Just some questions (again) please... I am planning to integrate the whole stack into a chassis and get the RCA out directly to a Hypex Ncore NC252MP stereo power amp without using a pre-amp. I was initially thinking of getting everything together in one box but perhaps best not to have all my eggs in one basket.
However, I only have 44.8mm internal height available in the chassis. Thus, I am hoping to split the stack into 2 and use perhaps one of these (Black 40-pin GPIO Ribbon Cable for Raspberry Pi – Pimoroni) to bridge the two stacks.
1. Raspberry Pi and Isolator 1.2
2. Katana micro-controller and DAC
I am thinking of powering the first stack (RPi and I) with a LT1083 5V 2.5A linear PSU (Linear Power supply Module DC with heat slug LT1083 1.2V / 19.5V 2.5A - Audiophonics), and the second stack (MC + D) with a dual output 5V 2x 1.5A 6x LT3045 (MPAudio | Produktseite), using one 1.5A output, and perhaps using the other 1.5A output for the 15V opamps if I can find a suitable converter. Both PSU to be supplied by toroidal (not found one yet).
Questions now...
1. Does this sound (no pun intended) reasonable? Quite expensive but is this overkill and I can save some money using cheaper PSUs and not compromise on sound quality?
2. I think one of the isolator docs suggests 6V DC rather than 5V?
3. What would be the wiring and jumper configuration? I am presuming just to the isolator for stack 1 and the MC for stack 2?
4. Does the isolator provider an ATX style on/off? I don't think it does and can I have recommendation on a suitable one and where does it actually get wired up?
5. How do I possibly get the switching on sequence correct?
6. I do want to add a 12V trigger control to turn on my power amp (I think it will accept as low as 5V but need to check on this), so how can I integrate this? I was initially considering the GPIO but I don't think I will have access to this other than from the back of the board?
7. I think there is an option for a hardwired volume pot in addition to the ALSA controlled volume control on the DAC? Where is it wired?
8. Has anyone any thoughts on how the RPi/Katana/Isolator might compare to a Linn Majik DS streamer in terms of sound quality? My fallback plan is to get a second hand Linn Majik DSi/M which will probably work out as slightly cheaper than what I am planning to do.
Thoughts / comments would be very welcome...
Thank you.
However, I only have 44.8mm internal height available in the chassis. Thus, I am hoping to split the stack into 2 and use perhaps one of these (Black 40-pin GPIO Ribbon Cable for Raspberry Pi – Pimoroni) to bridge the two stacks.
1. Raspberry Pi and Isolator 1.2
2. Katana micro-controller and DAC
I am thinking of powering the first stack (RPi and I) with a LT1083 5V 2.5A linear PSU (Linear Power supply Module DC with heat slug LT1083 1.2V / 19.5V 2.5A - Audiophonics), and the second stack (MC + D) with a dual output 5V 2x 1.5A 6x LT3045 (MPAudio | Produktseite), using one 1.5A output, and perhaps using the other 1.5A output for the 15V opamps if I can find a suitable converter. Both PSU to be supplied by toroidal (not found one yet).
Questions now...
1. Does this sound (no pun intended) reasonable? Quite expensive but is this overkill and I can save some money using cheaper PSUs and not compromise on sound quality?
2. I think one of the isolator docs suggests 6V DC rather than 5V?
3. What would be the wiring and jumper configuration? I am presuming just to the isolator for stack 1 and the MC for stack 2?
4. Does the isolator provider an ATX style on/off? I don't think it does and can I have recommendation on a suitable one and where does it actually get wired up?
5. How do I possibly get the switching on sequence correct?
6. I do want to add a 12V trigger control to turn on my power amp (I think it will accept as low as 5V but need to check on this), so how can I integrate this? I was initially considering the GPIO but I don't think I will have access to this other than from the back of the board?
7. I think there is an option for a hardwired volume pot in addition to the ALSA controlled volume control on the DAC? Where is it wired?
8. Has anyone any thoughts on how the RPi/Katana/Isolator might compare to a Linn Majik DS streamer in terms of sound quality? My fallback plan is to get a second hand Linn Majik DSi/M which will probably work out as slightly cheaper than what I am planning to do.
Thoughts / comments would be very welcome...
Thank you.
A couple of comments. Overall, I think it is OK to split the stack, but i think you will need a different cable - one that has female on one end and male on the other, not 2 females. I have one I found on Amazon (I was thinking of splitting my stack but have not tried it yet).
Also - the Allo isolator requires at least 6 volts on the isolated side because it has an onboard LT3042 5 Volt LDO, which in turn feeds an LP5907 3.3 Volt regulator which powers the isolated side of the SI Labs chips. The supply can go up to 15 Volts, I think. Since this is on the isolated side, you could probably share it with the DAC supplies (not sure what Allo recommends).
