solder them in place. Trim off any excess lead. Now you will need to make up the lead for the pot motor. Start by cutting a length of light-duty figure-8 cable so that it will reach from the rear of the pot over to the right-angle pin header on the remote board. Be a little generous, keeping in mind the orientation of the plug and the fact that you will need some slack in order to plug it in. Strip and separate the wires at both ends of this cable and crimp both wires at one end into two polarised header pins. We like to solder the wires after crimping (being careful not to get any solder outside of the crimp section) so that they can’t pull out. Next, push the pins into the polarised block using a small jeweller’s screwdriver. They should click into siliconchip.com.au The remote volume control PCB is attached to a single mounting point under the main PCB (see text). place. If they won’t go in, don’t force them; you may need to pull them out and straighten the “springy” section before they will go in properly. Now solder the other ends of the lead to the pot motor terminals (or to the capacitor leads which are already soldered to them). Unfortunately, there’s no good way to figure out the polarity so you’ll just have to pick one and then reverse the connection if it’s wrong but we’ll get to that later. Next, insert an M3 x 6mm machine screw through the sole mounting hole on the remote control board, head on the underside, with a shakeproof washer under the screw head. Place a Nylon washer on top and then screw it into an M3 x 9mm tapped spacer. Do it up nice and tight. Plug the remote board into the 4-pin header on the main board, then use another M3 machine screw and a flat washer to hold it in place via the provided mounting hole on the main board. Finally, plug the polarised header from the motorised pot into CON11 on the bottom of the remote board and you are ready to test it. Note that the pot motor lead should not be able to reach the mains switch which, in any case, should be completely covered in heatshrink tubing. The next step is to drill a 4mmdiameter hole in the front panel for the IR receiver. This 4mm hole should be positioned exactly 27mm to the left of the power LED (LED1). Having done that, leave the front panel off for the moment, so that you can set VR1 correctly and if necessary, swap the motor polarity. Initial power up & testing When we left off last month, we had built the PCB and plinth, wired up the WARNING! HIGH VOLTAGES High AC and DC voltages are present in this amplifier. In particular, mains voltages (230VAC) are present on the IEC socket and the primary side of the mains transformers (including the wiring to the power switch). In addition, the transformer secondaries together provide a 114VAC output and the power supply produces an HT voltage in excess of 300V DC which is present on various parts of the amplifier circuit (including the output transformers). Do not touch any part of the amplifier or power supply circuitry when power is applied otherwise you could get a severe or even fatal electric shock. The blue LEDs in the circuit indicate when high voltages are present. If they are lit, the power supply and various parts on amplifier board are potentially dangerous. The completed amplifier must be fitted with Perspex covers as described in Pt.3 this month, to ensure safety. January 2015  67 Parts List: Currawong Remote Control 1 double-sided PCB, code 01111144, 50 x 50mm 1 4-pin header, 2.54mm pitch (CON10) 1 4-pin female header, 2.54mm pitch (CON11) 1 1kΩ mini vertical trimpot (VR1) 1 4MHz crystal, HC-49 (low-profile if possible*) (X1) 1 3-pin header, 2.54mm pitch, with shorting block (LK7) 1 18-pin DIL IC socket 1 2-pin right-angle polarised header 1 2-pin polarised header plug with crimp pins 1 200mm length light-duty figure-8 cable 1 9mm tapped Nylon spacer 3 M3 x 6mm machine screws 1 M3 nut 1 3mm ID shakeproof washer 1 3mm ID flat washer 1 3mm ID Nylon flat washer 1 universal remote control (eg, Altronics A1012, Jaycar AR1719) power supply and mounted the PCB in place. Now it’s time to power it up without the valves in place and check that the power supply is working. Start by popping the fuseholder out of the mains input socket using a flat-bladed screwdriver, then fit the fuse (plus a spare) and re-install it. Leave LK4 & LK5 off the board for now. From this point on until the top cover is fitted, be careful to avoid putting either of your hands near any of the components on the top of the board – touch the assembly using insulated probes only. Now set your DMM to DC volts (with a range that goes up to