parallel feed transformers

0 Members and 1 Guest are viewing this topic. Read 4986 times.

steve f

  • Full Member
  • Posts: 682
parallel feed transformers
« on: 24 Aug 2013, 10:28 pm »
Here are the big questions. Why parallel feed? Why do most manufacturers avoid it?

galyons

  • Full Member
  • Posts: 477
Re: parallel feed transformers
« Reply #1 on: 25 Aug 2013, 12:03 am »
Well here is my take on the basics....

Parallel feed or shunt feed has been around along time, but more as a concept than practical application.  The recent incarnation was due to the efforts of Mike LaFevre of MagneQuest. Parallel feed, (parafeed for short!), was popularized by  Dan Schmalle  through Valve magazine and the technical expertise of Paul Joppa.  The original Paramour 2A3 amps from Electronic Tonalities/Bottlehead were the first commercial application of DHT SE parafeed in recent years. IMO, as kits, they were more readily accepted in the market as DIY'ers/kit builders tend to be more knowledgeable and experimental.

Parafeed splits the opposing demands on the typical SE transformer to handle both AC flux and DC flux.   Those opposing demands are DC current, from the power supply to the OPT and AC  current, the music signal from the output tubes. An SE transformer must be air gapped and have a large core to handle the DC current.  To maximize the music signal bandwidth, the AC wants to see a small core,  lower capacitance from the windings and high inductance.  Hence the compromises.

Parafeed splits the DC and AC between two parallel inductors, an air gapped plate choke for DC and a capacitor in parallel with an interleaved OPT.   The beauty is each component is designed for its current load, without the offsetting compromises from an OPT that must handle both.  I own Bottlehead Paramours, one set running 2A3’s and the other 45’s.  I also have  300B amps and 2A3 amps with traditional Transcendar OPT’s.  The Parafeed amps just seem to do everything right.  The traditional 2A3 monoblocks are a bit more detailed and, perhaps, a bit deeper in the bass, than the Paramour 2A3’s, but the Paramours have cheap Speco line transformers that cost less the 5% of the Transcendars.  My Paramour 45’s have Magnequest plate chokes and nickel OPT’s and are the "go to" amps for most music.

Most manufacturers make products to sell.   Most want to just market to, rather than educate, consumers.  Make a traditional SE amp and the typical buyer understands, to some degree, the technology and accepts the design.  Make a parafeed amp and you, as a manufacturer, must dedicate resources to  market, differentiate circuit architecture and educate the consumer on advantages.  IMO, it is simply easier and more profitable to make traditional amps.

Cheers,
Geary
« Last Edit: 25 Aug 2013, 04:44 am by galyons »

poseidonsvoice

  • Full Member
  • Posts: 4019
  • Science is not a democracy - Earl Geddes
    • 2 channel/7 channel setup
Re: parallel feed transformers
« Reply #2 on: 25 Aug 2013, 02:13 am »
Geary,

A fantastic and well delineated response. I agree with everything you have said. My first DIY amp was a Paramour 2A3 - it was lovely. Fast forward 10 years and my current phonostage's output stage is technically a cascode constant current source loaded triode that is transformer coupled to the outside world using the parallel feed configuration. The core is made with amorphous cobalt. Hopefully it will sound good!

Best,
Anand.

steve f

  • Full Member
  • Posts: 682
Re: parallel feed transformers
« Reply #3 on: 27 Aug 2013, 05:56 pm »
Thank you Geary. I assume the parallel cap is there to block DC from the interleaved OPT?

Steve

DaveC113

  • Industry Contributor
  • Posts: 4344
  • ZenWaveAudio.com
Re: parallel feed transformers
« Reply #4 on: 27 Aug 2013, 06:20 pm »
To play devil's advocate...

I haven't tried parafeed, and the main thing stopping me is the requirement for the plate choke and parafeed caps. It seems like the additional money needed for these parts, which isn't insignificant, would be put to better use in a SE OPT. I realize the interleaved OPT for parafeed don't need to be as expensive, but adding another coupling cap of a fairly large value doesn't seem like a great idea.

