The link to the post showed two schematics and naughtily I sort of added/ameded one a bit... hope nobody is offended. This is basically a solid state version of our Super Linear Cathode Follower, so we can call it a Super Linear Fet Follower.
Cheers, Joe
Attachments
Last edited:
Joe,
Any comments on the benefits of the SLCF compared to the simple cascoded source follower circuit that I use? This is the 2nd transformer output Buffalo DAC that I've built. The first one used 1:1 transformers and a bipolar diamond output buffer. And it was based on the original ES9008 Buffalo kit with simpler power supplies. This new version has better supplies, newer ESS chip, low value resistive loading to ground (thanks Joe), step up transformers, and the FET buffer. It sounds much "fuller" than the first version. Both have great detail but the first one was a bit lacking in body. This version seems to have fixed that and is very enjoyable.
---Gary
Hi Gary
I don't want to draw the wrong conclusion, but if you are using the fet follower with the constant current source and not the bootstrapped device I am showing, then there is quite a difference. Essentially the SLFF (and SLCF) runs in both constant current and constant voltage modes. It is common to see the CCS - but its also important that the actual follower element sees a constant voltage. What in reality is happening is that input (Grid or Gate), the output (Cathode or Source) and static element (Anode or Drain) are all seeing the same AC signal. Normally followers are one hundred feedback devices, you are using a trick to alters that state.
This is a quote from another forum:
"Good to hear from you John. It is really fun to watch the SLCF in action. If a 10.00 Vrms signal is applied to the grid, there will be a 10.00 Vrms signal coming out of the cathode, and the plate will also be following along with the 10.00 Vrms signal. They all move together as one. There is no gain or loss of signal, from in to out.... Paul Speltz"
Cheers, Joe
buffalo 3 dac
anyone can help me?
i have a output transformer, and i want to connect it to buffalo 3 dac o/p, here is the pic, which no.1,2,3 is correct?
anyone can help me?
i have a output transformer, and i want to connect it to buffalo 3 dac o/p, here is the pic, which no.1,2,3 is correct?
An externally hosted image should be here but it was not working when we last tested it.
Lowering the gain of the previous two output stages will likely yield similar (or maybe better) results.
I haven't yet reported back, but I did try Russ' suggestion. I modified the Legato balanced stage with ZFoil resistors, and the opamp stage with Dale PTF film resistors. The new output level of Legato is 600 mV FS. Sounds great! Unfortunately, now that it's burnt in I can't do a very good comparison from memory, and an AB comparison is difficult. Regardless though, it sounds so good I don't think I will switch back again anytime soon.
I know this is a little off topic, but I thought I should let you know I switched back from transformers and am quite happy.
How does "1" work. Or am I missing anything?
"3" will give you the full output of the DAC, assuming transformer is 1:1 ?
"2" will give somewhat reduced output, but give a fuller and more pleasing sound. I would ground the resistors, and even experiment with values.
Up to you, see what you like best.
Cheers, Joe R.
buffalo dac 3
hi, i have one question,
I have the buffalo dac 3 and would like to connect it direcly in my preamplifier,
My preamplifier is audio tekne push pull circuit with transformer 30k-30k
in cd imput.
i know that the buffalo dac is 781 ohm
I ask you the matching impedence is realizzated ?
the dc osfet it can a problem ?
thanks for the Help....
hi, i have one question,
I have the buffalo dac 3 and would like to connect it direcly in my preamplifier,
My preamplifier is audio tekne push pull circuit with transformer 30k-30k
in cd imput.
i know that the buffalo dac is 781 ohm
I ask you the matching impedence is realizzated ?
the dc osfet it can a problem ?
thanks for the Help....
I see, the D option was modified, and you did what I would have done, mentioned the source.
Cheers, Joe R.
I would like to thank Joe Rasmussen, because his idea lo charge the dac before the transformer is simple but brilliant. I have done my homework regarding this mod. For more than 2 years I have used a 1:10 transformer, loaded secondary with 1kOhm. Then I saw Joe's
drawing and began to think about it, followed by some work:
- I soldered a 1 Ohm to the primary, things opened up.
- I have a silver winding tube output transformer made by Audio Consulting. I use it the other way round, and this way it is a 1:25 step up. I have loaded my buffalo (TP24bit, battery driven, no regs) with a 0.07mm wire, about 25cm long, resistance is about 1.2Ohm. I measured the resistor with a ruler, not a meter.
This was a big step forward, and it performed very well, when we had a dac contest in my appartment, contenders were:
buffalo32, Metrum six, Meitner M1.
- A week after the meeting I thought about the discussion regarding grounding the primaries. Yesterday I soldered a ground wire to the middle point of the I/V wire, according to Joes drawing. Wow!!!:
Incredible resolution power, more micro dynamic, this is Sabre passive in current mode!
It is a pity that The Sabre chip sounds o.k. even in voltage mod, so often people use 1.1 trannies, which is still much better than most of the active I/V stages, but sounds just broken compared to (nearly) passive current mode, but there are people paying 1000 Dollars for a transformer to run a Sabre in voltage mode....