As far as powering a relay for your power amp, you could use a spare GPIO pin on the Pi and have a script control it. I am using this approach with piCorePlayer, which has this capability. Not sure the Katana has any provisions for this. Otherwise, you would need some sort of audio detector to drive a relay.
Cheers
Whit
Thanks, yes, I did not see that that cable was FF. 6V on the isolator - yep, that is probably why I saw it as such, but the Allo power settings docs says 5V though. Very confusing!
I agree Allo makes things a bit confusing. They could do better with documentation.
Take a look at the isolator block diagram - this should make things clearer (note that this is probably for the old version - but I don't think they changed the design much except to replace the DC connector with a USB (type C?):
Amazon Drive
Also this document talks about the power options for the isolator + Katana:
Amazon Drive
Good luck!
Cheers
Whit
Last edited:
So the new isolator requires ONLY 5 volts? The docs on the new isolator page show the old block diagram, and mention 6-15 volts on the isolated side. You need to update that webpage for sure!
Thanks
Whit
That's just half the story. And it reflects the ESS marketing.
IMO the first thing you need to get right is a proper DAC-Amp integration on the gain side.
Most people here will just use 0.1V or less of that 2V DAC output (Katana) because they run a much too high gain on the amp.
And it gets even worse if you run differential.
It's known that most people won't even run 1W max into their speakers (e.g. 88db/W at 4R).
The voltage required is:
U=SQRT(PxR)=2V
to achieve that.
At a quite low - assumed - amp gain of 20 you'll end up at above mentioned 0.1V max on the DAC output.
With a higher gain on the amp that value gets even lower.
There are amps offering overall gain adjustment.
E.g. NewclassD offers gain adjustment for their amps. You just replace some Rs.
E.g. I'm not sure if I can do that in a similar way on my Abletec amps. The datasheet indicates the gain structure. But I'm not sure if I can simply replace this or that R.
Once the overall gain structure is right, you can start playing with the digital V control.
I think that's also the story behind the Benchmark DAC-3 output analog adjust feature.
You get the base level right on the analog side (DAC->amp alignment) first and for the finetuning you run digital VC.
Of course Benchmark can not do anything about the amp gain. They'd still end up at 0.1V.
Last edited:
A few more words on Katana 1.2. We have added multiple audio grade capacitors on analog stage , one of them 1000uF. Now the all important analog stage shows no oscillations , and noise on rail is 50uV above floor.
Furthermore , to complicate things , we will offer 2 analog stages. One is Pure THD+N , for those that like the machine testing, and we have broken the old numbers , we are at about -113Thd+n 1Khz , A weighted
Second stage is called "Sound quality optimized" , it incorporates a Rasmussen mod (you can read about it on DIYaudio) (apart from all other extra caps) and in my opinion it sounds better....but THD_+N drops to -100 (same testing conditions)
Of course we will offer the option to buy both. (for about 50$ extra)
Questions ?
Furthermore , to complicate things , we will offer 2 analog stages. One is Pure THD+N , for those that like the machine testing, and we have broken the old numbers , we are at about -113Thd+n 1Khz , A weighted
Second stage is called "Sound quality optimized" , it incorporates a Rasmussen mod (you can read about it on DIYaudio) (apart from all other extra caps) and in my opinion it sounds better....but THD_+N drops to -100 (same testing conditions)
Of course we will offer the option to buy both. (for about 50$ extra)
Questions ?
does the second stage require separate power?A few more words on Katana 1.2. We have added multiple audio grade capacitors on analog stage , one of them 1000uF. Now the all important analog stage shows no oscillations , and noise on rail is 50uV above floor.
Furthermore , to complicate things , we will offer 2 analog stages. One is Pure THD+N , for those that like the machine testing, and we have broken the old numbers , we are at about -113Thd+n 1Khz , A weighted
Second stage is called "Sound quality optimized" , it incorporates a Rasmussen mod (you can read about it on DIYaudio) (apart from all other extra caps) and in my opinion it sounds better....but THD_+N drops to -100 (same testing conditions)
Of course we will offer the option to buy both. (for about 50$ extra)
Questions ?
Will there be changes in the manual fron the website?
Last edited:
Is this previous pre orders? Or is there going to be a new round of preorders?
If one were to go for a Katana + isolator, how many PSUs does one actually need as an absolute minimum? What is the recommended number of PSUs?
Thanks,
Rafa.
THD+N is useless, which has been known for probably half a century. It is used only because it is easy to measure.
For people that like testing, they need to know how much is HD, and how much is N. Also, for HD, they need to know low order or high order. How much is 2nd, 3rd, and how much higher harmonics. (Low order can sound good, or at least it can sound fairly unobjectionable. High order tends to sound bad.)
With that information they might be able to make an informed decision. Otherwise, giving THD+N is virtually the same as no information at all.
Last edited: