• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

To build or to buy

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I hate to say it, but it looks like I need some serious multi-variable help from the community.

I recently built out a new set of speakers (that I'm thrilled about) - and upgraded my phono situation (now a Rega P8 with a Hana ML with my current PC3 and awaiting some PCB's for an Emerald build.) I'm just going in deep and currently looking to charge the Pre/Amp setup in my system. Its great now, but I'm frankly researching to build "my" end-game pre/amp setup - the source components are sounding great - the speakers, killer, just need to fill in the gap in performance.

For starters, the speakers are relatively inefficient (more on that later) clocking in what I estimate to be about 84-86/db. @ 6ohms which makes them needy on paper but I'd be curious to better understand how these two are interrelated. I live in a small space, in what normal listening amounts to essentially "near-field) listening at relatively reasonable if not low SPLs.

I was first intrigued by building a Class A amp to power the speakers, I liked the idea of linearity and the Aleph J would represent a piece of audio history to participate in. After further consideration, the "stock" build doesn't offer too much in the way of wattage (not a con, per se) and the cost of the parts puts it into the realm of considering Schiit's Aegir with a DIY Pre, an existing pre or even a purchased one. Those first 10 watts are pure Class A, with the remainder in some proprietary "class a-like" performance scenario. To put it bluntly, the cost of an Aegir is about on par with the cost of the parts of an Aleph J. with all due respect to the Aleph amps, it's hard not to begin to consider that option,

Further down the rabbit hole, I recall my desire and passion for a tube amp - not usually reserved for lower efficiency systems - but still, a single-ended setup could offer that renowned "holographic" soundstage and maybe even a better performer at the now-obvious low volumes we can enjoy. So, plugged in a Maggie 9300 series restoration i did a few years ago (still the original tubes) which is meant to put out something like nearly 10W @ 8 ohms - ran this with a Sansui 555A from the pre-outputs - and it drove the speakers "just fine" from the 2 of 10 position on the volume pot. Was it perfect? Not at all, I think the low end suffered and the top seemed a touch sharp/grainy. But then again it was the Maggie with OG tubes. More to my point, is that those 10 watts gave me more than I wanted from an SPL perspective, but since I know that SPL isnt all there is to what is asked of an amp, then we continue.

For perspective, on these speakers, I am running a Sansui 555A, in full integrated mode - volume at 2 and it's pretty powerful, maybe could go to 3 if I wanted to rock it. Sansui AU 666, even more so, and the AU5900 - even more so. Yeah, I've got a bunch of Sansui's kicking around.

All of that being said - I've got a few options I'm considering (and I'd like to better understand how power really functions to better judge if I can run one of these tube systems as my main amp.)

1. Just pickup a Schiitt Freya+ with an Aegir and call it a game.

2. an Aleph J build and either a Schiit Freya+ or another DIY pre - maybe something from Elekit

3. Elekit 8600S or 8800 with Lundahl. Awfully tempting, if surprisingly ridden with hiccups for some builders - but based on the reading these are some hard-to-beat amps, for what seems like a reasonable price plus some elbow grease.

4. Prebuilt Tube systems - something like Decware's Zen TorIII Mark IV. This is of course high powered and definitely seems to fit the bill. I'd likely still like a pre-amp so

5. Pre-owned market - definitely would require some patience and shopping around but I get the sense that I could find something that punches well that someone would be passing down/along. Some such examples that came up in my brief query into the matter are things like - Cairn 4808, Musical Fidelity's M6S, Conrad Johnson LP70, and I'm sure there are plenty others I haven't considered.

Of course, the problem with any of this is the inability to test/trial them without committing to buying and/or reselling (yes Schiit, as well as Decware, have trial periods, but I'd rather hone in on something before spending a small wedge of the budget on shipping things to and fro.)

I guess amongst the things that I'm most curious to learn is how to better understand demand in the context of efficiency and amp/speaker synergy. It's often discussed that certain pairings work, or don't, but is there a way of simply and intelligently getting a better sense of the needs? Even some testing equipment that I could acquire to better understand the draw etc.

As you can see I'm a bit deep in the hole here, and surely some wisdom will prevail from the community.

Oh, and there's an obvious question of the budget as subtext - I'd say I was originally int he $2k neighborhood - for total spend at the moment - but after reading a bit I'd be willing to go as high as the Decware price point (3500+) and save a bit for a preamp further down. IN any case, there's a lead time on many of these choices.

Help!
 
You've answered your own questions.

For the budget you mentioned, go for the Decware Torii. You might be able to find one used. Send it back and pay the shipping and fee and you'll have a lifetime warranty.

