Suggestion for DIY Tube Amplifier: ~$500 excludng enclosure/tubes

[cotdt]

If the PPX3 SLAM has an output impedance of 30 ohms, then bass will be loose and flabby. High output impedance actually boosts bass response for most headphones, but it won't be high quality bass.

[deepak]

Speaker amps, yes, headphone amps, can't think of one. For speaker amps, see Amity and Karna.

I was thinking about something for my K1000. The Zanden that people loved at the national meet was 5687/300B (can't remember the rectifier).

[Nanoha]

All right, I need a second soldering iron as my old one is utter @#$! now. Shoot me options. Please don't get crazy on me as I dont' have much money to spare.

[aerius]

I was thinking about something for my K1000. The Zanden that people loved at the national meet was 5687/300B (can't remember the rectifier).

In that case, go about 60% of the way down this page till you get to the heading "Even less is even more". I'd swap the rectifier for a pair of 6CJ3 TV damper diodes as they are quieter and can pass more current than the normal 5xx4 rectifier bottles. Use the 6C45P as the input tube if the WE437A tubes ($350 each) are out of your price range, reports say they're nearly and possibly as good. Handily beats the 5687 in drive capability, and should kill it in sound quality. I'm thinking about a set of 6C45P's for my next project, don't know when I'll build it though.

All right, I need a second soldering iron as my old one is utter @#$! now. Shoot me options. Please don't get crazy on me as I dont' have much money to spare.

You need a Pace PRC2000 soldering & rework station. It's only $4500 or so these days, back when I was in the electronics assembly industry they cost something like $10,000. It'll be the last soldering iron you ever use!

Serious answer. I like Hakkos. They fit my hand the best and I haven't heard any complaints about them.

[Dusty Chalk]

All right, I need a second soldering iron as my old one is utter @#$! now. Shoot me options. Please don't get crazy on me as I dont' have much money to spare.

Audiokarma.org had some good suggestions, I'll go there later and find the specific thread that I'm thinking of, but if you want to go there and search under DIY, feel free. And they're constantly posting when the good ones go on sale.

[Nanoha]

Ah ha, thanks for the help Aerius and Dusty.

[Dusty Chalk]

Ah, shit, I forgot to look, it may be a couple days before I get on again. Feel free to remind me again.

[Nanoha]

Weeee, so many projects this summer. Beta 22 is also in my "to-build" list it seems.

[aerius]

what software is that, aerius?

It's called drawing by hand and scanning into Photoshop to adjust the contrast a bit. Oh, you mean the PSU. That's PSU Designer II

Do you know what a headphone would sound like if the impedances don't match up? (Output impedance too high for the low impedance cans?)

I want to see if it's the effect I'm hearing.

One of several things, or a combination thereof. Bass loses control and gets a bit loose. Dynamics get softened and lose impact. Tone shifts towards warm & fuzzy instead of neutral. Soundstage & imaging may not be as clean & precise.

Also, final changes to the schematic. Use a red or yellow low current low brightness LED to bias the ECC40. I'm currently using a yellow LED rated for 2.1V, 5mA, and 8mCd. I've thought about LED biasing the output stage as well, but if I'm using 6BX7's I'm going to end up with a string of something like a dozen LED's on each channel. That's a little much, even for me.

[cotdt]

that PSU designer tool looks useful. I'll eventually use it for my future tube amp, but it will be a while before I have the funds for it.

Impedance does not need to be matched. You just want the output impedance to be sufficiently low relative to the headphones. The output impedance being 10 times lower than the headphone's impedance is good, more would be even better. It's only when the output impedance is significantly higher than your headphone's impedance that you get the effects aerius mentioned. A good SS amp has an output impedance of 0.1 ohms, about 4-16 ohms for a transformer-coupled tube amp, and 30+ ohms for an OTL tube amp.

[Nanoha]

Hmm, so a question. What happens if you put a resistor in parallel or series with your headphone cable?

[cotdt]

Hmm, so a question. What happens if you put a resistor in parallel or series with your headphone cable?

you will lose output capability. well, adding a resistor in series is commonly done for the Koss KSC35. Since you increase output impedance, for some amps you'll get tighter bass and better clarity. But good amps should already drive them to their max potential without such tricks.

[Nanoha]

Aerius, do you have an idea as to the output impedance of your amplifier? Also, would it be possible to build that amplifier for say $350 USD max?

