Looking for Bluetooth DAC with vacuum tube buffer/pre-amp

[VA1DER]

I am looking for a DAC (hopefully with good bluetooth) with a vacuum tube pre-amp or buffer. Many DACS slash pre-amps use tubes for a "coloration" stage. I am not looking for this - this to me is using a tube to poison the audio. Problem is the signal to noise ratio is so poor searching for what I want that I can't find what I'm looking for, so I'm hoping for some recommendations.

What I'm looking for is a | DAC chip -> I-V stage -> Tube buffer | topology. This on the assumption that most DACs will quite happily drive a line out relatively directly after I->V conversion. If there is a truly fine DAC that doesn't have enough output to drive a -10dBV (or 2V P-t-P) line out and needs a gain stage, I would also happily use a | DAC -> I-V -> unity solid-state-buffer -> tube gain stage | topology.

Best case scenario:
Bluetooth: Built-in LDAC support is nice to have, but external through SPDIF is fine
DAC: R-2R or other voltage output design that has a tube friendly high impedance output
Tube: 12AU7 family (e.g., JAN‑5654) for a cathode‑follower or low‑gain buffer.

 

[wparks]

I can only really comment from the tube perspective- I'm not really sure your desire for a tube in this role is the best choice- You don't want to "poison" the audio with the tube so you are not looking for coloration here, and you are concerned with signal to noise ratio. The problem is that introducing a tube into a small design usually involves switching filament and high voltage supplies, that will kill your signal to noise no matter how spiffy the filtering. Only a much larger chassis that has a full linear power supply (without switching) and choke filtering to minimize ripple can give you the signal to noise you are looking for- then the tube buffer itself is far from "low impedance", and will have difficulty driving some modern equipment inputs that are designed for most modern sources that have very low output impedance. Having any gain in the tube stage while having an acceptable output impedance will require a transformer.

Seems to me the only easy way to get some gain with the accuracy with low noise you need would be solid state.

 

[VA1DER]

I'm not sure I agree the power supply to be a real issue. For this purpose a large transformer isn't needed. In fact, strictly speaking, for B+ a transformer isn't needed at all, even in North America (though I'd still use an isolation transformer). For most tubes a 150V plate voltage would be sufficient and you really only need to supply quiescent current. For a 12AU7 family, as an example, that's about 3ma total per triode. I can get that just with a small isolation transformer and some good filtering.

There are some commercial offerings. The most noteworthy I've found so far is the FiiO Warmer R2R. The stock tubes it uses, though, seem to be a bit of an impediment and I haven't seen any reviews with anyone who has upgraded it. Also the internal tubes sitting right by an R-2R network has to be awful. I think using .1% resistors to make a 24 bit R-2R is simply not credible anyway - even if they are calibrating it I don't see them getting more than 16 bits of actual precision. And that is likely being generous. With four little space heaters sitting right beside them, well, even cermet resistors will drift with that kind of heat. The reviews on this product seem to focus on the stock tubes as being its limiting factor. But as I look it over I suspect heat as a real killer here. They chose its name better than they knew.

Anyway, if anyone has any experience or even just knowledge of other offerings in this vein, I'd love to hear it.

Seems to me the only easy way to get some gain with the accuracy with low noise you need would be solid state.

You may be right. I've brainstormed some designs, and even considered making something. My motivation is this. I'm not looking for tube "coloration" per se, that is I don't want to artificially add tube "sound". The incidental attributes of a tube lends itself to the serendipity that where it departs from the "ideal" of perfect noise-free linearity, it does so in a way that is more pleasant. I have been willing to accept more total non-linearity because of this. My 1964 Philips Capella B7X43A is the best sounding piece of gear I own, beating out even my beloved STA-77.

I draw the line, however, at using a tube to intentionally distort the sound. But I think as a buffer or as very low gain stage, it could have a purpose. It would be operating in its strongest linear point, would be (with good power as you say) low noise, and I think makes enough sense that I'm willing to try it. But yes, despite my rough napkin diagrams, it may not be as practical as I think.

