Modded SB output level

Wayne1 · « **on:** 30 Dec 2006, 05:00 pm »

There has been some concern expressed about the output level of the modded SB. Part of the mod I preform is to remove most of the analog section of the SB and connect the analog output of the the DAC chip to the output jack through very high quality capacitors.

Without the op-amp in circuit, the voltage level is reduced to the 1.1 V max level of the DAC chip. The phase is also inverted in Firmware versions after 15.

With some systems, the reduced output level is a benefit. You can run the SB direct into your amp without any fear of overloading the input sections. There is no need to use external attenuators.

Quite a few systems use amplifiers that can reach full rated output with 1 volt or less driving them. Others may need more voltage to reach a desired listening level.

There has been a discussion about this on the Lab Circle here on AC. There have been a lot of good suggestions in that thread. http://www.audiocircle.com/index.php?topic=35101.0

I have been aware of this "problem" for some time. One of the local DAM members does like to play his music at levels a bit higher than his modded SB can hit at max output. mgalusha put together a 1:2 step-up transformer for him that satisfied his needs.

I finally have received a pair of very high quality 1:2 transformers and will be putting them in a case with input and output wiring for evaluation purposes. The step-up transformer offers a solution to a couple of "issues" of the modded SB while adding some more.

The first thing it solves is the voltage "issue". The max output is now 2.2 V. This should be enough to drive most amplifier to very loud levels. It also can solve the phase inversion "issue". The output of the transformer can be wired 180 degrees out of phase from the input.

The transformer also isolates the signal and ground of the SB from the amplifier. There has been some comment that the ground of the SB "might" be carrying some of the internal switching supply noise into the next stage.

The one "issue" that using a transformer alone doesn't solve is the amount of current and sensitivity to capacitance of the cables. If a transformer alone is used, the output cable should be very short and of a low capacitance design. Optimum would be no longer than .75 meter. Shorter would be even better.

A solution to this would be to use an active buffer circuit after the transformer. The buffer would not increase the gain but it would lower the output impedance. There would be quite a bit of current available to drive most cable designs up to 10 feet long. Multiple outputs would also be possible.

I am looking at various options to this idea of buffers and transformers. I have placed an order for a Burson Buffer to see how it performs.

I will be trying to make some time, in the next few weeks, to work on a prototype tube buffer circuit. My initial thoughts are to use an octal base, dual triode for the buffer tube. I am thinking of the 12SN7. I hope to be able to have the filament circuit also be able to support the 6SN7 family, for those who like to tube roll.

At some point in the new year, I hope to be able to compare the pure transformer step-up to the transformer and tube buffer to the Burson Buffer. Depending on the outcome of the comparison, BOLDER Cable may have a new add-on product to offer to compliment the modded SB.

Vinnie R. · « **Reply #1 on:** 30 Dec 2006, 05:16 pm »

Hi Wayne,

I understand that you want to try the 1:2 transformer to step up the voltage (which also steps down current output) and a buffer to lower the output impedance after the transformer. Why not use an opamp (or even some custom designed discrete transistor circuit) to do both? I remember running to the same concern with some customers a while back when I was modding SBs. Some needed more output voltage than the 1.1V max level of the DAC in order to get their power amps loud enough. This was especially true with those that had less efficient speakers and/or were playing music that was recorded at low levels.

There are a lot of opamps out there that sound much better than the stock one.
You can tweak the opamp output to a voltage level that you want (2.2V, 3V, etc) based on the Ri and Rf resistors AND the output impedance is low enough to drive long cable runs, amps with lower input impedance, etc.

Of course I did love the minimalist approach of just taking the voltage output off of the dac

The Tubed-SB does sound cool!

Have a Happy New Year!

Vinnie

Wayne1 · « **Reply #2 on:** 30 Dec 2006, 05:47 pm »

Quote from: Vinnie R. on 30 Dec 2006, 05:16 pm

Hi Wayne,
Why not use an opamp (or even some custom designed discrete transistor circuit) to do both?

Vinnie

The concept of using the Burson Buffer after the SB is exactly what that is. It is a custom designed, discrete, solid state circuit that can also add 6 db of gain. It has it's own power supply. http://www.bursonaudio.com/Burson%20HDAM%20Buffer.htm

The transformer coupled tube circuit is just a blue sky concept to keep the circuit path very minimal and provide isolation from the SB. We won't know what sounds better until I actually build one.

Happy New Year to you and yours.

Charles Calkins · « **Reply #3 on:** 30 Dec 2006, 05:56 pm »

Wayne:
I run my digital modded (By you) SB3 at 100% volume. From there it goes into the dac and then to an AVA T7ECR preamp. I use the preamp volume control. I'm very pleased with the results. I'm kind of leery bypassing the preamp.

