[tortugaranger]

For quite some time now (I think 2+ years qualifies) we've been modeling, prototyping and modifying various buffer designs. In fact one of my key goals for 2016 is the release of a buffer product. Time is indeed running short to accomplishment that.

Sooner or later you have to drive a stake in the ground, stop noodling around endlessly, commit to a design and produce it. That time has come for our buffer.

One of the key decisions that's been made is whether to go solid state (JFET) or tube.  TUBE it is! In fact the 6CG7. For the tubular among you the 6CG7 is the smaller 9 pin version of the venerable 6SN7 octal tube.

A key design criteria for our tube buffer is we wanted to fit the whole thing, including the power supplies, inside one of our existing 3x6x8 inch extruded aluminum cases. That meant no big black honk'n transformer riding on top of the enclosure either. Everything inside. Except the tubes themselves of course. A pair of 6CG7 tubes will be standing up through the case. Figuring out how to accomplish that took a fair bit of time. We are using a pos/neg voltage doubler to produce the main split voltage power for the tubes along with 3 separate small low-profile semi-toroidal transformers mounted on a board along with the rest of the power supply circuitry. There will be 2 separate boards, the buffer board and a power supply board.

Another key design criteria was the elimination of soul-sucking coupling caps from both the input and the output the buffer. To accomplish this the tubes are powered with a split voltage  +/-125 VDC  supply rather than a single sided supply. Configured as cathode followers each channel has a pair of triodes connected top to bottom with the output coming out between them at essentially 0 volts. A servo circuit is included to ensure the elimination of any remaining DC offset. In the unlikely event that the DC offset gets our of bounds there's a DC offset detection and safety isolation circuit that will disconnect the buffer output. This circuit also avoids any bumps or thumps during power up and will only connect the buffer outputs once the DC offset on the output is minimal and stable.

Since eliminating coupling caps required a DC offset eliminating servo we ended up with a high quality op amp  input stage using the LM49720 op amps. Hence the reference to this being a "hybrid" tube buffer. I love the LM48720's having used them before in several designs including as upgrades in a DAC I've been using. The input stage takes direct DC input and is nominally unity gain although the design does provide for switching out a couple of removable plug-in resistors if you want to boost the gain for any reason. All of this further complicated the power supply since now we needed a split +/- 15 V supply for the op amps. It's all in there!

Barring any surprises (those never happen...not!) I expect to have a near-final prototype running later this month (October) and final-final prototype in November for release by year end. I'll post some pics as soon as we have something presentable but imagine one of our LDR3.V2 preamps with a pair of 6CG7 tubes sticking out the top and you'll get the visual gist. 

To be clear, this hybrid tube buffer will NOT fit inside next to our LDR preamp board in our existing preamps. The tube buffer requires additional real estate. I do envision a possible combined buffered preamp in an single package but it will be twice the width of our existing V2 preamps which means ~ 3" tall by 12" wide by 8" deep.

Watch this space! 

[mute]

This is so nice!
Are you planning a balanced version of the buffer that can be used with LDR1B.V2?

[rajacat]

Its too bad that there aren't more current production 6cg7s. The EH 6cg7 doesn't have a very good rep SQ wise. The NOS market is thin and the prices are high especially for the ones with the inner shield between the plates. On the other hand, if you find a good NOS pair, they should last a long time.

[tortugaranger]

This is so nice!
Are you planning a balanced version of the buffer that can be used with LDR1B.V2?

Since a true balanced buffer requires an additional 2 channel tube board there's no way we can fit 2 of the existing 2-channel tube boards in a single enclosure.  Even if we could there's also the issue of doubling the heater current. We have 2 amp heater capacity and only use roughly 1.2 amp but doubling it would push it past the 2 amp limit.

I suspect that it comes down to a choice of either: 1) Use 2  separate buffer units for balanced @ twice the price, or; 2) Design a balanced version of both the  buffer power supply and the buffer board that still might fit in the single enclosure.

For now we're going to defer the decision on whether to go with option 2 until after the buffer has been released and has accumulated some commercial history.

[tortugaranger]

Its too bad that there aren't more current production 6cg7s. The EH 6cg7 doesn't have a very good rep SQ wise. The NOS market is thin and the prices are high especially for the ones with the inner shield between the plates. On the other hand, if you find a good NOS pair, they should last a long time.

Given the buffer topology we're using we may be able to accommodate a 6H30 as an alternative. Have not gotten into the details on that but intend to assess the possibility.

[33na3rd]

Its too bad that there aren't more current production 6cg7s. The EH 6cg7 doesn't have a very good rep SQ wise. The NOS market is thin and the prices are high especially for the ones with the inner shield between the plates. On the other hand, if you find a good NOS pair, they should last a long time.

I think the current EH 6CG7 can sound very good in modern equipment that was designed around it. I agree that it can sound a bit thin in equipment designed around NOS tubes.

[tortugaranger]

Looking at a very interesting variation using a single 6H30 tube instead of a pair of 6CG7's per stereo channel. If this holds up under testing/listening we'll be able to fit a balanced pre (2 x 6H30's) in the existing 3x6x8 inch enclosure. Another bright shiny object to slow things down.     But ultimately the results are what matter. 

[firedog]

Like the 6H30, and lots of variants are available.

[tortugaranger]

Up until recently we were pursuing a hybrid tube buffer with an op-amp input stage (with optional gain) followed by a 6CG7 cathode follower output stage using a split voltage power supply. Now we're evaluating a possible alternative approach except the input stage uses a 6H30 tube and the output stage is a proprietary solid state buffer that also uses a split voltage power supply albeit at a far lower voltage. The 6H30 input stage would have nominal unity gain with the ability to add 3-6 dB of gain if needed.  Both designs have a servo to keep the output DC offset to zero so no output coupling cap is needed. Excited about finishing the evaluation and getting the best buffer version out the door! Have high hopes for the 6H30 design approach which may yield extraordinary performance with a less complicated and costly power supply.  And will only need a single tube for single ended we should be able to fit a balanced version into the same enclosure using a pair of 6H30s.

