advertisement


Building a B4 as a headphone amp. Any idea where I can get 1N5287 and 1N5297 (or alternatives)

sam_cat

C'est Crounchifique!
Building a B4 up as a headphone amp, can source (or already have) everything except:
1N5287/J501 x4
1N5297/J505 x4

Any idea where to source, or suitable alternatives?

Thanks,
Sam
 
I definitely have some (1N5287/97) if keithsnaim doesn’t, but I won’t be anywhere near them for a week.
 
Perfect, thanks both. There is no rush! :) Slowly pottering away at a build.

Two full sets would be ideal, as I have an experiment in mind, but one set would be great. :)
 
Dumping some search results here for me to have a read through properly later... Pile of previous b4/Bootstrap buffer threads:


Oct 23 2006 - Preamp concept - https://pinkfishmedia.net/forum/threads/preamp-concept.27680/
Noc 02 2006 - Buffer Circuits - https://pinkfishmedia.net/forum/threads/buffer-circuits.28036/
Aug 08 2008 - Buffer recision - https://pinkfishmedia.net/forum/threads/buffer-revision.51463/
Jul 27 2009 - Did Anyone ever build PigletsDad's Bootstrap buffer stage? - https://pinkfishmedia.net/forum/thr...ild-pigletsdads-bootstrap-buffer-stage.66430/
Jan 03 2010 - Bootstrap Buffer Based preamp board - Group Buy - https://pinkfishmedia.net/forum/threads/bootstrap-buffer-based-preamp-board-group-buy.73412/
Nov 19 2011 - New B4 Bootstrap buffer build - https://pinkfishmedia.net/forum/threads/new-b4-bootstrap-buffer-build.109313/
Jan 27 2012 - Another B4 Thread - https://pinkfishmedia.net/forum/threads/another-b4-thread.113269/
Apr 17 2016 - Finally finished my B4 - https://pinkfishmedia.net/forum/threads/finally-finished-my-b4.187128/#post-2842587
 
Take a fet and connect a resistor, or for set up purposes a variable resistor (trimmer), between gate and source. The drain is your "anode" and the gate your "cathode" of the resulting "constant current diode". If you require pretty accurate current you will need to adjust each one for the correct current and then replace the trimmer with a suitable fixed value resistor. The sample variation between fet's in IDss is quite wide hence the need to adjust each one on test. Try a 1K trimmer and for lower currents increase from there. With no resistance (trimmer at one end of track) you have the IDss of the fet itself and this is the max CC for that fet. Obviously you can leave the trimmer in if you want a variable CCS.

The commercial J501 etc are generally just exactly as described above but in a 2 pin package. I guess they make them with say 100R, 500R, 1K etc resistor and select them afterwards for correct current ie ones found to be within 0.9 and 1.1mA would be given the type number of a 1mA part. The higher the resistor the lower the constant current and the more "ideal" the CCS....

To test/adjust it put + 12V or so (9 - 20 fine) on the anode and a DMM set to measure current in the expected range (use 20mA or so range) from cathode to ground.
Now put the red lead of your DMM back into the normal socket on your DMM before you forget and try to measure a voltage and blow either the fuse in your DMM or something else!!! It's the most commonly made mistake with a multimeter and we've all done it... several times DOH!



If you need a really good CCS and don't need it to be a two terminal (floating) device then other CCS types will easily outperform the "CCS diode" or home made version as above, especially for higher currents. Outperform? Yep, in terms of output impedance. A perfect CCS has infinite output impedance as it's a pure current source but if you're trying to get as much as say 5mA from a CCS of the type above it will have an impedance of say 10 - 100K... perfectly fine for many uses but other non floating types can give >1M quite easily. If you only want <1mA the fet type becomes much better. Don't fret over it as depending on the application 10K or more may be perfectly adequate. The beauty of the fet type CCS is it is floating:)
 
The reason the B4 uses the J50x series is that they maintain current regulation down to very, very low voltage overheads; a FET connected as above generally won't compete, needing 3-4v to work in (the B4 needs the currrent regulation within only c 1.7v of headroom)

ATI Semitec also do CRDs, Rapid Electronics even stocks them, but in very limited ranges ( currently 2 and 4mA) These are very good for many uses, pretty stiff output impedance for a single fet-based device* and can withstand up to c 80v; but I haven't tried them in the B4 circuit: again, they are running out of voltage overhead from the datasheet. Add another green LED in the two bootstrap rail tails to give yourself c 5v a side , and it'd be worth a go (NB untested as yet, and so not quite a direct drop-in as Pigletdad's circuit is published in links above.)

* the 2mA ATISemi part is around 250kohm over a usefully wide voltage compliance range. Not bad at all, actually... use it in the reference tail of a real CCS and you are going places :)
 
I'm not familiar with the B4 so didn't know of the voltage compliance issue.... If the CCS does not need to be floating then a basic bipolar CCS with diode drop ref should do the compliance and beat the fet/CRD type on output impedance.
 
If the CCS does not need to be floating then a basic bipolar CCS with diode drop ref should do the compliance and beat the fet/CRD type on output impedance.

I tend to agree with you on this after doing a lot of builds with both types, with and without plenty of voltage to play with, however measurements by Walter Jung tend to disagree, at least with respect to output impedance. Check out his excellent analysis here:

http://www.audioxpress.com/assets/upload/files/Sources_101_P1.pdf

Sorry to the OP if I'm going in to far here...

John
 
I'm familiar with Jung's work on this and it's generally good stuff. However my own measurements disagree with Jung's in certain areas.
I have also done a lot of work on this including designing and building a CCS test instrument which can measure impedance at any test frequency and voltage overhead and the fet types don't do very well unless set for low currents... it's the degeneration of the source resistor that helps.

With a bipolar a LED will give better results than a single diode but obviously reduces voltage compliance somewhat.
 
Another interesting thread:
https://pinkfishmedia.net/forum/threads/b4-buffer-as-a-headphone-amp.90290/

Oddly enough its about using B4 buffer as a headphone amp.... some useful stuff and some questions I had concerning output caps have now been answered.

Powered on and checked voltages last night, all looking good. Gonna stuff input and output caps into it (as source is PC and DOES spit DC sometimes, yes I could null the DC using trimmers but I want it to just be idiot proof, hence in and out caps)
Bit more wiring tonight hopefully.
 
Finished assembly and done first simple tests (powered up, check rail voltages, DC etc). Knocked up a simple 50ohm 'test headphone' using a pair of 5 watt WW resistors.







Last image is it sat with my DIY LM4780 (60w stereo) amp. Tempted to add a second pair of RCA's on the back panel to make it possible to use it as a pre-amp in the future.. Atm its destined for use on my PC so no need for Pre functionality, but may as well while its on the bench (would have been much easier before wiring, hindsight!)

Given half a chance tonight it will be scoped, test tones etc and then headphones hooked up once I am happy.

Amazing how projects take 10 times longer with a 4 year old in the house! :)
 
Last edited:


advertisement


Back
Top