1. Things you need to know about the new ‘Conversations’ PM system:

    a) DO NOT REPLY TO THE NOTIFICATION EMAIL! I get them, not the intended recipient. I get a lot of them and I do not want them! It is just a notification, log into the site and reply from there.

    b) To delete old conversations use the ‘Leave conversation’ option. This is just delete by another name.
    Dismiss Notice

HackerNAP clone - a messy prototype

Discussion in 'd.i.y.' started by john.luckins, Aug 26, 2010.

  1. john.luckins

    john.luckins pfm Member

    Hi All,

    I thought I would report on my HackerNAP clone, built while I am awaiting the HackerNAP parts. It is built on NAP140 clone boards from Jims audio on Ebay and is fitted with T03 output transistors. I have bodged 4 PFM Gyrator (Cap Multipliers) to fit and played around with a number of different PSU options based on what bits I have hanging around. Its not a pretty sight in its prototype form and weighs an absolute ton, but it does sound rather good and certainly better than my previous NAP135’s in both my opinion and that of others who have listened to both. It has evolved from birth over the last week and there have been no nasty accidents on the way. The cat knows to stay well clear or risk being fried…

    Circuit modifications:

    The Jims Audio boards are the standard NAP 140 circuit and good quality. The circuit is at the link below with modifications as follows:

    1. Using 2SD 756 for Vas stage current sink – higher gain, Ft and lower Cob so in theory a better device in this position.
    2. Using 2SB 716 for VAS transistor – same reasons as 1. Considerably better Hfe and speed than ZTX 753 it replaces.
    3. protection circuit not installed
    4. Still using Naim original design MJE 243 and 253 for the driver stage. Prefer these to MJE15030/1 as Hfe and gain linearity better at lower power.
    5. BUV 60 as output NPN – higher voltage and current rating to BDY58 and slightly lower Ft. I have BDY58’s to use in final version.
    6. Split the 22k resistor to the current sinks with a 4.7uF cap to –ve rail.

    Circuit additions:
    1. Powered LTP and VAS stage using cap multipliers with IRF 610 and 9610.
    2. Filtered the feed to cap multipliers through a 20mH choke, series 39 Ohm resistor and a 1000uF smoothing cap.
    3. Same supplies feed both channels.
    4. 2 X 30-0-30 200VA non screened transformers with bog standard rectifiers and snubbers (100R +0.1uF) feed front end.
    5. 2 X 432VA R core 27-0-27 with high speed 25JPF40/25CPF40 rectifiers and snubbers as above for power stage.
    6. 2 X BHC 22000uF 63 V feeding power stage
    7. 2 X BHC 15000uF 63 V feeding front end

    I started with just the two 200 VA transformers feeding both front end and power stage together and gradually upgraded the monster, listening as I went. I didn’t want to loose the characteristic Naim presentation so I have stuck with the Tantalum caps for now, and it seems to have worked. My aim was to see if I could get this build to exceed the performance of my NAP 135s which I sold 6 months ago. Initially the sound into my PMC DB1’s was like an unrefined Nait, with a slightly thin bass but with some of the Naim PRAT. Good for spinning a few disks but a bit tiresome after an hour or two.


    What made improvements to the sound:

    • Smoothing the mid point of the 22kohm feed to the two current sinks using the 4.7uF cap to the negative rail brought a worthwhile improvement to the depth of detail
    • Adding the cap multipliers and splitting the feed to the front end gave a much greater improvement in the detail and depth. This wasn’t subtle, and I became aware of detail and perspectives to the sound that I didn’t get with my 135’s. You know when you hear so many new things happening in a range of familiar recordings. The bass was still thin and lacking authority though.
    • Adding the choke made a small further improvement to the detail. Now sounding very tidy and precise.
    • Putting the snubber (100ohm + 0.1uF) across the AC terminals of the rectifier removed a great deal of the hash and glare that I thought was down to the slightly agricultural 200VA transformers I was using.
    • Finally, adding the 2 X 27-0-27 R Core 432 VA transformers and 22000uF caps to the feed the output stage alone brought back the bass weight and authority that I had missed from the 135’s.