[aerius]

In theory the output impedance should be a bit over 4 Ohms with 6BL7's and about half that with 6BX7's, however I've never measured it so I have no idea what it actually is in real life. I do know that it'll drive 6 Ohm speakers with either tube so the output Z can't be too bad.

[aerius]

New and improved, now with the wonderful 37 tube as the input. A 37 for each channel, then to an ECC40 (a dual triode) which does the phase-splitting for both channels, then to the output tubes.

This is much better for Grados as it significantly lowers the gain and makes the volume knob a lot more useful. You can also use a 76 which has a tad more gain, but nowhere near as much as the ECC40. These tubes are still pretty cheap, $10 give or take $5 or so, but you might have to hunt around a bit. I think they sound better than the ECC40 as well, but I'll have to borrow a better source to be sure.

My recommendations would be as follows. For Grados and other low impedance 'phones, build it with 37's. For Senns & AKG's, go with the higher gain all ECC40 version.

[mirumu]

37s, nice! Very tempting to build one I must say. Nice use of the toroidal too. From the pic you posted on page 1 it looks like you've built it on stripboard? I'd always just dismissed the idea myself thinking it wasn't really that suitable for tube designs given the socket sizes but if that isn't a problem then it would otherwise make prototyping much easier.

[aerius]

Thanks! It is indeed built on a piece of perfboard/stripboard or whatever they call a board with a bunch of holes in it in your part of the world. I just drilled the holes for the sockets with some large wood bits on my powerdrill and glued the sockets in place, then put the supports & standoffs near the sockets so I don't break the board when putting in the tubes.

[Nanoha]

Okay, I've decided I want one of those. 37? ECC40? I want.

Aerius, I might start trying to contact you a lot in the coming weeks. Hopefully I won't be too much of a bother.

[aerius]

Also, would it be possible to build that amplifier for say $350 USD max?

A quick back of envelope calculation says it should be entirely possible. I built mine for under $100 CDN, but I bought a lot of the parts used from a local surplus store. The big expenses will be the transformers, the choke, and the motor-run capacitors, that might add up to maybe $200-250, everything else is literally a few dozen dollars. I used Cat6 network cable to do all the wiring since I had a ton of it lying around. I'll draw up a schematic for the DC tube filament supply after I get mine dialed in, it's running a little higher than I'd like with the 37's, I want it to run at 5.9V-6V instead of the nominal 6.3V

[cotdt]

is 37 same as VT-37? those are extremely rare and hard to find. why make a design for tubes that are so hard to obtain?

[aerius]

is 37 same as VT-37? those are extremely rare and hard to find. why make a design for tubes that are so hard to obtain?

It's so rare and expensive that it only costs $8.50. Vacuumtubes.net is selling them for about the same price, and I recently had a dozen shipped to me for under $80 from a guy on diyaudio.com. These are not rare tubes by any stretch.

[cotdt]

wow! interesting. i'll see if i can gather the parts to build the amp.

[cotdt]

A quick back of envelope calculation says it should be entirely possible. I built mine for under $100 CDN, but I bought a lot of the parts used from a local surplus store. The big expenses will be the transformers, the choke, and the motor-run capacitors, that might add up to maybe $200-250, everything else is literally a few dozen dollars. I used Cat6 network cable to do all the wiring since I had a ton of it lying around. I'll draw up a schematic for the DC tube filament supply after I get mine dialed in, it's running a little higher than I'd like with the 37's, I want it to run at 5.9V-6V instead of the nominal 6.3V

i've had a good experience using a big 6V battery powering 6.3V filaments. the background noise was nonexistant.

[cotdt]

aerius, i'll build it with what i have on hand: 240V B+, 4k primary SE transformer, should be fine?

[mirumu]

wow! interesting. i'll see if i can gather the parts to build the amp.

I'm seriously thinking I should do the same. I have some long term plans for building an amp to drive electrostatics in the future. I'm thinking something DHT based like some GM70s, 845s or 805s for the output stage (either OTL or with a low step down ratio output transformer) but haven't really looked into what tubes would pair well with these for earlier stages as of yet. I definitely do need practise building and understanding tube circuits though and this really looks like a nice amp, especially with the nice short signal path and low cost part selection . 37s and ECC40,etc are also much more interesting to me than the typical 6SN7 or 12AX7 designs too.