advertisement


The art of Sen

Yes I think something is amiss.
No pre amp just power amp at 20K
Not sure what you mean about the phases though, if you mean are the two Buffalos working as a stereo pair then yes. I played a right / left channel, in phase / out of phase test track.
What happens to your rig if you disconnect the batteries?
 
I still have sound :eek:

As I suspected I have a goof some were.

I just did a little test with a different scheme

I fed both the Sen quads the same Vref by just putting both Vref + and - wires in the same Vref board and plugged the Vref board into Buffalo left channel AVCC regulator.

I removed the DAC output common ground connection as this fitted in with my plug socket scheme without re soldering.

I fed the left half of the Sen board with DACa and the right half with DACb of the left channel Buffalo.
So balanced output from the Sabre DACa and DACb, no DAC common ground, but there is a ground path via the common Vref grounds to the AVCC regulator.

This is just mono now as I have not built up the other Sen board.

A very quick listen with batteries connected and not connected and now I here a difference when batteries are connected, but still sound when batteries are not connected.

The batteries do draw 40mA per quad.
 
Hi Tony

I've never tried, but I'd expect to hear sound with the DAC powered, but SEN not. The FETS are depletion types, and so conduct when the gate-source junction isn't appropriately biassed- like when they're unpowered!

Just unpowered one channel- still hear program. Reduced, level, poor sonics- but output

A red herring, I think

Paul N
 
I original expected no sound, but early on I found I had sound and commented that battery disconnect gives an unexpected result. Although at this point I had not paid any attention to the quality of the sound.

I thought about this and was happy with what happened, the upper fets still conduct source to drain as you said or for that matter drain to source I think..

Later on I mistakenly fitted a set of uncharged batteries yet did not notice any change in sound quality.
This set some alarm bells ringing.

I later changed my scheme a little and noticed a slight change in SQ, though I only listened for a very short while in mono.

However Nic has no sound on battery removal, perhaps its due to his balanced output scheme.

For now I just glued up the remaining two quads, and will try to build up next week.
 
Here is the Sen connection scheme I used.
Beginning to doubt everything, which is often not a bad thing :)

senfront01.jpg
 
I don't think you need the Vref ground there. There should be no ground connection from Vref to the SEN board.
I will verify if I get sound with the Sen non-povered.
 
OK Legato disconnected and powered down, spare DAC outputs shorted to ground.
No change other than I have introduced a hum, never mind its just a hum and will get to look at that later.
Anyway those issues are now out of the picture.

Connected back to the original scheme above but dropped of the Vref ground as Nic suggested and now there is more of a difference between batteries connected or not.
No battery, volume a little lower, not a lot.
Sound looses some sparkle, sounds a little duller.

I had a scary moment when I disconnected the battery and heard a 600W power amp transformer start to bark, power amp obviously started to oscillate, I powered down just in time to save it.
Again a further indication something has changed as battery disconnect had little effect before.

Need a little time now to listen but still a little shook up at nearly trashing a HackerNAP, probably a good time to start on the Wine :)
 
It's always a good time to start on the wine Tony! Please don't zap your HackerNAP, it's a right old pain...
 
Well I do believe its working.

For me I have had bigger steps but it is a step and in the right direction. probably no huge steps left to take?

Also implementation is still poor and no BAL/SE stage of course just stealing half of the balanced output. Perhaps no bad thing but this is new territory.

I have somehow managed to create quite a loud mains hum around -45/55dB below my normal highish listening level and a whistling HF squeal that is just audible so not an ideal environment to appraise in.
The whistle seems to be just one channel so perhaps a filter cap problem.
Anyway lots of ideas for tracking down the noise and not for the first time.


I am hearing a little more detail in the low bass, also note that I don't have full range speakers.
The upper frequency harmonics are more delicate and even though big transients are still impressive they do not over shadow the underlying material, I can here a little more going on that sits under the main material in complex passages, a few more strands of complex material are coming through.

Piano stands out as very realistic as does top notch acoustic guitar fingering.

