[regal]

"
. soft/sharp switch on DF has been done as well. ?

How did you do the slow roll-off,  did you cut the PCB trace or lift the pin?  Did you hear an improvement?

With the bass issue,  you are probably picking up hum in the inductors,  try air wiring them to the side of the board.  A "silence" music track
is helpful in finding good placement for the inductors.

[zware]

"How did you do the slow roll-off,  did you cut the PCB trace or lift the pin?  Did you hear an improvement?
"

Was done week ago with lift the pin, then he made 3 wires connection to added switch on the back. Like it has been done on Spoiler I think.
It is not on my dac. I am expecting that first audition will be on the end of next week. If sucesfull I'll replicate work on my dac as well.

[regal]

I ordered these for the B+

JENSEN AUDIO GRADE ELECTROLYTIC CAPACITORS
Premium Dual Tube Amplifier Power Supply Capacitors.

1. 47+47/500VDC, 1&5/16" diameter, 2" tall, clamp mount type capacitor. Aluminum case with plastic jacket.  Solder lug terminals on bottom are stamped "+" or "-". Replaces the worn out 50+50/500v LCR caps found in British amps like Marshall, Vox and Hiwatt.  $35  EACH.


2. 100+100/500VDC, 1&5/16" diameter, 2&1/2" tall, clamp mount type capacitor. Aluminum case with plastic jacket. Solder lug terminals on bottom are stamped "+" or "-". Perfect for old British amps. $45 EACH.

[markC]

Thanks for the pic and info, Regal. I think I will do the change to the I/V circuit, but leave the filter as is. Yesterday, we had the opportunity to listen to 4 different Dacs in the same system. The Dac-60 was about 4-5 db louder than the other 3. I wonder how many db I would lose by just changing the resistors and not adding the 1k uf by-pass cap?

[regal]

You would lose a lot,  it would be  0.7V RMS.  Plus changing the resistor changes the filter !  Who knows what -3d and slope you would have.

If you want to avoid the by-pass caps you would need to design a bessel filter with a resistance to ground of at least 100 ohms.  This would give enough Vrms out (~1.25V).  I would not go lower.

Jim Haggerman's website has a bessel calculator,  Your R value would be 200 (2 in parallel give the 100 ohms.)

[markC]

Right, duh! It would change the filter. So, I guess if I want to reduce the load on the dac chip, I'll have to do the filter as well. I noticed Digi lists the Elna by-pass cap as non-stock. Did you get yours a while ago?

[regal]

I ended up getting Muze bipolar caps,  supposedly a bipolar cap works better on a cathode bypass because the signal is AC.

http://beezar.com/oscommerce2/catalog/product_info.php?cPath=24&products_id=29&osCsid=8b06e4a97642d1abc3c6d0a866ecd4cf


Also if you are ordering from Digikey you may want to try a different inductor.  It seems that TDK has changed the shielding on these since the last time I used them,  they pick up a little buzz if not air wired away from the board.  Inductors are relatively cheap and I plan on ordering a few different brands on my next order.


Anyone have info on replacing the rectifiers with XYZ Hexfreds ?

[regal]

Here is the choke calculation:

R35 is rated 1W at 150 ohms.

Maximum Power = 1.0 watt = V*I = (R*I)*I = R*I*I

Maximum Current = I = SQRT(1.0 watt/(150 ohms)^2) ~= 8mA

So, any choke with current rating greater than 10mA with DC resistance around 150 ohms will do.

A quick look at Hammond products yields the following candidates:

156L : 5H, 75ma, 135 ohms, 400VDC
159P : 10H, 125ma, 155 ohms, 500VDC
193B : 12H, 100ma, 155 ohms, 600VDC


Also I have just seen another SRRP tube DAC schematic and they used no filter,  just an I/V resistor.  Maybe worth trying.

Lastly I wanted to share this photo I found of the Dac-72 modified which has the same B+ circuit as our Dac-60.  I can't really make out what is going on here with the caps.  Anyone have an idea?

[markC]

Looks like 2 single caps stacked with a dual cap added to the singles to achieve the necessary uf value? R removed and replaced with a choke.

[Cappy]

Hi there guys,

I just read much of the latest reponses on this thread yet, just having discovered it again, but I've thought I'd comment since I've made some mods to my Lite Dac and am very pleased with them.  I have a Lite Dac 50, which is the PCM63 version.  The stock circuit is mostly identical to the Lite Dac 60.  It uses a 110 ohm i/v resistor because the PCM63 puts out twice the current, has an LCLR analog filter instead of LCLC, but in almost all other respects it is the same.

In my opinion the big win with swapping capacitors in the B+ is getting rid of poor sounding electrolytics.  So going to Jensens perhaps won't help much.  I replaced the final B+ cap with a 70 uF Diy Hi Fi Supply Obbligato Oil Cap and a 39 uF 250v (I think) Solen film cap.  The 70 uF oil cap is huge but it does just fit.  Now this is a good improvement.  Too bad there isn't room in the chassis for another 70 uF Obbligato as the first B+ cap.

I replaced the CRC resistor with the Hammond 159P.  This also was a very good change sonically.  It really reduces noise in the power supply and makes it react a lot quicker, based on PSUD II simulations.  Listening confirms that.  I think the stock CRC electrolytic based power supply is overdamped and sluggish sounding.  Ripple goes from 86 mV to 1.3 mV which is almost two orders of magitude lower.

I'd get the I/V resistor as low as possible.  The lower it gets the better the thing sounds.  Many of us have too much gain in our system anyway.  Particularly with those folks with active preamps.   I took out the stock metal film (I think) resistor and replaced it with a 15 ohm Mills.  With the Lite Dac 60, you probably won't be able to go that low, but 30-50 ohms will definitely help.  I put in little jumpers made out of high quality wire in place of the resistors, so I could clip in different brands and values.

I thought adding the cathode bypass made a nice sonic improvement.  I used 180 uF Sanyo Oscons based on the recommendation of "Ruach" on Head-Fi.

I used Mundorf ZN 2.2 uF as the output caps.  Going to a high uF for the output caps isn't advisable in my opinion.  But of course all this stuff is system and hearing preference dependent.

I like regal's idea of pulling out the analog filter.  I'm going to try that next.  Lukasz (at www.lampizator.eu) doesn't use an analog filter on his SRPP output stages. 

Another thing I'm going to do is try Cree rectifiers to replace the B+ stock rectifier bridge.  Why have low recovery Hexfreds when one can have no recovery diodes?  Will they sound better than Hexfreds?  Maybe, maybe not, who knows.

One big change I'm in the process of trying is trying an Aikido output stage.  This of course won't fit in the current chassis.  Hmmm, now that I think about it, since I'll have to expand the chassis maybe I can fit in another large Obbligato cap for the B+.  Here is a thread where I discuss my Aikido output stage, which I'm also going to try with the Twisted Pear Cod and Buffalo DACs:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=120119

Regards, Bill

[regal]

thanks for the info.


The ultimate upgrade for the analog section I think would be a Sowter 8347 transformer.  Takes care of the filtering and presents a low load to the DAC. Would also not require a cathode bypass.  Problem is they are $200 per pair.

[markC]

Now we're pushing $1K. I wonder how much we can polish this design. After listening to several dacs last Sat., I still give the edge to my friends diy 6n1p unit. Can't remember what chip exactly, but could easily find out if anyone is interested. It's Burr Brown. Maybe 1730. Definately a little cleaner than my modded Dac-60.

[regal]

 I've heard quite a few commercial DAC's and the modded DAC-60 is very high end.    One thing to keep in mind is the PCM1704K is considered by most to be the finest DAC chip ever made,  it will probably never be surpassed.  The chip makers have moved to cheaper inferior technology.  They only recently brought back the 1704 because scientific instrumentation designers couldn't match its performance with the newer chips.  So the heart of the DAC-60 really has unlimited potential.

[gitarretyp]

Now we're pushing $1K. I wonder how much we can polish this design. After listening to several dacs last Sat., I still give the edge to my friends diy 6n1p unit. Can't remember what chip exactly, but could easily find out if anyone is interested. It's Burr Brown. Maybe 1730. Definately a little cleaner than my modded Dac-60.

If you or your friend could post the details of the 6n1p dac, i'd be interested.

[markC]

It is the pcm1730 chip that he is using. I/V conversion is handled by opa627 opamps then handed off to the 6n1p's. All regulated power supplies with attention to grounding. I think the key here is that the single chip provides current out + and - for both channels independantly.

[regal]

The PCM1730 was not a great chip,  all sigma-deltas have differential out.  The reason you liked it best was probably the tube stage.


I put in the Jensen 2 pole caps(100+100 & 47+47) and the choke today.  It does seem to help the bass some but I can't really comment w/out being able to do an AB.

You can't cut the trace on the DF1704 to wire into slow mode,  the ground plan must extend deep.  I ordered a fine bevel soldering tip to attemp lifting the pin.  I am not really sure how Emperical Audio was able to do this,  the pins on the chip are very small.

I also ordered a Pass D1 clone kit so I will this project is about over for me.

[markC]

The reason I liked it best was because it sounded a little cleaner in the treble region. The upper highs were more pronounced and oh so clear. This is something I read b4 I started with the journey, (that the Dac-60 is a little soft in the top end). As close as I've come I'm still not quite there in terms of upper freaquency resolution.
The Jensen mod, for me, brought a more clear, in room type of presence with the mid range. Most noticable to me with female vocals.

[regal]

Marc what tube are you using?  I'm noticing the same issue.  But I think it is the NOS Tesla tubes I'm using.

Also did you change the tube bias?   I calculated the stock bias is over 20 mA which is over the operating range for 6922 tubes,  the 330 ohms resistors takes the bias down to 6 mA.

[markC]

I'm currently using Amperex orange label 6922. They are the best yet in terms of bass, midrange and lower treble that I have tried.  I only have about 15 hrs. on them as I just replaced the Amperex orange 6dj8's. They are a close 2nd, but don't have quite as much of the palpable sound of the 6922's.
I have not yet made the resistor changes. The more I think about it and consult with a friend the more I think that changing the 220r I/V resistor to 100r shouldn't affect the filter. Should the filter not work the same just at a lower voltage? Once I find a source for either Muse or Elna caps, I'll order up the 100r and 330r resistors.
Did you notice a blacker background when you added the choke?

[pbrstreetgang]

Agree on the tube sounds. American plant orange globes are my fav then harleen 6922 Orange PQ. The white BB 6DJ and orange globe harleen 6DJ8 are slightly softer in the dynamics and highs. An etereal change of pace for a good price are Mullard in the red national print are a good change of pace. Ediswans are pretty good(in between the Amprex and Mullard)