About a year ago, someone left a comment that really looked like spam on one of my pages advertising a web store selling tube audio gear and parts for guitar and audio amplifiers. Against everything I’ve ever learned about behaving on the Internet, and my better judgement, I ended up actually buying a kit from that vendor Mable Audio located in Shenzhen China. They have several kits available with a few options: 6V6 or EL84 for output, 6SL7 or 12AX7 for the input stage and also a beefier EL34/12AX7 amplifier for about double the price. They have a ton of components and guitar amplifier parts and kits as well.
I selected the 6V6/6SL7 kit because I like the larger tubes, and had a set of 1937-issue 6V6G tubes that look more interesting than the small GT-style tubes it came with.
The kit arrived as a box of miscellaneous parts and the chassis. There was no particular documentation included, but the schematic was sent to me via e-mail. Also, the silver decorative plate (in the center of the photo) is no longer included in these kits despite being pictured, as I was told after I purchased. Not the end of the world, though. The kit included a large power transformer, two output transformers (23% Ultralinear), two huge 470uF 450V filter capacitors, all the needed tubes, signal capacitors, connectors, controls, high quality ceramic sockets and lengths of 600V-rated wire to assemble it all with.
The amplifier itself has decent schematics, on paper anyway, although the output impedance is inaccurate on the datasheet: the datasheet, schematic, and transformer color code documents all specify multiple taps but the provided hardware only supports an 8-ohm speaker. This is acceptable, though, as most home theatre speakers are 8 ohm anyway.
INPUT IMPEDANCE : 100K ohm(RCA)
4ohm – 16 ohm8 Ohm
OUTPUT POWER: 12w x2/(ultralinear)class AB P-P
DAMPING FACTOR : >3
FREQUENCY RESPONCE: 20Hz-20KHz(REF.OUPUT)
TOTAL GAIN: 28dB
INPUT SENSITIVITY: 300mV-600mV
S/N: >89dB (HUM NOISE <3mV
CHANNEL BALANCE: <1dB 20Hz-20KHz (MAX.VOLUME)
CHANNEL SERPRATION:>65dB 20Hz-20KHz
TUBE COMPLEMENT: 6P6P(6V6) X4
POWER REQUIREMENT: AC220V OR 110V OR 240V + -5% 50~60Hz
Assembly wasn’t terrible, but the build quality of the chassis was marginal at best. The top surface of the chassis bolts to the chassis pan itself to hide mounting rails and dividers, and the holes on the top panel didn’t perfectly line up with the socket locations on the bottom panel. This meant that after mounting the sockets to the bottom where they belong the top cover wouldn’t slide over them because it was interfering by about 1mm on one of the sockets. It took significant pressure to force it on and made a very loud snapping noise when it did so, but the socket and cover were all intact. It’s definitely not coming off again, though. Additionally, the cover over the transformers is supposed to mount using several screws at the corners, but only two screw holes lined up.
The sockets are ceramic and the sockets gold-plated, and the transformers look appropriately sized, though, so it’s on to building past mounting up the parts. This project was a long time in the making, as I mounted the sockets and chassis up on July 12th 2011 and then did no further work until March 2012. I worked my way through the schematic, roughly right to left. This amplifier is a power supply, and two identical amplifier channels. I’d highlight on the schematic as I completed both mirror halves, and this made it easy to keep track of where I was going.
Some photos showing filling in the components:
After assembly, I had a few left-over parts:
A really weird assortment of leftover parts at that. I declined to connect the headphone output (on advice that it was probably not safe for any headphones that I might actually like due to the connection – not unique to this amp, many other circuits have similar issues) which accounts for 2 of those resistors, but there’s still 13 more + a small capacitor that weren’t called for anywhere. And there are two extra panel-mount RCA input jacks. And a ton of extra wire. But for all the extra they did include, there were 4 resistor types (8 resistors total) that were not included, so it was a trip down to Vetco Electronics in Bellevue to pick up replacements. The replacements are the Red- and Tan- bodied resistors. Not quite sure what to make of that.
I elected to replace the LED power indicator with an NE-2A neon bulb instead, so it would be the same color as the glow of the tube heaters.
There is a minor issue with this one, though. There is an anti-arcing capacitor across the switch. I placed the NE-2A between the switched side of the switch and the AC neutral (it runs directly from AC mains through a 150K 1/4W resistor). With the switch turned off, the capacitor allows AC leakage of a few mA to pass to the cold side of the switch. The current leakage is less than the excitation current of the transformer, so the transformer appears “open” to the low signal as it’s entirely eaten because it’s not strong enough to set up a magnetic field. This causes the current through the 150K resistor and neon….so the bulb lights up when the power is off, now. With the power on, the capacitor is out of the circuit and the transformer is an extremely low impedance, so all the current goes through the transformer primary and none into the 150K resistor and the NE-2A bulb. The result? The power switch indicator is backwards. The light comes on when the device is turned off, and goes out when you turn it on. I intend to fix this at some point, but am not entirely sure what the best approach might be. The volume control itself is abysmally low quality, probably the only part in the kit I actually think is just plain “bad”. The return spring is weak and the switch touchy and it binds up easily.
There was also another issue: I wired the volume control backwards. “Right” is lower, now, instead of “Left” as is the convention. I evaluated fixing this problem, but while poking around one of the volume control’s pins started coming out of its molding and I didn’t want to risk destroying the control, so that’s a problem that won’t be fixed in this iteration.
I did a quick visual for obvious shorts and powered it up the first time with the 6P6P (Chinese 6V6GT) tubes which were included, as I didn’t want to risk my classic tubes. If there’s going to be a problem it’d be on first power up, for any amplifier from anywhere. The transformers all made a great *THUNK * noise when they energized for the first time, but quickly de-energized again. I poked around some more and found I’d missed a solder connection on the bridge rectifier. I soldered it back on and tried again and everything was fine! First power-up successful, with no re-work needed.
Then again on the test bench with the 6V6G tubes:
Still good! And for some post-production glamour shots:
Now, I’ve moved the amplifier onto my desk and will use it for personal stereo when not using the Surround Sound receiver.
In conclusion: Mable Audio is a reputable supplier. Their product sounds great, and was of acceptable build quality for an entry-level amplifier. The chassis and platform itself will be a great starting point for more modification projects in the future. I rate this kit as a 3.5/5 due to the mechanical issues with the chassis fit and the power switch, but the electronics quality and circuit design seem to be solid. I would recommend this kit to anyone who wants to build an inexpensive tube amp from parts – but, given the complete lack of instructions or documentation beyond a schematic and some color code diagrams, make sure you go slowly and check your work and are somewhat familiar with how to read electrical schematics.
- Pro: Inexpensive for what it is.
- Pro: Good quality components, for the most part.
- Pro: Attractive styling.
- Pro: Fairly straightforward build.
- Con: Terrible quality power switch. Just terrible.
- Con: Mechanical fit on mine wasn’t the best.
- Con: Only 99% of the parts in my kit were the right ones.
For $175 + shipping from China, it’s not a bad deal. Similar kits from more well-known sources, or U.S./European makers, run at least double the price. Once assembled, it’s very attractive and has a lot of potential for future upgrades. This was a fun and rewarding project and a nice change of pace from vintage gear while still keeping the tube connection.
A local friend is building a rat rod out of 1920s-1950s parts, a custom collection that ultimately will turn into a very fast car powered by a huge V8. He found a vintage car radio to go with it, the perfect addition and gave it to me to fix up. He requested to leave the metal cabinet alone so he could paint it to match after the car’s color scheme is finalized, so don’t worry too much about the finish.
This radio, the 4-B-31 “Roamer” was built by Firestone Tire & Rubber, the same company that today makes tires interestingly enough – they used to have a bigger product line when consumer buying habits favored combination stores. It’s a six-tube radio with a broad RF amplifier stage. Most likely the radio bolted up under a pickup truck’s dash and connected in the back to the firewall.
The tubes are 6SK7GT 6SA7GT 6SK7GT 6SQ7 6V6GT 6X5GT. The radio operates off a 6V car battery. With the low voltages it’s only about 1.2W of output power so will never be that loud, but when highways were new it was a lot quieter on the road and probably sounded better.
The battery directly powers the 6.3V filaments of the tubes, and the high voltage is provided with the help of a vibrator power supply. The 6V is fed into the electromechanical device which rapidly vibrates between two contact points turning the DC into a square-wave AC which is fed through a transformer to step the voltage up, then into a conventional rectifier power supply.
