Sunday, October 19, 2014

Bottleneck Meet October 2014

We met as planned at Steve's house to hold the competition. Since both of my entries were still not working, we just chose to listen to the various amps that turned up. No competition. If you click on the pictures, you can enlarge them.

My first entry was a 4P1L push-pull parafeed amp with light bulb loads. The driver for each channel is a single 4P1L and the other amps are also 4P1L tubes. I had tried to use the 4P1L tubes back in the Russian line-stage competition but without success. My inspiration came from Andy Evans single ended amp. (schematic here). For the DC heaters I used the switched mode DC supplies mentioned in the forum by Henry's Cat. If you look at Andy's schematic, you will see that he uses filament bias which is supposed to sound better than cathode resistor bias. I initially tried that with some 15 ohm 50 watt resistors mounted on heat sinks that I mounted on standoffs. The little DC supplies appeared to work fine.

The heat sinks got quite hot which I thought could prove problematic with the polycarbonate that I often use for my amps. The 4P1L tube can be sensitive to vibration so I did not want to use steel or aluminium. I then tried cathode resistor bias and it worked briefly in one channel but there were problems in the other channel that I was not able to solve in time. It was a real mess to wire.

The interstage transformers that do the phase splitting are a couple of cheap ones from Allied that I had in my stash. You can read about them here

For the parafeed push-pull, I used this schematic with lightbulbs substituted for the DN2540 as resistors. The output transformers were Antek 2212 toroids. They are huge 200w units and were salvaged from another project. I had used toroids in the Ikea 6V6 competition and although they take a while to break in, they work fine in a parafeed arrangement.

The case was an old receiver that died when the cat vomited on the power supply. Catty-watty lived as you can see in the first photo. I kept running out of space so the huge toroid power transformer and choke were suspended underneath. Those are tomato cans from Aldi that hold it up. I also had to extend the amp vertically. Weight 41 lb.

 I am now going to dismantle it and turn it into a single ended amp with filament bias and much bigger heatsinks. I won't use the old receiver box since it is too difficult to work down from the top.

Years ago when we used to meet in Elkton, Doug gave me a power transformer and a few organ amps and chassis. John Hogan gave me a couple of Webster organ output transformers. With an old Hammond 193M choke, it weighs 47 pounds. The caps on the left are non functional other than making it easy to tip the amp over.

You might notice there are no input tubes. I was trying Bill Chater's cascode power-amplifier circuit from Glass Audio Vol 10 Number 5, 1998. It's fascinated me for some time so here was my chance to try it out. He used some KT88 tubes which I don't have so I used some 6HJ5 tubes that I had. I finally managed to fire it up the day before the competition but there was the magic puff of smoke. I have a rule that I wait at least a day after a smoke event so that I can calm down so the amp became the other display item. I am going to persevere with this amp. I figure I can do a John Hogan amp imitation by covering the holes with duct tape. Unfortunately John passed away about 10 years ago but his amps have become collector's items. He was a nice bloke.

We started with Steve's Econowaves.

Roscoe actually built an amp using an old Jolida chassis.

It's a very interesting direct coupled design. Roscoe has given me permission to post the values. The output tubes are 5881 and the inputs are 5965.  Please note Roscoe's followup:

Please make the changes mentioned below before posting... A couple of comments based on some more measurements I took this AM.
  • C4 & C5 should probably be upped to 250v types, as the voltage can be higher than 190V depending on how the output tubes bias up.  The value probably would be fine at anything at or over 220uF (the values in your junk box are always the best ones to use, right?).
  • The power supply turns out to be much higher than I thought, at about 500Vdc.  That runs the dissipation up to higher than 6L6/5881 limits, so I'd recommend either building with a B+ at more like 420-450Vdc, or running 6550 output tubes (which bias up just fine, tried it this AM).
  • Old Jolida amps don't cost all that much, especially if you can find one that needs work, and they make a good base platform w/o the need to do metal work, but keep in mind that you're pretty much stuck with a nearly 500Vdc B+ if you go that route, unless you do some serious work on the power supply filtering.

Considering that Roscoe finished it a couple of hours before the meet, it sounded very good, much better than the digital amp we started with.

Some of the DACs etc we used. The top photo shows an early Bottlehead design.

I had asked Roscoe to bring his open baffle speakers since I also have the Goldwood 18" GW-1858 high QTS speakers which I currently have in H frames that my wife objects to. The other speaker is a Tangband W8-1808.

The design comes from JE Labs and is based on the old Briggs sand filled baffle speaker design. More about it here.


The handles are useful for carrying the beasts.

We were mightily impressed by how low the bass would go with these beasts. I liked the sound of the Tangbands and it all integrated very well. I've told my wife that I'm going to build the Mondrian version that you can see at the bottom of the link. Since the Tangbands are quite expensive I will think about some alternatives.

Roscoe also brought along the Western Electric design amp that he demoed at the prior meet. It was richer and fuller than the Jolida chassis amp, but remember that the latter was only finished that morning. Later Roscoe substituted the 5881 tubes and there was a smoke event.

David brought along an ST-70. It had the traditional tube sound with a weaker treble and bass than Roscoe's amps. The ST-70 front end was traditionally thought to be its weak point and now I can understand why.

Tony bought along a set of Climax speakers that he was planning to wire up during the meet.

The crossover box.

Hard at work. Tony also brought along a few headphones to try out which I found very useful since I spend a few months each year travelling now that I am retired. The inner ear phones just don't work for me and I think I will investigate inexpensive closed back over ear models that can take a beating.

 Unfortunately I was not able to hand around to hear the Climaxs. It was my 39th wedding anniversary and my wife was preparing a special dinner. Salmon rat coffins (salmon in puff pastry) and a fig jam souffle for desert.

There were only five of us there but we had a great time and heard plenty of interesting music. We are thinking of holding another meet in January. It will be a last minute scheduling affair depending on the weather. Thanks to Steve for hosting.

For those who are interested, here is a link for my train trip from Narvik in Norway to Singapore which occurred earlier this year.