at least 300V), plug in the mains cord, switch on and observe the LEDs. The four blue LEDs adjacent to output transformers T3 & T4 (LEDs3-6) should immediately light. Blue LED2, next to the headphone socket should remain off while LED1 (power) should be red. If your amplifier doesn’t display this behaviour, switch off immediately and wait for the HT voltage to drop to a safe level before troubleshooting. This can be monitored by connecting the negative probe of your DMM to one of the valve socket mounting screws and the positive to the cathode (striped end) of D1. Wait for it to drop below 40V 68  Silicon Chip Semiconductors 1 PIC16F88-I/P programmed with 0111114A.HEX (IC2) 1 infrared receiver (IRD1) 1 7805 5V linear regulator (REG2) 2 BC327 PNP transistors (Q10,Q12) 2 BC337 NPN transistors (Q11,Q13) 1 BC547 NPN transistor (Q14) 2 1N4148 signal diodes (D7,D8) Capacitors 3 100µF 16V electrolytic 2 1µF monolithic multi-layer ceramic 3 100nF monolithic multi-layer ceramic 2 22pF disc ceramic Resistors (all 0.25W, 1%) 1 18kΩ 1 100Ω 5 10kΩ 1 22Ω 4 1kΩ 1 10Ω * If using a full-height can crystal, add 1 x 20mm length of 10mmdiameter heatshrink tubing before touching the board and to 10V before doing any soldering or other work on the board. Assuming blue LEDs3-6 are working properly, these indicate the state of the HT rail. They will be glow brightly when dangerous voltages are present and dim significantly once the HT capacitors have discharged to a safe level. Note that they will continue to produce a small amount of light for a long time after switch-off but will be quite dim by the time the HT rail drops below 10V or so. If these LEDs do not light up, one or more could be installed with the wrong polarity or might be faulty. Once the HT has discharged, you can connect a current-limited voltage source across each LED to check them. Some (but not all) multimeters can light blue LEDs when set on diode test mode. If LEDs3-6 are working but LED1 does not come on, this points to a possible fault in the low-voltage AC wiring, the regulator section or a problem with IC1 or Q5-Q8 and associated components. Check these areas, starting by measuring the voltage between pins 4 & 5 (the two topmost pins) of one of the 9-pin valve sockets, which should be stable at just above 12V and proceed from there. On the other hand, if LED2 is on, that suggests a fault in Q9 or its base resistor or a short circuit in that section of the board. Assuming that you get the correct LEDs lighting, LED1 should turn green about 20 seconds after switch-on. During this time, you can check that the various voltage rails are correct. First, measure the DC voltage between pins 4 & 5 of the 9-pin valve sockets as mentioned above and check that you get close to 12.3V. You can also confirm that there isn’t too much ripple on the regulated supply by measuring the AC voltage between these pins; it should be below 100mV. Now check the unfiltered HT supply voltage, between the cathode of D1 and one of the valve socket mounting screws. You should get a reading close to 320V. The filtered HT voltage can be measured between pin 3 of any 8-pin valve socket and one of the earthed mounting screws. Pin 3 is the pin closest to you, on the right – see Fig.6 in Pt.2 last month. This should give a low reading (a few volts) initially while LED1 is red and then it should shoot up to 318V or so (ie, a couple of volts below the unfiltered HT rail) as soon as LED1 turns green. The other filtered HT rails can also be checked, at pins 1 & 6 of each 9-pin valve socket (lower-right and upperleft respectively). With the valves not yet fitted, these should all be pretty close to the main filtered HT rail at around 318V although they will rise more slowly after LED1 turns green. Testing the remote board If you have fitted the remote control board, this is a good time to test it now that you have determined that the power supply is working properly. First, set your remote control to one of the supported codes. For the Altronics A1012, this is either 023 or 089. For the Jaycar AR1719, use 97948 (Philips 02 CJ 412 TV). Now aim the remote control at the receiver and hold down the volume up or volume down button. You should see the acknowledge LED (LED1) flash and the pot shaft rotate. If nothing happens and you have definitely programmed the remote for the correct code then that suggests either a fault on the remote control board or an improperly programmed PIC micro. Check that the board’s 4-pin siliconchip.com.au header (CON11) is plugged in correctly to the main board and that there is around 15V between pins 1 & 