 

[Isotonic]

The moment the music enters into an electronic state, the music is "Poisoned". It will never sound the same. Pick your poison. Some like Vodka, some like Whiskey...

 

[FlaCharlie]

The incidental attributes of a tube lends itself to the serendipity that where it departs from the "ideal" of perfect noise-free linearity, it does so in a way that is more pleasant. I have been willing to accept more total non-linearity because of this.

Based on your incredibly eloquent description (!) it seems that you do actually want the tube to influence the sound to some degree. That is, you want it to sound different (subjectively better) with the tube in the signal path. It seems to me that your concern is just that the tube’s influence on the sound is not extreme.

What type of amp do you plan to use this with? What’s the input impedance?

I’ve built a couple of simple, low gain, tube preamps. If you’re up for a DIY project they might work for you. They can be used with either SS / Class D or tube amps.

Thread 'The Flow Pre4D - A DIY Tube Preamp'

Apr 15, 2025

I’m always encouraging people to build something themselves that uses tubes. My most recent amp threads feature simple designs that don’t even require the use of solder. And they use a SMPS (switch mode power supply), which also simplifies the DIY process.

Here I’m presenting a similar DIY project for those who are interested in tube preamps.

The popularity of tube preamps has been steadily rising for years now. Perhaps the best proof of this on AK is the massive thread over on the New Gear - Values forum about the little Chinese FX Tube 01 preamp and its various spin-off...

• FlaCharlie
• Replies: 45
• Forum: Tube Audio

• 

1626 Preamp Build

Thought I'd post about my recent preamp build using the 1626. First, some background . . . I've been breadboarding a basic preamp circuit with a variety of tubes. This has gone on, somewhat sporadically, for the past couple of years. In all I think I tried at least 8 different tube types...

www.audiokarma.org

FWIW, I recently picked up a Fiio BR15 R2R DAC which is essentially the same as a Fiio Warmer but without the tube section. Since I already have a couple of tube preamps this seemed like the best option, for me. The combo works well and, of course, since the preamp is a separate, stand-alone piece of gear there is always the option of pulling the tubes out of the signal path if I want. The Warmer doesn’t appeal to me at all, for several reasons.

 

[VA1DER]

Based on your incredibly eloquent description (!) it seems that you do actually want the tube to influence the sound to some degree. That is, you want it to sound different (subjectively better) with the tube in the signal path. It seems to me that your concern is just that the tube’s influence on the sound is not extreme.

Ultimately, I want my DAC to be 100% transparent. There just aren't any that are. Delta-sigma chips are really good some of them, but still tend to be a little harsh and/or their I->V op-amp stage afterwards is. So if they are a finger on the scale in one direction, then using a tube for the required gain (if any is needed) is a way to offset that. It's a fine distinction, I suppose. I don't want "tube sound". I want perfect sound. Tubes just happen to be one good tool to get that.

FWIW, I recently picked up a Fiio BR15 R2R DAC

It's funny that you bring that up. I was actually -><- that close to pulling the trigger on a FiiO K13 R2R. Resistor ladders are themselves a bit like tubes, in that the kind of non-linearity you see from resistor ratio error tends to be even order. So, again, where they depart from perfection they naturally do it in a way that is more pleasant sounding. That could be why the reviews found the FiiO Warmer to be too much. R2R + Tube = too much. When I thought about it, it seemed that an R2R without the tube might be just what I was looking for. I came really close to ordering one.