I will give you ideas to run a Buffalo in (nearly)current mod with transformers:
Buy the best transformer you can afford, you will be rewarded!
It should be from 1:10 up to 1:30.
The most interesting item at the moment is Lundahl's new MC step-up with silver winding, if it can stand the higher voltage.
Use the best I/V resistor: a piece of thin silver wire. Now the math:
I have 1.2Ohms, middlepoint to ground is mandatory, then we'll have: 7mVolt x 25(1:25)= 175mV output, 1.2x625=750Ohms (about) on the secondary side. Very low output indeed, no problem for my TVC and 105db speakers. If you need more output, you can increase the I/V a bit, but think of the output resistance. If you sill need more, you choose a "mild" active stage, gain about 12db, I think nobody really needs 2 Volts!
regards André
Buy the best transformer you can afford, you will be rewarded!
It should be from 1:10 up to 1:30.
The most interesting item at the moment is Lundahl's new MC step-up with silver winding, if it can stand the higher voltage.
Use the best I/V resistor: a piece of thin silver wire. Now the math:
I have 1.2Ohms, middlepoint to ground is mandatory, then we'll have: 7mVolt x 25(1:25)= 175mV output, 1.2x625=750Ohms (about) on the secondary side. Very low output indeed, no problem for my TVC and 105db speakers. If you need more output, you can increase the I/V a bit, but think of the output resistance. If you sill need more, you choose a "mild" active stage, gain about 12db, I think nobody really needs 2 Volts!
regards André
I much prefer grounding load resistance, too - same as Joe Rasumussen's "C" schematic
Without that, I feel like losing a little focus and dynamics (but still sounds good)
I use two 22R load resister at primary of feastrex line transformer.
22R load lowers output voltage a lot, but isn't a problem as I have more than enough gain at following audio chain.
I can not tell you, I have about 1.2 Ohm totally, a soldered wire.
Try it out and report!
Sorry, but i was just wondering if i should go for SUT's or to buy the Legato's to my Preamp/Dac-project (same box for everything). And since i don't even have the Buffalo yet it is hard.
I just wondered if anybody had tried, and if it was a big difference. Which would help me choose if i go for SUT's vs Active.
Thinking about using Lundahl LL1941 with 2,7R I/V and 1:16 step-up, which would be ~1400R output resistance (Should be approx ~1850R incl. SUT internal impedance) and approx 0,69V on the secondary - and i would end with approx 1,75VRMS after line-stage gain.
My third alternative would be even smaller I/V resistors with a pair of tubes between the dac and my linestage, but this i will probably skip for other reasons.
I just wondered if anybody had tried, and if it was a big difference. Which would help me choose if i go for SUT's vs Active.
Thinking about using Lundahl LL1941 with 2,7R I/V and 1:16 step-up, which would be ~1400R output resistance (Should be approx ~1850R incl. SUT internal impedance) and approx 0,69V on the secondary - and i would end with approx 1,75VRMS after line-stage gain.
My third alternative would be even smaller I/V resistors with a pair of tubes between the dac and my linestage, but this i will probably skip for other reasons.
Last edited:
If you have a linestage, then the Lundahl without an active stage will be perfect, but ask if it can stand the higher voltages involved here.
about 2k output is O.K., your linestage has probably 47k Ohms input. I would not be worried about the higher I/V resistor. Build it this way, and if there is sufficient gain, you can reduce the resistor. A silver wire is even better than a Texas resistor, but it would be difficult to get it in 0.07mm. It has to be thin, otherwise you need a meters! I have taken copper until I find silver.
Build it and enjoy it!
regards André
While this is about transformers, I think the schematic below works along similar principles - that you can choose a step-up transformer for the necessary "gain" or an actual gain stage. They key is to understand what is done to the Sabre DAC itself: This, in my view, is the way to get the most out of it, bring that load value as low as you can get away with.
Cheers, Joe R.
Cheers, Joe R.
Attachments
But, since my line-stage is balanced with 5,4dB gain on each phase - then wouldn't it be better to connect output from each phase with a 2.7R resistor to GND, then maybe the cap between both phases (what does this do in this specific case?) and then into the step-up transformer.
In my case the output from the transformer would go straight into the volume attenuator (50k R2-R log attenuator between the phases not down to GND) with approx 15cm wire, twisted and then maybe 15cm wire into my tube-line-stage.
Thare shouldn't be any trouble running from the step-up into the volume right? since impedance matching seems quite fine ~1.85k into 50k.
Also Joe, wouldn't it be better skipping the GND connection on the output, at least when i run transformer - to keep DAC-GND from my Line-stage GND?
Thanks alot for taking your time helping in this thread so far
In my case the output from the transformer would go straight into the volume attenuator (50k R2-R log attenuator between the phases not down to GND) with approx 15cm wire, twisted and then maybe 15cm wire into my tube-line-stage.
Thare shouldn't be any trouble running from the step-up into the volume right? since impedance matching seems quite fine ~1.85k into 50k.
Also Joe, wouldn't it be better skipping the GND connection on the output, at least when i run transformer - to keep DAC-GND from my Line-stage GND?
Thanks alot for taking your time helping in this thread so far
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.