Anybody selling used gear should let you listen in person if they are local and there is nothing wrong with the gear. That's been my experience. People selling on DIY Audio usually back their sales by offering to fix any problems, but you usually pay shipping. Also, I feel sorry for all the DIY folks who don't live near a shop that will let you return something you buy for store credit or a refund minus restocking fee. Join/form a club. Take a electronics class at your local community college. Attend a convention when covid is a memory.

If you live in a large town, be nice to your neighbors. After work one evening, I was walking home from the bus stop and saw tubes glowing from the retired neighbors window. I asked him about his tube amps and got to be good friends with him. A few years later, he gave me over 1000 tubes and two tube testers. That's how I fell into the vacuum tube rabbit hole, but it's been a fun adventure.
 
I would say step #1 is to quantify your average "just fine" listening level in measured SPL. To do this, you'd need an SPL meter, which are easily had for not a lot of cash.

When you get this number, then you can figure out your speakers efficiency, i.e. how many watts it takes to achieve your "just fine" listening level. That would take more measuring equipment, if you dont have this stuff already.

Then knowing the Watts it takes to power your speakers most of the time, you can decide on something called headroom. This is the excess and undistorted power the amplifier has beyond what you'd run it at for your ordinary listening level.

Some say 10X; normal listening level takes 10W, need 100W amplifier. I'm sure there are amplifiers (AB) that run class A for 10W, then jump to class B for those occasional musical transients that might clip a 10W amp, like your "Maggie".

That's one way to whittle it down to be able to shop for something.
 
Last edited:
If you're willing to go in an unconventional route, I have a modular amp design that might fit the bill.

I made PCBs for line, buffer, VA/PI, Driver, VA/PI/Driver in one, 4 different output boards that work with over 30 tube types, RIAA, 6E2 meter etc.

My idea is to make it easy to customize for a certain purpose.

Using a pair of sweep tubes in triode mode per channel, you can make between 15W to 80W depending on the tubes and OPTs - all using an easy power supply (made boards for that, too).

It uses off the shelf toroidal power transformers (2 per channel) as the OPTs and therefore the entire build shouldn't cost more than about 1500 US, less if you try. I made the 15W version for my dad. Total cost was about 250$. I also built a version in a cake pan for myself as an experiment for 100$. Very cheap by tube standards but the performance is great.

FWIW - I have a pair of 83db speakers that I can bottom out with the 80W version.

As jjasniew said, figure out how much sound you need, but in my case it's variable.

Sometimes, my 7 watt single ended MOSFET amp is plenty loud, other times my 300W Yorkville isn't loud enough. It has been known to be linked to blood alcohol level in my case though :D

EDIT: Added a pic of some of the boards in a build I'm working on. The three most central boards are not my design.
 

Attachments

  • 2021-01-24 12-31-40 (1).jpg
    2021-01-24 12-31-40 (1).jpg
    1.1 MB · Views: 214
Last edited:
jjasniew - I think you're right - I'll do some measuring this evening and try to get some numbers on paper for what the needs are and work with some established grounds.

Kodabmx - very interesting indeed - what, if any, would you say are the compromises in your design? Do you ultimately retain a SET topology or some sort of hybrid?

Definitely a tempting idea hmm
 
The topology is as in this schematic. It's push pull. The values are from an older revision, and I don't recommend this exact recipe anymore, but the topology is sound. R7 and R8 should be 33k, R6 should be 240. Also, I bias the 6P36S in this schematic at 80mA instead of 60mA now. The 10R resistors are specced to be 5W incase of a shorted tube. A bad tube shouldn't take out the rest of the circuitry IMHO.

By changing the output stage to suite, the front end can be used for many designs :)
 

Attachments

  • 6F12P-6N8S-6P36S-Black-edition. (1).png
    6F12P-6N8S-6P36S-Black-edition. (1).png
    66.2 KB · Views: 186
Last edited:
Brief update -

I did some a few measurements - and it seems like the highest demand "uncomfortably" loud signal I could push through that was most demanding was 125hz which drew a continuous 12W. As expected, other high frequencies, though painful, drew much less power.

In a practical listening, I picked a bass heavy electronic track with a very consistent bass-line. In dynamic draw I wasnt able to simulate anything higher than 9W at those uncomfortable values - and if we throw in some margin for error I think that 12W is fair.

So, i've heard/read different ideas here - 3db of headroom puts me somewhere like 24W on the shopping cart - and I am going to call that 24W conitnuous or RMS for the sake of clarity. If we do 10X - then we'd say something like 90-120W RMS? or is that 10X meant to be RMS to Peak calc?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.