Cheers
Charlie

Wayne1 · « **Reply #4 on:** 30 Dec 2006, 06:06 pm »

Charlie,

When I first started modding the SB, mgalusha and I tried it with both a preamp in circuit and without. We both thought the sound "improved" without the extra circuits and connections involved with a pre-amp. Quite a few folks have tried this since then. VinnieR has also remarked on the "improvement" of the sonics. Zybar has recently posted that his analog modded SB2 going straight into his power amps seems to be an improvement over running a digital modded SB into his modded TacT and then into the amps.

All I can say is to try it out yourself. Be Careful. For your system, as you are using a DAC after the SB, you might want to try a set of stepped attenuators to limit the max voltage into your amps.

With the analog modded SB, there is usually not a problem of too much voltage.

Daygloworange · « **Reply #5 on:** 30 Dec 2006, 06:26 pm »

Wayne,

Great news.

I'm looking forward to seeing and hearing what you come up with. Is this going to address those of us who also need another set of outputs to run to a powered sub?

Cheers

Vinnie R. · « **Reply #6 on:** 30 Dec 2006, 06:38 pm »

Quote

The concept of using the Burson Buffer after the SB is exactly what that is. It is a custom designed, discrete, solid state circuit that can also add 6 db of gain.

Cool

robert1325 · « **Reply #7 on:** 30 Dec 2006, 06:53 pm »

Looking forward to hear the outcome of this! my integrated has a passive pre, it needs a higher voltage to "wake up" . I found that my dvd - player and record player sound more dynamic and full......

Steve Eddy · « **Reply #8 on:** 30 Dec 2006, 07:11 pm »

Quote from: Vinnie R. on 30 Dec 2006, 05:16 pm

I understand that you want to try the 1:2 transformer to step up the voltage (which also steps down current output) and a buffer to lower the output impedance after the transformer. Why not use an opamp (or even some custom designed discrete transistor circuit) to do both?

'Cause transformers do voltage gain better'n anythain'!

Anyway, just seems a bit odd to do a mod which removes an opamp, only to turn around and essentially put one back in. While it may be better than the stock opamp, at the end of the day it's still... well, an opamp.

I've had very good results using transformers for all the voltage gain and active devices for impedance transfromation. Sort of like an inside-out SET amp where the active devices provide the voltage gain and the output transformer provides the impedance transformation.

Here's one very simple solution. The input transformer provides the voltage gain, and the JFETs provide the current gain. The DCR of L1 is used to self-bias the JFETs. It's fully balanced from input to output and it can use a single 12 volt power supply, same as the SB. In fact, the two could use the same supply. Preferably just a battery.

The only caveat is that it can only drive balanced inputs. However that can be amoeliorated by replacing L1 with a 1:1 center-tapped output transformer which would allow it to drive balanced or unbalanced inputs.

Edit: Oh, and if you want to use the SB for volume control, you can just replace R1 and R1 with fixed resistors.

se

chadh · « **Reply #9 on:** 30 Dec 2006, 08:26 pm »

Sorry for what's probably a stupid question, but...

If you're going to add some sort of gain stage/buffer after the SB, why not just use a pre-amp? I guess the buffer solution doesn't involve any volume control. But, is that the only advantage?

Put another way: if many listening tests have suggested the modified SB sounds better direct to amplifier than running through a pre-amp, is there any reason to believe that this buffer/gain stage won't produce the same sort of degradation in the sound?

(This question comes from someone who currently runs his Bolder modified SB2 through a tubed pre-amp, then to a pair of monoblocks that give 32dB gain, and speakers than are something like 94dB sensitivity. I know I don't need the pre-amp, but I haven't yet had time to do the appropriate listening comparisons to convince me to take the pre-amp out of the system).

Chad

zybar · « **Reply #10 on:** 30 Dec 2006, 08:48 pm »

Quote from: chadh on 30 Dec 2006, 08:26 pm

Sorry for what's probably a stupid question, but...

If you're going to add some sort of gain stage/buffer after the SB, why not just use a pre-amp? I guess the buffer solution doesn't involve any volume control. But, is that the only advantage?

Put another way: if many listening tests have suggested the modified SB sounds better direct to amplifier than running through a pre-amp, is there any reason to believe that this buffer/gain stage won't produce the same sort of degradation in the sound?

(This question comes from someone who currently runs his Bolder modified SB2 through a tubed pre-amp, then to a pair of monoblocks that give 32dB gain, and speakers than are something like 94dB sensitivity. I know I don't need the pre-amp, but I haven't yet had time to do the appropriate listening comparisons to convince me to take the pre-amp out of the system).

Chad

Chad,

Depending on the level of mods performed on your SB2 by Wayne, running the SB2 directly into your amps and bypassing your preamp "can" be a sonic revelation.

Why do I say "can"?

Well, you might like the sound of your system better with your tube preamp in the path. It might add just the right sonic flavoring that you like.

However, I am very confident in saying that removing the preamp and the ic's from the signal path will do the following:

1. Improve clarity, definition, and detail
2. Improve space and soundstage - make things more 3-D like
3. Removal of a level of digital haze

Give it a whirl, I know I am very happy that I did.