    It may be an ugly beast but my family and friends agree that it outperforms my previous 135’s at all volumes in my 5 metre square listening room where it is driven by a CDS3 and a Starfish. My son could tell it was better from upstairs. I’m not sure that the R cores are best suited to driving the output stage compared with toroids which have less series inductance. I’m certain that the two other transformers can be improved upon. Their rating is excessive and R cores would be better here than on the output stage. Higher voltage lower power R cores for the front end are hard to find. The benefit of isolating the voltage feed to the front end is considerable and I think Naim missed a trick by not doing this themselves. It would have been considerably cheaper to do in the NAP250 and 135’s than the full power regulator.

    What is next:

    • Improve the earthing where the speaker and Zobel return earths run common mode with the front end earth for about 2 inches – a big no-no.
    • Load up the Cap multipliers with 1 watt resistors to get their output resistance down (greater current = higher Gm = lower Rout)
    • Tweak the Cap multiplier voltages to maximize them, possibly feeding the gate of the FET from a separate transformer/rectifier to eliminate interference
    • Clean the boards of all traces of flux.
    • Use non wirewound emitter resistors in the output stage
    • A Polypropelene Zobel Cap.
    • Measure the transformer inductances to establish the best values for the snubbers - tricky!
    • Source better quality Tantalum capacitors - pricey!
    • Wind new inductors for the Cap multiplier supply.
    • Consider a separate regulated feed to the driver transistors (MJE243/253). I have a suitable +- 40 volt 500mA shunt regulator
    • Build up my second set of boards and bridge the amp to remove all high signal currents from the earth rails. I understand this is what the NAP500 does. Need to consider adjusting the output impedance and damping factor as there will be two amps in series.
    • Use what I have learnt from this to improve my forthcoming HackerNAP build.

    Anyway, back to listen again….and a big thanks to Hacker for his NAP and the PFM Gyrator boards.

    Here's some piccies...

    http://s869.photobucket.com/albums/ab257/John_Luckins/

    John
     
  2. James Evans

    James Evans Bedroom Bodger

    Nice! I've got a pair of those jim's audio boards here as well. Populated except for output transistors. One day I'll get the sods rigged up and working.
     
  3. john.luckins

    john.luckins pfm Member

    I think you'll find them to be pretty good James. I am pleasantly surprised at how good they sound, but results do depend heavily on the PSU. I haven't examined the layout compared with the HackerNAP or the original Naim boards. The BUV60 output transistors are a bit of a find.

    I've actually got two different types of Jims Audio boards. Both have the same circuit layout but one has tinned pads and the other not. The tinned ones (labelled XR00008312V1.1) were just a bit easier to get soldered up.

    I've also set up a dual rail 500mA shunt Reg kit that I got from Jims Audio. Good quality parts and worked straight off. Got good after sales service too.

    See what you can do with a couple of days off when the kids have neary grown up!

    John
     
  4. James Evans

    James Evans Bedroom Bodger

    I reckon some of my projects might end up having to wait for that day :)
     
  5. TonyK

    TonyK pfm Member

    Hi John
    Where did you get the R core's from?

    Tony
     
  6. John_73

    John_73 pfm Member

    I found with Jim's boards that the polarity of the input and feedback cap (C1 and C5 on the board) appears to be reversed in direct comparison with my genuine NAP140 baords. Whether they're connected properly underneath I don't know, so all may be well, but its worth checking...
     
  7. john.luckins

    john.luckins pfm Member

    I got the R Cores from a company in Germany called HiFi Tuning. They don't sell them any more I'm afraid and these were the last of thier stock.

    A French company Selectronic at http://www.selectronic.fr/ has similar.

    Valveheaduk, thanks for pointing that out. You are correct about the polarities marked on the boards for these two caps being wrong and I've checked the reverse of the boards to confirm this. The white splodge in the cap symbol on the board indicates the negative terminal. Interestingly the Avondale Audio schematic from 2001 below gets this wrong too. I got away with getting it wrong as Tantalums can tolerate up to 15% of their rated voltage as reverse bias, and they are only reverse biased by about 100mV. It works quite happily this way so I wonder just how many are running permanently reverse biased across the planet?

    [​IMG]

    John
     

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice