advertisement


ALWSR for NAXO queries

Project stalled for a while, as I was concerned about how to wire in the ALWSR. Seems simple enough, two options in the ALWSR user manual, Super Regulator + Tracking pre-regulator, local sense and Super Regulator + Tracking pre-regulator, remote sense. In general, as I understand it, using local sense avoids some potential instability issues associated with remote sense and is the default option. Well even I could see that using the local sense wiring with multiple ALWSR in the NAXO combined with the NAXO ground plane might lead to troubles:

ALWSR local sense NAXO ground loop wiring

So I thought I would go for the remote sense, but somewhat wary of this decision I contacted @Andrew L Weekes. Thanks for your time and patience :)

Andrew pointed out two things: I should stick with local sense, but allow the NAXO ground plane to form the sole return path from the load; that the SRGND is the star point ground on the ALWSR PCB so wire in the power to the ALWSR there. So the new wiring scheme is as follows:

Improved ALWSR local sense NAXO wiring

So next daft question, what size hook up wire?
 
Last edited:
Thanks! I guess that should have been my question what would match the Naim wiring, now I just need to be careful to wire it in the correct direction :p
 
Spent a bit of time checking the Neil McBride schematics too. They are correct for NAXO for SBLs except the high pass sections have split power rails between the input buffer and output sections with two separate power 27R/47uF filters. Which is good, because I have already copied them into LTspice for simulating the filters. More of that later ...
 
Thanks Andrew, it owes a lot to TimH and your good self, he showed the way with the TeddyReg NAXO, you provided the ALWSR design, PCB and user manual. Still lots of work to go, in some order that may or may not resemble:

1) Sweep test current NAXO and compare with LTspice simulations for a benchmark for modifications
2) Recap the ac coupling tants and 47u electrolytic on relay circuit so it is good to go for another decade
3) Remove the 27R/47u and wire in the ALWSRs, fit Hicap with modified 28V output regulator assembly
4) Rewire the DIN sockets for separate power from Hicap and signal from NAC72
5) Add the Walt Jung "One Vbe" modification to the constant current sinks
6) REW UMIK-1 measure SBLs
7) Tweak pots to optimise response as measured in REW and replace pots with fixed resistors
8) Improve channel matching (standard 5% resistors in 3rd order filters? probably not very close at the moment ...)
9) um ... depends how long this lockdown goes on ...

Think I will call it SDD-NAXO for Super Dee Dooper (Dee Booper Special Deluxe a la Peter T. Hooper).
 
Well it has been a while, but lockdown working from home got me a new toy, a Picoscope 2000, no great shakes, but adequate for some sweeps of a NAXO.

NAXO sweep ch1, on Flickr

NAXO sweep ch2, on Flickr

EDIT: Couple of interesting (to me) points, channel 1 and 2 are not quite the same. The pots on mine are odded adjusted so ch1 LF and HF are c. 2.5 dB down at maximum output, a further c.5.6 dB down at a c. 2.3 kHz cross over.

While channel 2 has LF c. 2.5 dB down at maximum output, HF c. 3 dB down at the maximum output, LF a further c. 5.4 dB down at the c. 2.4 kHz cross over, HF a further c. 6 dB down at the cross over.

When I measured the pots in circuit I found channel 1 LF opt was c. 13k35, HF pot c. 11k12, while channel 2 LF pot was c. 13k31, and HF pot was 12K53.
 
Here is the schematic in LTspice generated from a combination of Neil McBride's schematic and poking about the PCB. I added in the separation of power rails with 27R/47uF as they are in a NAXO and the pots on the input to each section. I also reorganised the schematic a little so it better matches other layouts of 3rd order filters with a 1st order filter followed by a 2nd order sallen key.

NAXO SBL schematichttps://www.flickr.com/photos/184147951@N07/, on Flickr

In this case I have chucked in 13K for the LF pot and 12K for the HF pot to give something reasonably representative for the simulation, LTSpice has no pot component and I couldn't be bothered to make one so they are just resistors in the schematic. Here is the simulation.

NAXO simulationhttps://www.flickr.com/photos/184147951@N07/, on Flickr

I think it is fair to say that they match quite well. Which is as it should be but good to know.
 
Mynamemynaim said:
Teddy (Pardo) says that the more regulation the better ...ie there is gains to be had from feeding a regulated supply into another regulator close to the board being feed...and he HAS done a lot of work in this area...
Click to expand...
AIUI, that's not only about noise. There is a supply impedance advantage to having the final regulator as close as possible to the circuit its feeding.

Then again, there's also a big advantage in using a primary circuit which has good PSRR, but that's another matter...
 
Last edited:
DoctorRad said:
AIUI, that's not only about noise. There is a supply impedance advantage to having the final regulator as close as possible to the circuit is feeding.

Then again, there's also a big advantage in using a primary circuit which has good PSRR, but that's another matter...
Click to expand...

Shunt regulators are awesome, located close to circuitry. They eliminate the need for much of the perceived wisdom when it comes to audio PSU design, but are wasteful of power and ideally need their operating parameters adjusted for the circuit they power, if not to be wasteful and generate excess heat, which is why they aren't used as much as they should be. They aren't a drop-in replacement for an existing linear regulator.
 
You must log in or register to reply here.


advertisement


Back
Top