One of the pins was broken on this original vibrator, so it was the first to go. I replaced it with a solid-state replacement that uses a few transistors in a multivibrator circuit to accomplish the same effect, and should never need to be replaced again. I also replaced the 6X5 with a pair of 1N4007 diodes in an octal tube base, although this isn’t shown in any photos.
The chassis was decent to work on. It had open sides which made it easier to get things in with tight tolerances. The resistors tested decently, but all caps did need to be replaced as always. Several had blown their ends off already.
This radio was of course designed to be used in a car, and that means used with a car radio antenna which is a specific length and has certain transmission line characteristics – not quite as simple as just stringing out a long-wire. It’s a standard antenna, though, so I ordered a replacement that cost something like $10 with free shipping from Crutchfield.
It arrived in interesting packaging. The box was clearly broken in half, but both halves made it to my door without actually being connected somehow.
The antenna was in the bigger section. Go UPS?
A terminal strip in the radio was broken. This was a problem because the broken terminal happens to be the positive power lead-in and it couldn’t be salvaged. Only one terminal broke, though, so I improvised, screwing a screw lead to the mounting bracket and securing as shown, then running the wire out of the case.
Reassembled and testing with a bench power supply that was okay to check functionality. The switching power supply introduces too much hash to receive any stations, but it was good enough to do an alignment with a signal generator by injection. I then switched to a lantern battery for final tweaks which had a disappointing life of about 10 minutes. Clearly these were meant to be run off lead-acid batteries or linear power supplies only. It draws around 4A.
I also rewound the dial indicator. The dial tuning drum was still wound properly but the dial indicator string had broken so the pointer no longer moved. I used string that was a bit too thick but it worked out okay and is perfectly functional. No photos of that available though, it was pretty quick. The service manual had a full dial string diagram and pointer adjustment procedure. Unfortunately I ran into a problem as I was reassembling everything: the volume suddenly dropped off massively even with the control maxed out and it wasn’t coming back for anything. A check of the voltages showed that I had tens of volts on the screens of most all the tubes, where there was supposed to be a few hundred. I was at a loss about why this happened and finally resorted to the poke test.
The poke test is what it sounds like: poking or tapping on pretty much every part in the radio. I gave decent raps on all of the solder joints, tube pins, tie points and finally came to one that would make the volume cut back and forth: R7, the B+ dropping resistor for the screen voltages, a 15K 1W carbon resistor. Apparently it was internally cracked or otherwise defective. I replaced it with two 30K resistors in parallel to form a 15K 2W resistor, and a few others that shared the same tie point or were otherwise looking rattier than I really like even if they were in spec.
With that repair completed, the radio fired up perfectly with loud volume. This was a fun project, but power supply issues mean I don’t think I’ll take on too many of these in the future.
I’m looking for a new summer project, and think I’m going to branch off from repairing radios to building some of my own gear. I’m interested in making a tube hi-fi stereo amplifier to go with my old speakers – and it’s something I can build. Audiophiles, guitar enthusiasts, and other people into tube audio have driven the price of most of the good audio tubes higher than I’m interested in paying, though.
The 6V6 audio tube is known and loved by enthusiasts everywhere, and came in a variety of other versions with different ratings, different glass, different power supply requirements but pretty much all produce the same output. I’m interested in making a miniature-tube amplifier, so I’ll be starting with the 6AQ5A tube – but these are pretty expensive. An alternately powered version, the 5AQ5 goes for only about $2/each so that looks like a winner.
I’ll be basing my amplifier on the Silvertone 69A, a dual-channel push-pull stereo amplifier manufcatured in 1960. This amplifier used a 5U4 rectifier tube to supply the high voltage. By replacing the tube rectifier (right, above) with a pair of 1N4007 silicon diodes (left, above) I’ll free up the 5V power supply. A set of four 5AQ5s draws 2.4A of current; a 5U4 draws 3A of current so there’s more power to spare and a guarantee of being able to find cheap hi-fi tubes for years to come without a lot of competition.
The schematic – with some bias and output mods – not showing the power supply or output tube modifications – that I’m basing the design is here (click for larger version!):
This should be a good sounding little amp if I get around to actually completing construction. I have most of the iron [transformers] in stock already, it’s just a question of buying some hardware and spending the time required to wire it. We’ll see.