In the end, though, I found something close to what I was originally looking for. In fact I just picked up a couple Aiyima T8's. I don't know how it got on my radar, I've been looking at so many things. But I do remember when I first looked at it I completely dismissed it as "tube as an affectation budget crap". But a Russian review made me take a serious look at it (and my snobbiness). The device is marketed wrong, as a pre-amp instead of a DAC, and priced awkwardly - in a zone that's too high for bulk consumer, and too low for most audiophiles to consider. But the closer I looked at it, the better it seemed. The signal path runs from BT/USB/SPDIF -> ES9018K2M DAC -> OPA1656 I-V -> 6N3 tube for mild gain -> RCA Out. The tube is used appropriately, for actual gain. It's still burning in, but as I write this it's playing through an STA-77 / Minimus 7Ws and sounding outstanding doing it. Noise is a little more than 105dB down. THD is poor by today's standards, at about .2%. But Realistically (hah) no more than what my beloved STA-77 has. And a lot of that is likely the crappy Chinese tube that came with it. I look forward to measuring THD once I get a Northern Electric (Go Canada!) 396a in it.

What type of amp do you plan to use this with?

I am a bit of a Realistic buff, and gravitate to some of their underdogs. The STA-77 I'm listening to right now is, I think, a work of art. I also have a QTA-753 (Foster), QTA-790 (Kenwood), and JVC RX-7000V as my AVR. All of them are going to get some DAC love at some point. My STA-77 and QTA-753 are, right now, my main listeners.

What’s the input impedance?

On the STA-77 it's 100k. I honestly don't know the others. It's been too long since I've seen them on the inside.

I’ve built a couple of simple, low gain, tube preamps. If you’re up for a DIY project...

Indeed! The T8 has quenched my immediate thirst, but I'd love to sink my teeth into something like that.

 

[FlaCharlie]

Indeed! The T8 has quenched my immediate thirst, but I'd love to sink my teeth into something like that.

Glad you found something that works for you in your system.

I think you might enjoy DIY. It allows you a lot of freedom. To me the most important aspect of the DIY route is the process of breadboarding which lets you listen to different tubes and operating points and choose what sounds best to you. Between the two preamps linked above and several amp builds I’ve been able to listen to dozens of types of preamp tubes and output tubes.

Large commercial manufacturers can only use a limited number of tube types due to supply constraints. General cost restraints and marketing strategy considerations become the driver of their design decisions. This results in circuits that are (both objectively and subjectively) compromised not optimized.

For example, they usually operate the tubes at voltages that are considerably lower than the tube data sheet recommends. It’s almost impossible to obtain a schematic that documents the operating point of the tube so, if you really want to know how linear they are operating the tube itself, you have to measure them and figure it out yourself.

Rather than concern yourself with measurements I would suggest that you just listen and enjoy. If it tastes good just enjoy the sausage, don’t ask about the ingredients. If you want something that's "seasoned to taste", that's what DIY is all about.

Tube: 12AU7 family (e.g., JAN‑5654)

The 12AU7 and the 5654 are not related in any way. The 12AU7 is a 9 pin dual triode while the 5654 is a 7 pin pentode that is used in a lot of Chinese gear where it is wired in triode. The 12AU7 is widely reviled by tech types as being very non-linear when used as a preamp tube.

And, just so you aren’t confused, just because tubes have some characteristics in common, like the tubes in the 12AU7 “family”, that doesn’t mean they are interchangeable. Circuit changes are necessary if you're at all concerned about optimization / linearity.

And, speaking of incompatibility . . . . a word of caution. Aiyima lists a variety of tube types that they claim can be used as substitutes for the stock 6n3. That list is not accurate. A quick look at the data sheets shows that two of them cannot be used.

One tube they list as a GE5657. There doesn’t seem to be a tube with that designation. Perhaps the “GE” prefix refers to General Electric?? The 5657 exists but it’s not even an audio tube, it’s a Magnetron that’s used in microwave devices and radar. It doesn’t even look anything like a normal tube.

They also list the 12A4B on their website. There is a 12A4 but nothing with a B suffix that I’ve been able to find. Perhaps it’s a misprint or maybe the person at Aiyima who wrote that is dyslexic? I say that because, on Amazon. they list 12B4A, which is a legitimate tube number. Whichever one they mean, neither of them will work since they are both single triodes. If you plugged one in the tube would blow up and it’s possible that the T8 might also be damaged.

It doesn’t exactly inspire confidence when the manufacturer gives such totally inaccurate information.

