Tube amplifier for headphone use
design was found on Jan Deckers' website. He named it Blue Ears. This site is in Dutch but the schematics and the photographs contain much information.
Loudspeakers cannot be used on this amplifier. The design is simple with a very limited number of parts. See the schematic diagrams of the amplifier and the power supply. The amplifier is of the single-ended triode type and is based on the ECC99 triode made by JJ. The headphone impedance should fit the output transformers. I use a 32 Ohm headphone, but this does not seem to be very critical (as discovered later). Total price of the amplifier will be around 500 euros. Since the number of parts is small, I took high-quality parts. With lower quality parts, the price will be slightly less.
Where did I buy the parts?
| Mains transformer, output transformers and chokes for 'Blue Ears' | AE |
| ECC99 triodes | AE |
| ALPS Blue potentiometer | Triodes |
| Resistors for the power supply (Kiwame) | Triodes |
| Grid resistors (VTC) | Triodes |
| Cathode resistors (RikenOhm) | Triodes |
| Cathode capacitors (Black Gate) | Triodes |
| Ceramic tube sockets with gold-plated contacts | Triodes |
| EZ80 rectifier tube | Rotte |
| Gold-plated RCA chassis parts (for connecting the interlinks) | Art Speak |
| Power supply capacitors (JJ) | AE |
| Wire, Silver wire, Teflon (PTFE) tubing | AE |
| Silver solder (WBT) | Speakerland |
All other things can be obtained from regular shops for electronics parts. Since Triodes recently changed the assortment, part of the above parts probably cannot be obtained from them anymore. Searching the internet will surely yield alternative suppliers.
The top plate was made from aluminum plate, 2 mm thick, 250 mm * 350 mm. After having made the layout, I first drilled all holes in a 9-mm MDF sheet of the same size. This was used as a template. This prevented the need to make markings on the aluminum plate. Besides, there was no risk that the drill would move over the aluminum surface before drilling the hole.
The larger openings were made with a power jigsaw. The openings for the tube sockets were made with a manual jigsaw.
After abrading and primer coating, several paint coatings were applied.
For the aluminum part on the backside of the amplifier, the same techniques were used.
Below are some photo's. All photo's enlarge after clicking.
Next, the transformer was primer coated and coated with semi-gloss black paint. The orginal top cap was replaced by a cap in the color of the output transformers. This was mounted using threaded rods and stainless steel closed nuts. The transformer is now ready to be mounted on the top plate. The original cap was re-used inside the amplifier to shield the transformer.
Yellow pine (18 mm) was used for the housing. The local wood supplier did the sawing. I used a wood router to make a nice fitting on the wood edges in the corners. After glueing, the router was used to make round outer edges.
Some wood parts were glued to the inside wall to support the top plate. Final treatment was with teak oil.
The four holes for the threaded rods of the means transformer got a rubber insert for supporting
Self-adhesive soft neoprene tape was used to close the gap between transformer and top plate. The four nuts should not be tightened too much. Some kit is used to prevent them from loosening.
The soft neoprene tape was also used on a rack that I made from some MDF. The rack supports the top plate during soldering the parts.
I made an aluminum angle section under the top plate to prevent sagging.
After finishing the top plate, soldering could start. Connection layouts of the voltage regulators and the tubes are available in datasheet LM317, pin layout EZ80 and pin layout ECC99
The LM317's are used for the DC heater voltage of the ECC99's. The schematic diagram for the heater voltage supply (one for each channel) is not given on Jan Decker's site but can be found on the first page in the LM317 datasheet (bottom left). For C1 and C2, I used 10 uF, 16 V elco's. R1 is 125 Ohm, R2 is a 1 kOhm potentiometer. The 9 V AC from the mains transformer is rectified using a 2 A full wave rectifier and a 4700 uF, 16 V elco. Note that an isolating sheet under the LM317 is needed.
On the photo below, the rectifiers and LM317's are mounted.
On the threaded rods, the original top cap of the transformer is mounted. On top of it the heat sink for cooling the LM 317's.
Here is the heating current wiring for the two triodes. The minus of the DC supply is connected to the star ground point.
The DC heating voltage for the triodes was set at 6.3 V. Then I placed the triodes, so that current would flow. This caused a surpringly hot heat sink within several minutes. Since there will be little to no ventilation inside the finished amplifier, I decided to move the LM317's to the aluminum angle section so that the complete top plate would be the heat sink.
After completion of the wiring, the photo's below were made. Some remarks:
Here, the wood housing is present. The amplifier is still on the MDF rack.
Two photo's from the finished amplifier. The two interlinks are made from remnants from the silver and teflon tubing, again without shielding.
Colleague Rob S. offered to do some measurements on the finished amplifier. I was quite curious how the amplifier would perform, and it was an additional check to see if everything was done well during the construction.
The measurement data are in an MS Excel spreadsheet. The frequency response was nicely flat, with only in the very low frequncies a slight difference between the channels. This is not audible.
Quite surprisingly, the amplifier also worked well with a high-ohmic load (250 Ohm). This was confirmed by listening after the measurement session. Both my 32 Ohm headphone and Rob's 250 Ohm impedance headphone gave great sound.
After finishing the amplifier, I could plug in the mains plug and swich it on. Quite a special moment. Will it work or not?
First, I used it without a source. I heard absolutely nothing in the headphone, even not with the volume full open. No noise, no hum, just nothing. Was it non-working, or was it just perfectly silent? Furtunately, the latter was true.