Powerfully moody music is more emotional, just listened to Mark Isham - Blue Sun from the Barcelona album.

A lot better than yesterday, just one wire too many, was that it ?
Will be going back and putting some more thought into that extra wire.
 
Hi

Just had a quick listen to balanced out re SE (taken from + and gd). Both go through the transformer I use for BAL to SE. Tricky because of the 6 dB level differential, but I'm struggling to hear any difference otherwise! If anything, I seem to be preferring the SE.... If so, could have saved myself a lot of time and expense! But this was using pretty simple source material, something more complex might reveal more

Paul N
 
Hi Tony

I've just been listening to a piano piece- Bartok's Sonatina. Now the balanced out sounds clearly superior to the (single phase + gd) SE. In particular, higher notes are much more realistic, with natural sounding harmonics.

I too noticed a low level hum on the SE out. Buffalo /SEN are in a mains grounded Aluminium case; and the buffalo board's ground is connected to the case (at the power input the digital pre-reg feeds). I did this as otherwise the whole circuit would be floating, whichever output type I used (because of the transformer), and buzzed when unearthed.

The higher frequency noise may be an rf issue. Try touching some of the wiring, and/or the top of the battery boxes, with one hand; and trying grounding the other hand while you do this. If this sometimes removes the problem you need to find a way to simulate the effect while freeing you up! I stuck some self adhesive copper foil (not even grounded) to the top of the battery boxes, which did the trick

Paul N
 
Thanks for the info Paul.
It does look like the battery box is a problem with noise and will investigate later.
Spent quite some time having an enjoyable listen yesterday.

Also worth a mention is that I am single ended of the Sabre of course and wondered if there may be any gain by fitting the second quad and pulling DACa and DACb in a balanced manor.

You trials always used DACa and DACb I assume.
What was your DAC ground connection when you used this balanced scheme from the Sabre.
I could do with getting the charging system working next.

Also worth noting I have very long interconnects from the Buffalo to the power amps, perhaps six meters or more.
 
Hi Tony

I have a stereo buffalo II, with both + and - (DACa and DACb) feeding their own SEN "block". These feed (tiny) XLR sockets for balanced out. In addition, the just the + out of each channel feeds a phono socket; the latter is also connected to that channel's gd (at the IV R). Both SE and BAL out feed my preamp, and go through a four pole double throw switch before the Jensen transformer, my (LDR) attenuator and line stage. I just flick the switch to select between the output types. So , in SE mode, the unused outputs are still loaded by a SEN block.

The output impedance of SEN is whatever value IV R you use- so in our case 150 R- low, by my standards! One of the many obvious superiorities of SEN over Legato, in my system, was the higher quality and extension at low frequencies. It drives the next stage very well indeed to my ears.

Sort out the buzz. Get the charger working and install the other SEN blocks. Then have another listen. Then try installing e.g. the BAL to SE stage from Legato. This should sound very fine for now. Then- assuming a transformer BAL to SE stage is anathema- await XCEN!

Best wishes

Paul
 
Tony,
What values are you using for SEN R1 and R2?
When I was shopping for the build I actually succeeded in overlooking that the values for these are in kilo ohms:eek:
Cheers,
Nicv
 
Yes 680K and 470K, I know I am not setting the battery in perfect sysmetry for the 1/2AVCC but did not have a 681k
Something I will get back to on the tidyup
 
Thinking of upgrading the Riv as something easy for a change.

Considering
Welwyn RC55Y-150RBI
and
Arcol MRA0207150RBTA015

Both Metal Film 1/4W 150R ±0.1% ±15ppm/°C

Or should I be considering something more exotix rom say the Hificollective, Mills, Charcroft or dafter?
 
Riv is certainly the most important passive component. I'm still using some fairly standard PRP resistors for this, but the Charcroft will come in ones I understand what output swing I need.
 
For R1/R2 would not worry to much about precision. I'm still using those calculated for 16V with my 22.5V rails and it sounds mighty fine to me. They do affect DC offset a tiny bit, but nothing to worry about whatsoever.
 
You must log in or register to reply here.


advertisement


Back
Top