I wonder if any of the “reviewers” have mentioned this? I doubt it. Most of them are just shills who are paid with free gear to promote stuff.

My 1964 Philips Capella B7X43A is the best sounding piece of gear I own, beating out even my beloved STA-77.

I looked up your radio and found a schematic. Like many European radios the amplifier stage is a SE (single-ended / one tube per channel) using EL84 / 6BQ5 tubes. Essentially the same as the little Magnavox (and other brands) console stereo amps that are so popular with DIYers. So it's not surprising that it sounds good.

It looks like it may have auxiliary inputs. If so, you might want to pick up a cheap Bluetooth receiver (probably $10-15) and stream to it.

 

[gadget73]

sounds like something like a Bottlehead Foreplay with a decent bluetooth receiver added to it would get it done. I was actually contemplating this just last night. I have a Foreplay that serves as my shop preamp, and I have a decent BT receiver module that can also operate as a DAC when connected to a PC via the USB jack. I'm considering upgrading the transformer on the Foreplay anyway, and swapping to something that I can ge the 5vdc needed for the BT module would be easy. Mount the module inside, use a short USB-C extension that has a chassis-mount connector, and now I've got 3 analog inputs, and one digital that can operate as either a DAC or a BT receiver. The selector on a Foreplay is 5 positions, but its only got 3 sets of jacks. Absolutely trivial to make use of one of those un-wired positions.

The 6v winding on the Foreplay is good for 600mA, which is what a pair of 12AU7 consume. I'd like a little more so I can run 6CG7 tubes in it, or at least be able to rectify the supply since it has a very small amount of residual hum that i can't quite eliminate.

 

[DeadTrader]

Saw this in the tube forum Schiit Audio: Audio Products Made in USA a true tube buffer for what I'd call cheap. Put it in your current chain, might be what you need.

 

[VA1DER]

Put it in your current chain, might be what you need.

Thank-you for the link. While I don't want just an added "neutral" tube in my signal chain, maybe the hair I'm spitting is too fine to matter.

I wonder if any of the “reviewers” have mentioned this? I doubt it. Most of them are just shills who are paid with free gear to promote stuff.

I was a little confused by the list myself and spent some time scratching my head over the GE5657 myself. I haven't seen mention of the mistakes in any reviews. The best review I've seen is a Russian one (translation), it's seems pretty honest, and has some measurements to back it up, but doesn't speak toward other tubes other than to mention he believes a better tube would reduce the THD.

Most reviewers I've seen that were clearly "paid" to look at it actually seemed to have been a little confused by the device. The T8 is sold and marketed as a pre-amp and more than one review just hooked up analog source to its analog inputs. The manufacturer's clear (English) language difficulties did not help them market this, and it was really awkwardly positioned.

But I've been running this device for the better part of two months now, and I have to say I adore it. Call that selection bias, but it actually fits the bill quite nicely. It sounds outstanding.

I have done some experimenting with a few tubes now. What I have, in order of (perceived) transparency in my T8 are:

• Northern Electric (Canadian subsidiary of Western Electric) 396A, 1970
• GE JAN 5670W, 1984
• JJ E88CC (with an adapter I made)
• Stock Chinese 6N3

The first two are outstanding, with the last two being good and ok. The JJ E88CC I only bought because I needed a tube to make up minimum free-shipping cost for the 396A, but it was a fun little experiment and surprisingly is better than the Chinese.

I'll do some THD measurements when I can. My two Canadian 396A's just arrived today - ironically having to journey from Doug's Tubes in the States to get back home. But now that they've arrived I can get to work on gathering some hard data.

But going by my ole Mark 1 earball, I honestly can't say enough good things about this little DAC. The only real drawback to me was its lack of LDAC on BT, but I made that up with an Akliam DD2 LDAC->TOSLINK receiver. I found my T8s at Walmart's online store here in Canada for $40CDN. The device has just disappeared off Aiyima's web site, so I expect I picked up mine as a dump of the rest of their stock. At that price I bought three. One, a spare, and one to experiment with.