Hello,
But some people says the choke will give you a better regulation. The transformer, rectifier will have an easier life.
I like to use choke input. Stefan doesnt like it.
You can always try. I am sure you can find some information about it in French language too.
Sincere greetings, Edward
But some people says the choke will give you a better regulation. The transformer, rectifier will have an easier life.
I like to use choke input. Stefan doesnt like it.
You can always try. I am sure you can find some information about it in French language too.
Sincere greetings, Edward
But if a 4 board stack draws 1.3 amp (as measured by Stefan), doesn't this provide enough current to make the choke work properly?
I would also like to concider a higher bleeder value (draw less current & waste less heat).
James, how did you come up with 24ohm for the bleeder? In order to draw +/- 1Amp?
Happy to learn the LCLC work well for you.
Hi,
With four shunted boards there is no real need for a bleeder. The induction in the choke is higher with more current running through. A higher induction will give you better regulation.
The Lundahls are made for max. 1,5A so if you run close to 1,5A the regulation is optimal.
So if you use one board a bleeder can help.
With four shunted boards there is no real need for a bleeder. The induction in the choke is higher with more current running through. A higher induction will give you better regulation.
The Lundahls are made for max. 1,5A so if you run close to 1,5A the regulation is optimal.
So if you use one board a bleeder can help.
Hi Stefan,
Very nice job indeed.
I am very happy with my actual configuration with 4 decks, the rpi and the S03 isolator/reclock.
I like to upgrade step by step and take the time to "taste" the mods one by one and the next will be the shunts that are on the way
I have another DAC that I build for my daughter with only one board and I want to use it to try the shunt regulators before to invest on 16 regs for the next step in my DAC. Reading in the technical notes in the tent website is clearly mentioned that when the shunt is installed near the chip only one 10uF is needed in the input side and NO caps are required in the output. Looking in your board seems to me that the shunt ARE near the chip but you still have installed the Wima caps. I see that you use some pins to easily try different values. You don't try without this caps? I know that you are a guy very precise and if you use them you should have a reason that maybe I miss somewhere in the thread?
I want to attach some pictures of my DAC but tonight my camera have no battery and the 2 attached are the only two that I managed to do
Thanks and Regards,
Enrico
Attachments
Hello,
The attachment is in French so i will translate the essential. The graphic shows the inductance related to the current in two different chokes. A is a choke with normal dimensions and 4 ohm serie resistance and B is a smaller one with 1.4 ohm serie resistance. The smaller one has a smaller air gap that will give a higher induction with low current but once the current gets high there is not much induction left. The bigger one has a bigger air gap so induction value will be less dependant on the current. In the text you can read ( if you can read French that is) that with the same choke you will obtain 200mH at 200mA and only 16mH at 1500mA.
So there are several things that can determine the inductance in a choke.
In the text they first talk about the little one and then about the big one( but both they refer to it as model B but the second B must be an A)
They prefer the bigger one ( the one on top in the graphic)
Yes more induction will give a better regulation but more current will never give higher induction.
This is a copy from French l'Audiophile magazine.
I would always use a bleeder to make it function like a choke input with or without the dac attached. If i am right if you put a 100 ohm bleeder there will be enough according to the formula.
Sincere greetings, Edward
Attachments
Hi Enrico,
Nice that you got it running with the Pi and the isolator.
My dac is now sort of run in, so in the coming time I will try out different wima caps and also without the caps on analog side, in theory however it is best to keep some wima here because the shunt is feeding three dac entries and I do not want them interacting. But its nice to try and listen what the effect is.
The shunts also have a wima on the output.
Regards,
Nice that you got it running with the Pi and the isolator.
My dac is now sort of run in, so in the coming time I will try out different wima caps and also without the caps on analog side, in theory however it is best to keep some wima here because the shunt is feeding three dac entries and I do not want them interacting. But its nice to try and listen what the effect is.
The shunts also have a wima on the output.
Regards,
Gentlemen, just a quick message to share my excitement.
I've just installed 4*Belleson to my stock single deck, with the 12V PSU modded.