Then I put the amplifier two weeks on the mains and played music regularly to breaking it in.
In combination with my Philips HP910 headphone I got an impressively nice sound, very rich in details that is very joyful to listen to. The only direct comparison I could make was by using the headphone socket of my preamplifier. Although this preamplifier is not really low-end, the difference was huge, and the tube amplifier was the better of the two.
Unclear to me is if this high-quality sonics originates from the simple design, the quality of the parts or the single-ended-triode principle or a combination of these items. My conclusion is that my first amplifierr project yielded a very nice result that makes listening to a headphone much more attractive. I will use it regularly. But the downside is that the project is finished, since the construction work was nice to do....
Here I present my apprach on record cleaning with fluid suction. My record cleaner is a Knosti, filled with Quality Service Vinyl Cleaner. The drawback is that you have no fluid suction. Records need a long time to dry and there is a risk that loosened particles redeposit on the record after the fluid has evaporated. Vacuum suction was realized by modifiying a vacuum cleaner crevice tool as described on http://www.teresaudio.com/haven/cleaner/cleaner.html . A 'leak' was drilled in it to get the correct suction performance. I use it with our normal household vacuum cleaner. The very low amount of liquid will not damage the vacuum cleaner.
Turning of the record should be done by hand, but that is hardly a drawback. The vacuum fluid suction works very well in combination with the Knosti and is much cheaper than the cheapest record cleaner with fluid suction on the market.
In the photo series, the record is not wet. Nevertheless, the photo's will make clear how it works.
The player is of the non-suspended type, which means that there is no spring-supported sub-chassis. Vibrations are kept to the minimum by making a non-resonant high-mass plinth. Since the total construction becomes quite heavy, a very stable audio rack is needed. Best approach is to make a stable wall-mounted support, although a stable table is also possible.
The DC motor is in a separate housing, thus minimizing the risk that motor vibrations reach the player. The platter is driven with a very thin thread. Not some special thread, it is just sewing thread.
The bearing is a relatively expensive part, but very important for the quality of the player. The Scheu bearing supports the platter on a ceramic ball. The bearing is oil-lubricated.
The Cantus unipivot arm is a 9" arm (needle to pivot 229 mm, pivot to platter centre 212 mm). An MC-to-MM amplifier is placed as close as possible to the player. Its electronics is discussed in a separate chapter. This amplifier feeds the MM phono input of a highly tweaked Rotel preamp (a.o. better opamps, Black Gate capacitors, better internal cabling, all done by Artspeak).
I preferred a design with round shapes only. I built the plinth from a stack of eight parts made from 12 mm MDF. One template was made, which was copied eight times using a wood router. In the middle four sheets, an opening was made. After glueing the parts, this could be filled with lead beads (approx. 10 kg). I obtained the lead from Thomas Scheu, but it is readily available from diving sport shops as well. On top of this a lid is fixed with bolts. Note that my design allows for the use of two arms. At the moment, I have one arm and I have no plans for a second arm right now. However, the extra arm support was no extra work and it keeps the possibility open to use an additional arm.
The round forms are also used for the motor housing and for the MC-to-MM amplifier housing. These enclosures were made from 125 mm PVC tubing (for sewers) with a CD (120 mm) on top. The motor housing was made stable by putting in a high mass (some pieces of M18 steel threaded rods and concrete). All MDF parts for the tubular housing were made using a circle cutter in a workbench drilling machine. On the motor housing, two knobs were needed for the potentiometers for 33 and 45 rpm speed control. Since I could not find nice chromium-plated knobs a made them from a towel rack that I got for almost nothing at the local second-hand store.
All parts were painted black by either using spray painting or a roller. For getting an equal coating, the tubular parts were spinned slowly on an old 2 euro record player from the second-hand store during spraying.
Since the player does not have a subchassis, it requires a rigid support. I made a support from 18 mm yellow pine. Inside this support, I used three 18 mm plywood sheets. Total thickness of the wood part is 72 mm.
The shelve of the furniture where I would like to position the player was not really rigid. It was a single sheet of 18 mm fiber board. Therefore, I drilled four holes in this shelve. These allow for threaded rods between the wood support and a support frame that is fixed with large screws onto the concrete rear wall. In this way, it looks as if the wood support lies on the furniture shelve while it is actually fitted to the wall. The nuts on the threaded rods are used for leveling the record player.
After having built everything, the final setup had to be done. Guidelines how to setup the Scheu Cantus tone arm can be found here and more general guidelines on cartridge setup can be found on the internet, e.g. here. What has be done?:
Setting up the motor is simple. Make a string from the sewing thread by making a single knot in the thread. Shift the motor housing such that the tension on the thread is minimal but still enough to drive the platter. For setting the correct motor speed you will need a strobe disk and preferably a strobe lamp as well.
Ready for listening! The sound of this player is impressive and I think building a record player yourself gives the possibility to realize a great player for a relatively modest price. And I hope I made clear that building a record player is less difficult than it initially may seem to be.
Ever since the introduction of the CD, I kept my record player in the living room. From time to time, I listened to my records although not too often. Some months ago I bought a large number of records via the internet and that was the start of my renewed interest in record players.
Quite many attractive record players are available. But during my search, I happened to find out that a very limited number of suppliers also sell parts for DIY activities. This opened the possibility to realize a high-quality record player for a relatively modest price. Moreover, it gave the opportunity to realize the visual appearance that I preferred. And it gives a good feeling to build something yourself. Some inspiring examples of DIY record players and useful information can be found, e.g., on the Krishu website.