 

[tortugaranger]

We've made considerable progress recently on our hybrid tube buffer design which is now in its 2nd stage prototype design. I'm hopeful only one more design iteration will be needed before final production sign-off but that's admittedly optimistic.

We made some key design tradeoff decisions last week. After considerable back and forth we opted for the hybrid design shown below rather than the cathode follower variant we'd been pursuing. This resulted in swapping around the tube and solid state ends of the hybrid such that we now have a tube input stage (with some optional gain) followed by a solid state output stage. We reluctantly gave up on eliminating the output coupling cap due to the zero DC offset design getting way too complex. This design also eliminated a very bulky and oddball split voltage power supply needed by the cathode follower. The result is an elegantly simple hybrid tube buffer that sounds  absolutely fantastic while keeping the parts count and cost down and fitting within our existing enclosure design envelope.

This buffer sounds wonderful when combined with our LDR attenuator on the front end. I believe we've been able to achieve our primary goal of enhancing the already great sound of our LDR passive preamp while above all else doing no harm.

To that end we've opted to go with a Belleson super-regulator for our main DC power in our base offering rather than as an upgrade option. However, there will be a few other upgrade option paths in the final base design.

Once our 2nd prototype has been fully vetted later this month I'll be able to determine pricing and come up with a release date. As much as I'd like to release the buffer before year end I'm thinking 1st or 2nd week of January is more likely but that's still very soft.

The buffer will initially be sold as a stand alone unit that can be used with any existing Tortuga passive preamps or even other passive attenuators.

How does it sound?  Adding the buffer to the output of our existing LDR passive preamp does change the sound somewhat. This isn't surprising. Most tubes impart some coloration. However, in the case of this buffer it's subtle instead of being overwhelming or overly warm. Added euphonics are there typical of the 2nd order harmonics present in all tubes. Perhaps most noticeable is how the sound is a bit bolder and punchier.  All while preserving the amazing clarity, detail and unveiled natural sound typical of our LDR preamps. Same yet different. Better? I could get used to this sound!! 

[33na3rd]

Looks good! I like the option of adding just a little gain. This really opens up the choice of power amp nicely.

[sruffle]

Very interesting.  I have thought of trying a Decware tube buffer between my Tortuga and Decware SET amp.  The act of adding another power cord and set of ICs seems like it might take away some detail.  This definitely seems worth trying though if it adds a little more punch to the sound.

[33na3rd]

To that end we've opted to go with a Belleson super-regulator for our main DC power in our base offering rather than as an upgrade option.

Morten,

Would the buffer be able to provide power to the LDR unit, and would there be any added benefit?

[tortugaranger]

Morten,

Would the buffer be able to provide power to the LDR unit, and would there be any added benefit?

The buffer will have a 12V regulator and power takeoff that can be used to power an existing Tortuga preamp. This regulator can be optionally upgraded to a Belleson regulator as well which can benefit existing preamps. Users of our PS12B battery power supply have reported improved performance so upgraded power from the buffer should provide similar benefits.


The new V3 preamps (when finally released)  will also be able to be powered via the buffer although due to the V3 having a different power arrangement a Belleson upgrade will need to be done at the V3 controller board instead of the buffers 12V supply.

[rajacat]

Will the buffer be available in kit form that can be retrofitted to existing kit built  Tortuga pre?

[tortugaranger]

Will the buffer be available in kit form that can be retrofitted to existing kit built  Tortuga pre?

The idea of offering the buffer in kit form has really not been considered. My focus has been on design and performance and more broadly on how the buffer will fit into our current and future preamp products. What I can say for sure at this point is there will be no kit version initially. No doubt we will revisit thst down the road.

[rajacat]

The idea of offering the buffer in kit form has really not been considered. My focus has been on design and performance and more broadly on how the buffer will fit into our current and future preamp products. What I can say for sure at this point is there will be no kit version initially. No doubt we will revisit thst down the road.

I expected you'd say that. I certainly understand. Oh well... I tried. 
 Actually, I've been getting by very well without a buffer since the Tortuga really doesn't need one unless you have unusually low output voltage on the source. My source voltage is marginal but adequate. I expect to be getting a new DAC soon. I'll definitely make sure that it'll have more voltage. It would be nice to experiment though.

[tortugaranger]

I expected you'd say that. I certainly understand. Oh well... I tried. 
 Actually, I've been getting by very well without a buffer since the Tortuga really doesn't need one unless you have unusually low output voltage on the source. My source voltage is marginal but adequate. I expect to be getting a new DAC soon. I'll definitely make sure that it'll have more voltage. It would be nice to experiment though.

Adding a tube buffer downstream of the LDR attenuator adds some "tube flavor" and perhaps some more punch but in switching back and forth between pure passive and buffered I think it comes down to personal preference in terms of which one is better or more pleasing to listen to. I'll have more to say on that subject once we have our latest prototype up and running. Design details do matter so I'm reserving my final opinions until the final design is ....well...er....final. Very soon I hope. 

[jtsnead]

Now your talking, I was interested in your LDR designs but really liked the sound
of the 6H30 (Thanks Gary Dood for your buffer), so I wanted to upgrade to a line
stage with this tube and also balanced since I am right under a radio station, so I
got a Audio Research LS17SE and I am loving it, have been looking at upgrading
to the LS27 since it has adjustable gain, the gain I have now is a little to high with
my class d amps, but it looks like I might have to wait for this!

[33na3rd]

Any "ballpark" figures for input/output impedance?