4. If the pot rotates in the wrong direction, you will need to switch off and reverse the motor connections (once the HT rail has discharged sufficiently). This can be done by using a fine flat-bladed screwdriver to press in the retention tabs on the polarised header pins, then sliding the pins out of the housing (while holding the tabs down) and refitting them so that they are swapped around. Once you have the pot motor rotating correctly, press the mute button and check that the pot rotates to the fully anti-clockwise position and stops. If it doesn’t stop, turn VR1 clockwise until it does. If it only rotates part of the way, turn VR1 anticlockwise until it mutes properly. Ideally, VR1 should be set about mid-way between the too-low and too-high positions, to avoid later malfunctions if the pot shaft’s mechanical resistance increases slightly. Note that you may need to manually rotate VR1 clockwise to about half-way (or use the volume up button) before the mute function can be tested. Once it has been set up, you can refit the front panel and if necessary, bend the leads of IRD1 and LEDs1&2 so that they line up with their respective holes. Fitting the valves Assuming that the voltages check out, switch off the power and wait for the HT capacitors to discharge, then plug in all the valves. The sockets will probably be very stiff the first time they are fitted; a small amount of contact cleaner on the pins can help ease them in. Don’t push them too hard; you will need to wiggle them in and it’s better to push down on the octal valves by holding the base rather than the envelope. The 12AX7s have no base but they should require less insertion force anyway. The glass envelopes are pretty strong but they can be broken with enough force and there’s also the possibility of the glue holding the envelope to the base giving way. So slowly wiggle the valves in. After the first couple of insertions, the sockets will loosen up and fitting/removing the valves will be a lot easier. This may seem obvious but we should point out that V5-V8 will get siliconchip.com.au This is the laser-cut clear acrylic top cover for the main PCB assembly (the white colour is a reflection). Not shown are the front cover and the four pieces that are attached as shield plates to guard the output transformer connections. Acknowledgements: we’d like to thank Ada Lim and the people at Sydney hackerspace “Robots & Dinosaurs” for their help with the laser cutter. very hot during operation and you should not touch them! Even brief contact can result in a painful burn. Consider that with the glass envelopes and about 25W dissipation, they are similar to an incandescent light bulb – ie, they get very hot! Now, while we have provided a minimal output load on the PCB (~470Ω per channel), it’s still a good idea to hook up a “proper” dummy load until you’re ready to connect some speakers, to prevent flash-over due to excessive voltage when the amplifier is lightly loaded. A couple of 10Ω 5W resistors connected across the speaker terminals will do, although any value in the range of 3.9-100Ω is acceptable. Turn the volume control right down initially. If you have an oscilloscope and signal generator, you can feed sinewave signals into the inputs, power the unit up, advance the volume control and check the shape of the output waveforms on each channel. Otherwise, all you can really do is hook up a signal source (eg, a CD player) and some speakers and listen to it. Note that there won’t be much output (if any) until several seconds after HT has been applied (ie, LED1 has turned green), as it takes time for the various bias voltages in the circuit to stabilise. And it takes several more seconds until the amplifier can deliver a significant portion of its rated power. The warm-up is complete and the full ~10W/channel is available around 10 minutes after switch-on. Before that, you’ll probably run into clipping at 8-9W per channel. It simply takes that long for the valves to reach operating temperature. Assuming it appears to be working normally, switch off, turn the volume control back down and fit shorting blocks to LK4 and LK5 to enable global feedback. This dramatically lowers distortion, from around 0.5-1% down to 0.05-0.1% (ie, by at least an order of magnitude) so we definitely recommend operating the amplifier with these links in. Now switch the amplifier back on, slowly turn the volume back up and check that it’s still working properly. If you get a high-pitched squeal, you may have wired the output transformers improperly, turning the negative feedback into positive feedback and causing