George

Turk · « **Reply #11 on:** 30 Dec 2006, 08:59 pm »

George is spot on about the SB2 direct. Mike G and I have listened extensively with pretty fine preamps and always go back to the direct. We were very early into Waynes mods and it was very early on that we made this discovery. Also, Mike installed the pair of Jensen trannies for 6db of gain and they are used with 1 meter ICs wqith no problems and a switch to put them in or out. Yes, I listen to my music louder than most. While YMMV, I have been unable to discern any difference between the Trannies in or out, just more gain:D

TomS · « **Reply #12 on:** 30 Dec 2006, 09:16 pm »

To avoid building a transformer/box for my SB2, I just use the Sonic Euphoria PLC set wide open (+4db for now). Because it's so simple with no active circuitry, just the autoformers, it preserves what is so great about the 'nekked' Bolder analog SB2 really well (dead quiet, extraordinarily transparent, smooth, very dynamic, bass still improving with break-in). What you don't get vs. the transformer solution is the isolation and ability to reverse the polarity as Wayne suggested. So far I'm quite happy with it, though I may still have them bump the gain a little to +6 or +8db. I don't expect to go back to a "traditional" preamp either.

Tom

robert1325 · « **Reply #13 on:** 30 Dec 2006, 09:26 pm »

I might get these Jensen Trafo's sometime , instead of spending my money on a cheap tubed pre....

Could I use these with endler's volume attenuators?

Daygloworange · « **Reply #14 on:** 30 Dec 2006, 09:32 pm »

If you click on the link Wayne provides at the beginning of this thread, there was some discussion on the Endlers, and various applications.

Cheers

Steve Eddy · « **Reply #15 on:** 30 Dec 2006, 10:04 pm »

Quote from: robert1325 on 30 Dec 2006, 09:26 pm

I might get these Jensen Trafo's sometime , instead of spending my money on a cheap tubed pre....

I'd recommend going with CineMag instead of Jensen. Every bit the quality of Jensen, but cost significantly less. Any of their CMOQ output trannies would work fine. They're a quadfilar, 150 x 4, so you can wire them up however you want, i.e. 1:2, 1:1, 2:1. The CMOQs come in 1, 2, 3, and 4 models all available in low nickel (50/50) or high nickel (80/20).

Quote

Could I use these with endler's volume attenuators?

Don't see why not.

se

playntheblues · « **Reply #16 on:** 30 Dec 2006, 10:34 pm »

Wayne your said "Be Careful. For your system, as you are using a DAC after the SB, you might want to try a set of stepped attenuators to limit the max voltage into your amps." I too run a DAC after my SBII, will the volume control of the SB still work in this configuration? Guy

Turk · « **Reply #17 on:** 30 Dec 2006, 11:10 pm »

Steve,

Great deal on transformers. Thanks for the link. I'll save a bunch next time.

Wayne1 · « **Reply #18 on:** 30 Dec 2006, 11:30 pm »

Guy,

The digital attenuation of the SB does work on the digital outs. I suggested that Charlie be carefull because he is currently running the SB output without any attenuation. Depending on what the input sensitivity on his amp is, he may want to reduce the output level of his DAC so he can run the SB attenuation above -40.

I contacted David Geren at CineMag after I read Steve Eddy's suggestion on the other thread. Part of his comments to me were :"For an output transformer to work right, it must be fed by a buffer amp."

The modded SB is getting it's output directly from the DAC chip. The TI specs do not give output impedance but suggest a load above 5Kohm would be best.

Steve Eddy · « **Reply #19 on:** 31 Dec 2006, 12:51 am »

Quote from: Wayne1 on 30 Dec 2006, 11:30 pm

I contacted David Geren at CineMag after I read Steve Eddy's suggestion on the other thread. Part of his comments to me were :"For an output transformer to work right, it must be fed by a buffer amp."

While that speaks well of CineMag, it doesn't speak so well of David's social life.

Er, or was this on Friday?

Quote

The modded SB is getting it's output directly from the DAC chip. The TI specs do not give output impedance but suggest a load above 5Kohm would be best.

I'd suspect that the chip has a reasonably low output impedance. While any transformer will perform best when fed from the lowest output impedance, the output trannies still perform well with source impedances up to 1k.

As for the load on the DAC, because a transformer transforms impedances as the square of its turns ratio, in order to avoid presenting it with a load less than 5k, the transformer will need to be driving a load of at least 20k.

The only other possible issue, which I believe I mentioned in the other thread, is the coupling capacitance. Because output transformers have a smaller primary inductance than is typical of input transformers (and the primary inductance will be smaller still when you parallel the two primaries for 1:2 operation), if there's too little coupling capacitance, you can end up with some low frequency peaking. Of course that might be a plus for some.

How much coupling capacitance are you using on the DAC's output?

se