Project files: The EL2k headamp

What is it?
The PCB files for the EL2k headamp in the previous post (here)

How big are the boards?
The board measures 3.95″ x 3.0″ (app. 100 x 76 mm.) and contains one channel.

What is the status of the boards?
The files are v1.1 of the board. I corrected a few minor issues in the prototype such as the footprint of the input capacitor and the package dimension for the CRD.

Does it use any special/expensive/hard-to-find parts?
Yes.

  • The Elantec buffers (EL2008/EL2009) that for the basis of the EL2k have been out of production for a long time. You can find them in various places online, both as pulls from scrapped equipment and as unused new-old-stock (NOS) ones. I bought a single pair of EL2008s from a long time diyaudio-user and member of the headphone community who was cleaning out parts so I have no doubts that mine are genuine, but if you do try to source from elsewhere always beware of fakes or substandard parts.
    That said, demand for 15-year old analog video buffers in TO-220 packages probably isn’t at an all-time high now, so I would judge the chance of getting genuine ones as quite good. Unlike various power transistors and small signal JFETs that are regularly counterfeited – even in channels where you would never suspect it – there probably isn’t a lot of money to be made from faking these in the first place.
  • The heat sink is not hard to find as such, but it must be a specific one (the SK68/75 from Fischer). It’s quite common in Europe with Reichelt and TME both stocking it, but I am not really aware of any other sources.

Anything else I need to know?

  • When you mount the buffer as shown, the heat sink will be grounded (because the tab-pin is grounded on the PCB). Not a problem in itself, but it does mean you should be careful of flying PSU leads 🙂
  • Mounting the buffer is a bit of a hassle, firstly because the electrolytics can get in the way unless you wait with mounting those (I didn’t…) and secondly because it doesn’t line up 100% true with the grooves in the heat sink. You can either have the buffer sit flush with the board as I did and use the M2.5 slot, or you can use the M3 slot and have it raised quite a bit above the board. In the first case, there is barely enough room for the screw and in the second it might be necessary to bend the buffer leads a bit to make them reach the holes. It works, but it isn’t as elegant as I would have liked 🙂
  • The PSU for this board should be +/-12 to +/-15V. 15V is recommended unless you are using an opamp that doesn’t tolerate 15V. In theory, the PSU should be capable of at least 2A output currents, but in any normal audio application 1A should be more than enough.
  • If used, the recommended value for the CRD is app. 2mA. I used a J508 because I had a pair left over, but there are other options out there. You can also use a resistor instead or skip this entirely. See this page about class A bias of opamps.

Downloads:
Download design files here

Edit May 5th 2016: Download a BoM from here.

Related information:
See the two original posts for some more information and links.
Note: Always read the “intro post” for additional important information about my designs.

More old silicon – the EL2k headamp…

After my adventures with the obsolete BUF03 buffer comes a design with another discontinued buffer-chip – the Elantec EL2008 😀

The EL2008 is definitely an overkill-device for audio duty. It’s got high bandwidth, high slew-rate and a 1A current capability with a built-in limiter. Like the BUF03 it was originally intended for video applications and like the BUF03, it has also been discontinued for a decade or so. There is also an EL2009 with even more impressive specs for bandwidth and slew-rate if you want and since both ICs should work in this design coming up with the name was pretty simple – the EL2k.

The design itself is a fairly straightforward “buffered op-amp” circuit, the only “tricks” being a resistor between the opamp and the buffer (because the buffer needs a controlled source impedance) and a current-regulation diode (CRD) to bias the opamp into class A (something else that was very popular in diy headphone amplifiers 10 years ago – my age is starting to show here I guess…).

I’ve possibly gone a bit overboard with the heatsinking, but this was the type of layout I had in mind and it seems to work well (electrically at least, there are some mechanical niggles). The input caps are ClarityCaps ESA which probably also qualifies as overkill, but I am sure they do no harm 😀

The opamps used are LME49710s because I had a pair to hand when I did the test. However, in my opinion the obvious choice for this build would be another one of “yesterday’s heroes” – the OPA627. Back in the day when the EL2k buffers were around, the OPA627 was pretty much the king of the audio-grade opamps so I think that match is sort of meant to be. The OPA627 also has a warmer, less clinical sound signature that might offset a bit of brightness here.

I have only listened briefly to this design so far, but the immediate impression is “detail, loads of detail” Whether this turns into listening fatigue in longer sessions I don’t know yet but we’ll see when I get some more time. In any case: the guy that sold me the Elantec buffers told me “to build something awesome with them” – not sure I succeeded, but I definitely tried 😀

Project files: BUF-03 revisited

What is it?
As I mentioned in my post on my still work-in-progress “Ring” amplifier, I am planning to use (at least initially) a version of my BUF03-buffer as the input stage. Since I have now build and tested the revised boards and they seem to work well, I thought I’d publish the files as an update to the original. Key changes are:

  • In/Out connections are now grouped together as single 3-pin headers instead of the previous 2 x 2 configuration.
  • PSU connector has been uprated to a screw clamp type which is easier to work with. Also, there is now only one PSU entry instead of two.
  • The diameter of the decoupling electrolytics has been reduced to 8mm which should still be plenty for low-ESR types (Panasonic FC/FM are what I normally use). The smaller caps make the board look a little less cramped.
  • There is now a provision for a resistor at the input. This sets the input impedance and may be useful in applications where no volume pot is used. It is optional if you either want to use a volume pot immediately before the buffer or just want the previous stage to see the natural input impedance of the BUF03 (which is quite high).

