advertisement


DIY NAP250

Zobel ground shouldn't go to the input 0V, separate connection to the PSU is best.

Neil McBride has board layouts for 135 clones.

Pete
 
You know there are Gerber's out the already right? With better grounding and separate front end power options like the HackerNap or boards ready made from Avondale.
 
I've corrected another issue: I had removed the input coupling cap because my preamps all have them on the output. But that allowed the non-inverted LTP input to "see" the 100K resistor leaving it with a different impedance from the inverted input.

So I've added provisions for coupling cap options: an above-board 10V tant or an under-board polyprop. That should afford for another good comparison.

I'll post updated pics tomorrow.
 
Draft 2:

1) separated IPS/VAS ground from OS ground
2) more copper for OS
3) replaced input coupling cap (with options for above-board tant or below-board polyprop)
4) added more info on schematic

 
Moved Zobel ground to keep it from coupling with FE traces.
Fixed input resistor value and input coupling cap polarity.
Added jumpers so a single-supply version could easily be built.




 
Looking good. Which layout package do you use?

Might be worth running group buy, there might be quite a few interested in a build.
 
I use Kicad.

I'd be willing to do a group buy if there's interest. Either way I'll post the source files once I have something that works.

Cheers,
Jeff.
 
Never used Kicad myself, I use Eagle. Your renders look nice. Eagle works well with Fusion for renders but is a bit of a pain to get your head around.
 
I do a lot of Pass designs which while being simple, don't fit within the free version of Eagle's board limits. This one, for instance, is 14" x 1-7/8":

 
An update on the board designs. Main changes are to shorten paths in CFP and to join +/- outputs before feedback take-off point.

Regulators (both Naim and inverted-Naim) also get shroter paths in CFP along with more compact compensation networks.

Compensation schemes remain traditional Naim: Miller + lag for amp and Miller/dominant pole hybrid for regulators.



 
I bought 25 BC239Cs from Langrex. The theory is that Naim selected these for Vec (since the BC239 datasheet only claims 25V), and then for matching.

So I avalanche tested mine with a 0 to +86V supply through a 33K resistor to the collector, and the emitter to ground.

With the base also tied to ground, they all held tight all the way up to 86V.

With the base floating, about 1/3 avalanched just above 50V, 1/3 avalanched between 50V and 60V, and 1/3 had not yet avalanched at 61V (I didn't run them all the way up to 86V in this test).

So I don't know that there was a lot of selecting to be done....
 
For many years, pre CAT testing, the transistors will selected by careful (and time consuming) checking of an amp's DC offset by putting the device under test as one of the matched pair. All with the same offset were deemed matched.
 
This site contains affiliate links for which pink fish media may be compensated.
You must log in or register to reply here.


advertisement


Back
Top