oscillation. You’ll have to switch off and check the transformer wiring and feedback components (resistors/capacitors). Making the top cover The top cover is vital since contact with some of the components during operation could be lethal. We’ve designed a clear acrylic top cover to suit the plinth as described last month, so you can still see all the circuitry while keeping it safe. It also helps to keep dust and dirt off the board (although January 2015  69 ACRYLIC SHIELD PLATES INSULATE  ALL TRANSFORMER TERMINALS  WITH A DOUBLE LAYER OF HEATSHRINK  TUBING NEUTRAL-CURE SILICONE The two shield plates for each output transformer are glued together at right angles and then glued at right angles to the main cover. Some neutral-cure silicone is also used to provide further protection and to help hold the acrylic shield plates in place. Another close-up view of the output transformer shield. Don’t leave this shield out – the transformer terminals operate at 308V DC so it’s an important safety feature. not entirely, since there are cooling slots cut into it). Technically, acrylic plastic is polymethyl methacrylate and is sold (with some variations in the formulation) under several brand names, including Plexiglas, Perspex and Lucite. The cover panel likely won’t be included in any kits but you can purchase it direct from SILICON CHIP (eg, via our online shop). Alternatively, if you have access to a laser cutter with a bed of at least 300 x 300mm, you could cut it yourself. The cutting file is available on our website in various formats including DXF, SVG and PDF (as a free download for subscribers). 70  Silicon Chip We used a laser cutter with a 50W CO2 laser and found that we got good results cutting the 3mm acrylic using two passes at 50% power. Once you have your cover, check which way around it goes (the cutouts are not symmetrical), then slip it over the top of the assembly to make sure that it fits in place and that the plinth mounting holes are not too far out of their expected positions. Leave the protective film on for the time being. If you’re using valves with large envelopes (eg, KT66s) then you may have to remove them in order to fit the cover. 6L6s can be left in place. Push it down until it sits on top of the low-profile 39µF capacitors. If it won’t go all the way down, chances are you haven’t positioned transformers T3 & T4 in the middle of their mounting locations. It’s possible to carefully loosen their mounting screws, just enough to move the transformers, then tighten them again without having to remove the board. Now remove the cover and peel the protective film off the five pieces to be glued. These all have crenellated edges (like a castle rampart, with a series of square protrusions). While super glue (cyanoacrylate) is suitable for gluing acrylic, we strongly recommend that you use a proper, solvent-based adhesive as this will give a much stronger bond. We used SciGrip Weld-On 16, fastsetting “clear, medium-bodied solvent cement”. This states on the label that it’s suited for Butyrate, Polycarbonate, Styrene and Acrylics. You are unlikely to find this type of adhesive in a hardware store but should be able to get it from a plastics supplier. Ours came from Plastix [Sydney (02) 9567 4261; Sydney Northern Beaches (02) 9939 0555]. This forms a strong bond quickly so you only have about 30 seconds to mate the pieces and ensure that they are square before it’s too stiff to manipulate. Full strength is achieved after about 24 hours. The bond is clear but you don’t want to get excess adhesive on the material as it will affect the surface finish and you definitely don’t want to drip it on the cover. It tends to get a bit “stringy” (sort of like melted mozzarella) after coming in contact with the acrylic. In fact, to give yourself the best chance of getting a clean-looking bond, we’d recommend squeezing some of the adhesive out onto a smooth piece of timber or metal (not plastic!) and using a small paintbrush (hair, not Nylon) to apply it to the acrylic. This makes it easier to control how much you are applying compared to using the tube directly. You’ll also need a clean rag on hand. Start by gluing the two pairs of transformer shield plates together. Before applying any adhesive, figure out which surfaces will be in contact (they are on two faces). That done, apply a thin layer of adhesive to all those surfaces, then press the two pieces together. Make sure that they are at a 90° angle and that the tabs are fully inserted into the slots. Wipe off any excess adhesive and be careful not to get it on areas of the acrylic away from the join. You can then lay the part on its side to cure. Do the same for the other identical piece. Note that while there are two different orientations in which these pieces can be glued together, it doesn’t matter which way you do it as they are symmetrical. Once you’ve done those, you can move onto gluing the front and top sections together. This is a much larger join but the technique is basically the same. However, the orientation does matter in this case – be sure to glue the front section on such that when the cover is in place, it hangs down rather than sticks up. Acrylic adhesive is very strong so if you get it wrong, you probably won’t be to get them apart siliconchip.com.au This view shows the amplifier with the acrylic cover in place. It provides an attractive finish while protecting against dangerous voltages. Note that the output valves get hot so be sure to place the amplifier away from young children and where there is plenty of ventilation.   Before Switching On •  Check that the IEC socket’s Earth pin is connected to all exposed metalwork. •  Check the isolation between the Active & Earth pins and Neutral & Earth pins of the IEC socket.   • Check the output transformer and mains switch insulation. The output transformer terminals must be fully insulated with a double layer of heatshrink. •  Don’t touch any parts if the unit is being tested without the cover. •  Be sure to fit the cover when testing is complete. again without breaking something. Again, it’s important to make sure that the sections are at right angles and pushed fully together to get a neat result. You will need to peel away the protective film from the top cover near the front but it’s a good idea to leave it in place on the rest of the panel to protect it during gluing. The best way to do this is to peel back the film around the area to be joined and then use a pair of scissors to cut a strip of it away, so the rest can be laid back down on the surface. Once you’ve joined those parts, leave it for a few minutes and it should then be strong enough to allow you to glue the two transformer cover pieces prepared earlier into the crenellated siliconchip.com.au sections at the front of the transformer cut-outs. Glue the pieces in so that the horizontal pieces at the top project out over the cut-out areas in the top cover below (ie, not pointing towards the front of the panel). Fitting the top cover While full strength won’t be achieved for 24 hours, the joins should be strong enough after about 10 minutes to allow you to (carefully) fit the cover to the amplifier. Again, if using KT66s or other valves with envelopes larger than the 6L6s, remove them first. Lower the cover until it’s resting on top of the five low-profile capacitors. Take care to avoid touching the underside as this may leave visible fingerprints. If you do get fingerprints, polish them off with a soft cloth. You may need to push down on it gently but firmly to get it to go all the way down. If it won’t go, re-check the positioning of T3 and T4 and move them slightly if necessary. You can then mark out the seven mounting hole positions around the perimeter of the cover and drill 2mm pilot holes a few millimetres deep in each location. You can remove the cover to do this if you want to (which makes it easier to remove the resulting wood particles), however it isn’t strictly necessary. Next, peel the protective film off seven of the small doughnut-shaped laser-cut pieces. Once you’ve cleared the area around each hole, slip these “doughnut” spacers under the cover and push them into place (eg, using a screwdriver). You can then feed a 4G x 12mm self-tapping screw in from the top and do it up until the top panel is resting on the spacer. You may want to do up all seven screws loosely and then slightly adjust the top cover position before making them all tight to hold it in place. All that’s left now is to squeeze a small bead of neutral-cure silicone sealant into the gap at the upper-left corner of each output transformer. This helps hold the acrylic covers in place and also prevents small fingers or other objects from being pushed into this gap (see photo). The easiest way to do this is to cut a thin strip of plastic from a take-away container lid or similar, place a bead of silicone on the end and use it like a trowel to push it into the gap and wipe off any excess. Once it has all dried you can plug the valves back into their sockets and the amplifier is ready to go! Note that the output valves get hot in operation so be sure to place the amplifier where SC there is plenty of ventilation. January 2015  71