How big are the boards?
The board measures 2.55×1.45″ (app. 65x37mm) – same as the original.

What is the status of the boards?
The board is now in rev. 1.5 as I’ve only made minor changes (described above). Both the original version and this one have been tested and work fine.

Does it use any special/expensive/hard-to-find parts?
Yes. The BUF-03 IC was discontinued several years ago. They are often available as used or NOS on ebay, but as always be wary of the risk of counterfeit/substandard parts in ebay listings.

Anything else I need to know?

  • DO NOT attempt to use this board with the BUF634 or any other buffer IC as that will not work.
  • The BUF-03 ICs are internally biased to class A and will get (quite) hot during normal operation. Small TO-5 heatsinks are recommended if you can find them.
  • The board can be used with the (equally hard to find) BUF-04 from Analog Devices, as long as you note the different pinouts (different location of the DC offset adjust pins).
  • Remember to measure and adjust the DC-offset with the inputs shorted, preferably like this: Apply the power, measure offset and adjust to zero with the trimmer. Once done, disconnect power and let the buffers cool down. Once they are cold, reapply power and check that the offset does not “drift” too much when they heat up (leave them on for 5-10 mins. at least).

Downloads:
Download design files here

Related information:
See the two original posts for some more information and links.
Note: Always read the “intro post” for additional important information about my designs.

 

Project files: BUF-03 stereo buffer

What is it?
This is the board for the BUF-03 based stereo buffer described here. The board is a stereo buffer that can be used as a preamp or headphone amp. The board requires a dual, regulated 12-15VDC power supply capable of supplying at least 150mA.

How big are the boards?
The board measures 2.55×1.45″ (app. 65x37mm).

What is the status of the boards?
This board is in version 1.0. It has been prototyped and tested without any errors found.

Does it use any special/expensive/hard-to-find parts?
Yes. The BUF-03 IC was discontinued several years ago and is not easy to get. They are often available as used or NOS on ebay, but as always be wary of the risk of counterfeit/substandard parts in ebay listings.

Anything else I need to know?

  • DO NOT attempt to use this board with the BUF634 or any other buffer IC as that will not work.
  • The BUF-03 ICs are internally biased to class A and will get (quite) hot during normal operation. Small TO-5 heatsinks are recommended if you can find them.
  • The board can be used with the (equally hard to find) BUF-04 from Analog Devices, as long as you note the different pinouts (different location of the DC offset adjust pins).

Downloads:
Download design files here

Related information:
See the links referenced in the original blogpost for build articles etc.
Note: Always read the “intro post” for additional important information about my designs.

buf03

Buffers from the past…

A design from some of my earliest encounters with the audio DIY hobby that I have always wanted to try is Eldon Pawliw’s headphone buffer. This is based on an original article called “Aunt Corey’s Buffered Passive Preamp” by Stereophile reviewer Corey Greenberg. The original article appeared in Stereophile in 1991 and is still available at Stereophile online – go and have a look!. Aside: While reading it, note a few funny references that make you realise how different the world was back then (looking in databooks for parts information, sending SAEs for information, not a single mention of websites, emails or Goggle anywhere etc. 🙂 )

The design for this circuit is exceedingly simple as there is just one active part per channel – the BUF03 closed-loop buffer IC by Analog Devices. The “closed-loop” part is important, because it means it can be connected outside the feedback loop of an opamp: Nesting the buffer in the feedback loop s what you’d have to do with newer buffer ICs such as the BUF634 or the LME49600. The BUF03s have been discontinued for ages (mine have 1991 date codes…), but are now starting to show up on ebay as NOS or pulls from scrapped equipment. I bought mine as NOS a couple of years ago from a diyaudio.com member but haven’t done anything with them until now.

To try out the design I made a board layout for a stereo buffer. I stuck to the original design but decided to change the bypass scheme a bit. The BUF03 is a fairly wideband device, so I thought that using 100nF ceramics as the first stage bypass was better than the 1uF electrolytic used in the article. The other bypass caps are Elna Cerafine 220uFs (because I had them). The board is two layers and measures 36x65mm.

The boards worked fine during my quick tests 🙂 The BUF03s are internally biased to class A and run very hot. In fact so hot that I need to find some small heatsinks for them I think. This also means there is a risk of thermal drift in the offset, so that will have to be adjusted later on. This also means I haven’t got any detailed listening impressions yet, but I am looking forward to trying the buffers out for real – hopefully very soon 😀