@steve f, yes

galyons

  • Full Member
  • Posts: 477
Re: parallel feed transformers
« Reply #5 on: 28 Aug 2013, 05:14 am »
The devil's in the details!  There are always trade-offs.  Gonna need chokes and caps in most any well designed amp circuit.  Many use plate chokes in traditional SE designs.  Good SE OPT are not cheap.  Caps are not that bad if one wants ear candy versus eye candy.  Bottlehead parafeed amps sound damn good with some embarrassingly modest components.  You might want to give a listen to parafeed.

Cheers,
Geary


nullspace

Re: parallel feed transformers
« Reply #6 on: 28 Aug 2013, 02:31 pm »
The cost for a parafeed output amp is fairly comparable to series feed amp. To wit, if we're shopping at Magnequest for for iron to go into a 2A3 SE amp:
-- Parafeed -- EXO-003 plate choke (99 ea.), EXO-35 output (80 ea.), Obbligato cap (~24 ea.) for $204 total
-- Series feed -- DS-025 output at $225 ea.

In regards to the large coupling cap, Mike @ MQ would tell you that in series feed the last cap in the power supply, usually a big electrolytic, is in the output current loop while it is not with parafeed. Likely a good tradeoff. Of course, you could Ultrapath the output in series feed, but now you've got an additional large coupling cap.

Admittedly, I'm a fan and am close enough to MQ headquarters that I pick up my stuff rather than have Mike ship. I've gone so far as to build a parafeed PP amp, where gains from separating AC & DC is even smaller.

Regards,
John

galyons

  • Full Member
  • Posts: 477
Re: parallel feed transformers
« Reply #7 on: 28 Aug 2013, 04:19 pm »
Jon, you are quite right about the last cap in the PS!  I changed out the last electrolytic, in my series 2A3 amps, for a CSC metalized polyprop in oil.  Huge improvement in detail, timbre and imaging.  I have Sprague Vitamin Q PIO caps as my parafeed cap in my Bottlehead Paramour 45's.  I tried several boutique caps, Mundorf silver coated polyprop in oil, EVO Aluminum coated polyprop in oil, Jensen Aluminum foil in oil, Clarity cap  as well as some Russian KBG PIO and MBO Aluminum and waxpaper.  The Vitamin Q's were better overall, then slightly behind were the KBG's, MGO, then the Mundorf and finally the Clarity, (although not very far behind the Mundorf's!).

Parafeed is as good, or better, than comparably outfitted series.  I replaced the Speco's in my Bottlehead 2A3's with Altec 15708 transformers. Major positive impact with a cost of $40 including shipping.  But MQ iron still rules parafeed!  Again, the devil is in the details.

Cheers,
Geary

steve f

  • Full Member
  • Posts: 682
Re: parallel feed transformers
« Reply #8 on: 1 Sep 2013, 11:17 pm »
Looks like I'll have to try a parafeed amp. I'm really an OTL guy. The problems there are lots of tubes, and class A tops out at 3-4  WPC. I'm hoping that parafeed can get me closer to OTL sound and save tubes. I build fairly efficient speakers, so the idea of a few watts from one tube is attractive. I could build a Bottlehead kit I like the 2A3 triode, and have a 13EM7 , Jonokuchi amp that's quite nice,  Hmm, might be fun to convert it to parafeed if possible.

Steve

J-Pak

  • Jr. Member
  • Posts: 259
Re: parallel feed transformers
« Reply #9 on: 2 Sep 2013, 01:44 am »
Parafeed can also work great with headphone amps where not a lot of power or gain is needed. Magnequest blogged about parafeed OPTs to be used with a 45 or 71a DHT tube (transformers $500/pair). If Magnequest were willing to wind those with say a 32 ohm secondary, I bet they'd sound really nice in a headphone amp!

As to the series feed cap in the PSU in series feed SET amps, that's why manufacturers like Thomas Mayer use oil caps only.

Lynn Olson has some good reading on a somewhat similar note here: http://www.nutshellhifi.com/library/ETF.html