I'm experiencing some heavy butt kicks... This is so great I'm delaying the Choke input assembly.
Thank you Eduard and Stefan for your explanations about bleeder resistor.
I've just installed 4*Belleson to my stock single deck, with the 12V PSU modded.
I'm experiencing some heavy butt kicks... This is so great I'm delaying the Choke input assembly.
Thank you Eduard and Stefan for your explanations about bleeder resistor.
Hello Supersurfer,
I never did learn about chokes at school. But the air gap is there to allow more current without saturation and this is important because satuaration will make the inductance drop.
It appears that it isn't the easiest thing to design a good choke for choke input. You can use a normal overrated one but there will be a considerable voltage drop across the choke which it must be able to withstand. For this Dac inductance is low, so is the voltage only high current. BUT i remember with tube gear that the Tango chokes designed for choke input were much better than an every day choke. Lundahl isn't an everyday one.
If i am right the shunts will keep the output impedance low so maybe for the things ahead of it you should go for better filtering? More mH? more caps? Carefull chosing the rectifier and transformer ( static screen?)
Sincere greetings, Edward
I never did learn about chokes at school. But the air gap is there to allow more current without saturation and this is important because satuaration will make the inductance drop.
It appears that it isn't the easiest thing to design a good choke for choke input. You can use a normal overrated one but there will be a considerable voltage drop across the choke which it must be able to withstand. For this Dac inductance is low, so is the voltage only high current. BUT i remember with tube gear that the Tango chokes designed for choke input were much better than an every day choke. Lundahl isn't an everyday one.
If i am right the shunts will keep the output impedance low so maybe for the things ahead of it you should go for better filtering? More mH? more caps? Carefull chosing the rectifier and transformer ( static screen?)
Sincere greetings, Edward
Thanks again. I emailed Guido today and he said the 8v shunt regs are fine with anything over 10v input.
I also asked him about what the shunt regs ideally prefer on their input and if that should change depending on what's being driven at the output. He said OSCON caps would be good for all.
Thanks, I'll try this and see how it works. Hopefully I can get my USB scope hooked up as I'd like to see what difference these changes make.
I've got a few different values of bleeder resistors and some proper heatsinks ordered today, so you won't need to worry about my little firestarter any more
Last edited:
Hello James,
If you happen to have 2 identical resistors you could try which works better connecting the bleeder at the cap after the first choke or at the cap after the second choke. Because the bleeder only needs to sink down about 100mA to make it a true choke input you dont have to worry to much about heat and you can take more about shorter '' power lines '' for the circuit where most of the current is flowing.
It could take a few months before i will buy the DAC. Maybe i will even wait for a new board that will be better for use with the shunts.
I have at least 3 different transformers in '' stock ''that i could use for it. Also have some chokes. Only have to order some caps but the values used are rather small compared to amps like an Hiraga. Doubling the value is just a few euros. BUT it could be less is better. Panasonic has some new ultra low ESR caps on the market.
They have audio grade, low impedance, low ESR. I remember in the Hiraga amp decade usually they did focus on ESR.
Sincere greetings, Edward
If you happen to have 2 identical resistors you could try which works better connecting the bleeder at the cap after the first choke or at the cap after the second choke. Because the bleeder only needs to sink down about 100mA to make it a true choke input you dont have to worry to much about heat and you can take more about shorter '' power lines '' for the circuit where most of the current is flowing.
It could take a few months before i will buy the DAC. Maybe i will even wait for a new board that will be better for use with the shunts.
I have at least 3 different transformers in '' stock ''that i could use for it. Also have some chokes. Only have to order some caps but the values used are rather small compared to amps like an Hiraga. Doubling the value is just a few euros. BUT it could be less is better. Panasonic has some new ultra low ESR caps on the market.
They have audio grade, low impedance, low ESR. I remember in the Hiraga amp decade usually they did focus on ESR.
Sincere greetings, Edward
Hi Dusty, I noticed nobody had comment on you post. You’ve made some good observations. Getting a low jitter high quality I2S feed is definitely proving a challenge and there is still a big scope for improvement in this area. Having said that even though on paper jitter figures from the Rpi and (to a lesser extend) the WaveIO are not super stellar, in reality they sound better then their humble origins might suggest. In a direct comparison the Rpi was able to sound better (imo & and a few others present at the time) against (Supersurfers) tricked out DC player (connected via SPdif to the DDDAC).
Most people participating in the discussion in this thread seem to be focusing on the powersupply at the moment. Hence your comments were lost in the chatter.
There are other discussion on DIY audio focussing on a low jitter cape for the Beaglebone black, which might intrest you, which will hopefully work with the DDDAC in the future.
Please contribute if you have more suggestions in order to improve at the I2S input side.
Last edited:
I am particularly interested in this isolator from the given link. Has anyone here experienced it with this Raspberry Pi and whether it will work with Doede's DDDAC? Is US$125 for the unit a little steep? Very little infor about the master clock nor is visible from advertised images.
An externally hosted image should be here but it was not working when we last tested it.
Having both Raspberry Pi model B and BBB rev C here, I hope to get the I2S connecting direct to DDDAC. Right now am using them via USB to WaveIO with Volumio, and still finding the sound better than my existing dual PC setup on WinServer 2012 with JPlay 5.2 at Hibernate mode.
Stefan's unregulated PS for DDDAC is my next project, and here to thanks you all for the inputs in especially Stefan, Edward, James, and all.
For that price I would at least expect some technical info about it! I would not pay $125 for it unless I got some details of what is in it - it may just have two $10USD isolators and power supply components? Did you inquire for more technical details from the vendor? Thanks.
I now have 24ohm, 33ohm and 47ohm 50w resistors, plus some heatsinks to play with
If I only need to draw 100mA to get the choke input to work properly, I might try to use the 47 and 33 in series to give 80ohm and pull 12v / 80ohm = 150mA. I also have some 1uf caps to try placing before the first choke.I can change the location of the bleeder no problem, but if I attach it after the first choke, then the second choke won't see the load will it? When you say this may work better, is there something I can test or measure to see which works out best?
Also, I haven't decided proper locations for the power supply components yet. I can easily mount that new strip of board on the front of the case so that the wires for the bleeder are much shorter if you think that will help?
New panasonic caps sound interesting. Do you have details of the specific model?
I see Panasonic now offer an OSCON solid polymer cap with reasonably high capacitance and very low ESR called the SEPC
OS-CON Aluminum-Polymer Solid Capacitors - Panasonic
If what Guido says is right and a shunt regulator prefers an OSCON type cap before it regardless of whether it's powering an analog or a digital stage, maybe these are a good option for the mainboard and most of the DAC electrolytics for those using shunt regs?
Yes, I meant to reply to this as well, but it got lost along the way
When we compared the I2S direct out of the Pi compared to the via USB over the WaveIO we were shocked. In theory it should be really crappy as the Pi clock is not all that great, but the benefits of using I2S direct outweigh this and the sound is easily as good, if not better than going via the WaveIO, plus it's less complicated, less expensive, less components etc. give it a go and see what you reckon, it's very easy to do. Especially when you have software like the PiCorePlayer which supports it all out the box with no software modifications needed. You just have to solder a header on, which is a very simple job.
I have mine going via a simple isolator too (using the free DIYINHK board), but honestly I haven't done back to back testing and I'm not convinced it makes any real difference.
Hi Palmito,
I did send them an online technical query via their website. There was no email address provided but online form via contact us. Til date, no response!
I hope Miero & Russ can come up with a suitable BBB Cape that would work with Doede's DDDAC. For the mean time, I'd be grateful for the guidance getting my Pi and DDDAC via I2S with a decent isolator. I read Enrico has finally has his Amanero Isolator/Reclocker from Acko up and running after much efforts. Is there an easier alternative?
Regards,
Chanh
Hi Chanh, the first step is to get the Pi talking to the DDDAC over I2S without using any isolator. Did you get that far yet?