Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › LM317 or discrete transistor?
New Posts  All Forums:Forum Nav:

LM317 or discrete transistor?

post #1 of 12
9/26/09 at 2:09am
Thread Starter 
After completed this Sijosae's amp using LM317 as CCS. I may be trying out using the zener and a discrete transistor to replace the LM317 portion. Read somewhere stated that LM317 only allow 10kHz to flow through and thus affected the overall high frequency sound as human could hear frequency as high as 20kHz.

http://www.audioxpress.com/magsdirx/...a/jung2779.pdf

LM317 and resistor


For transistor like the MJE15030, it allowable frequency is 30MHz; which is far belong the human listening range of 20KHz. So, I presume it should be better than using the LM317. Please correct me if I am wrong. Also, anyone tried using such zener and MJE15030 concept before? Please comments. Thanks.

Zener and Transistor
Reply
post #2 of 12
9/26/09 at 2:38am
I have never been a fan of using a LM317 as a CCS in this kind of application. I have no doubt that the transistor implementation would be better. As far as the 30MHz spec is concerned, that is the fT (the frequency at which the transistor ceases to have any gain). You will see the performance deteriorate well before that point (perhaps at 1/10th the frequency), but that's still far above audio range.
Reply
post #3 of 12
9/26/09 at 6:09am
Thread Starter 
Thanks amb, I will start purchasing components to get ready for the zener and transistor MJE15030 approach.

Anyway, I read somewhere which claimed zener are noisy. But looking at the cct for this application, the noise if any should not contribute to the signal path.

Zener which I already have and will be using...
http://docs-asia.electrocomponents.c...6b80028a5d.pdf

Transistor MJE15030 which I am getting...
http://docs-asia.electrocomponents.c...6b8002efc0.pdf

Hope I get the right components...
Reply
post #4 of 12
9/26/09 at 6:14am
If you want simple - go with a CRD - couple of dollars from Mouser. If you want "better" but more complex, measure IDSS of a couple of JFETs and use them as CSS.
Reply
post #5 of 12
9/26/09 at 6:24am
^no you cant.

The CRD must supply at least 100mA for Class A operation.

Only discrete BJT/MOSFET based CCS with heatsinks will properly work.
Reply
post #6 of 12
9/26/09 at 6:38am
Thread Starter 
Thanks TzeYang, wondering if the heatsink I am using enough for the MJE15030? When I use the heatsink for the LM317, it is not even warm; very cool infact.

Anyway, below are the pictures of the heatsink and the LM317...on the right.


Reply
post #7 of 12
9/26/09 at 8:49am
OH MY GOODNESS LOOK AT THE GAUGE OF THE WIRES.

why not just mount them on the casing?
Reply
post #8 of 12
9/26/09 at 2:31pm
arcording to Walt Jung's measurements the 317 CCS has >1 KOhm Z at 100 KHz - not at all a problem for the audio performance of this circuit

http://waltjung.org/PDFs/Sources_101_P2.pdf

the 2 transistor feedback regulated current source would be the next step I'd take - with a smaller (faster, lower C) pass transistor

is the plate load CCS diode ref really a Red LED? - low Vdrop could give poor regualtion - use Green and a bigger emitter R
Reply
post #9 of 12
9/26/09 at 3:16pm
I have used both the 317 and 338 in the very similar aikido circuit as current source for both a mosfet(depletion mode DN2540) and a 317 as output device(like it is in the stock aikido). Sounded pretty good to me both ways.
Reply
post #10 of 12
9/26/09 at 8:37pm
Thread Starter 
Quote:
Originally Posted by TzeYang View Post
OH MY GOODNESS LOOK AT THE GAUGE OF THE WIRES.

why not just mount them on the casing?
Those wires are single core and 20AWG. UPOCC type.
The link is here

Ok, mounting to casing is a good idea too.
Reply
post #11 of 12
9/26/09 at 9:12pm
Thread Starter 
Quote:
Originally Posted by jcx View Post
arcording to Walt Jung's measurements the 317 CCS has >1 KOhm Z at 100 KHz - not at all a problem for the audio performance of this circuit

is the plate load CCS diode ref really a Red LED? - low Vdrop could give poor regualtion - use Green and a bigger emitter R
Thanks jcx, great news for me, I am kind of lazy to redo my amp if Walt Jung's measurements the 317 CCS has >1 KOhm Z at 100 KHz.

Yes, I am using the Red LED as the plate load 'CCS diode'. I am not sure what is the spec of this Red LED I am using. It is those 100 Degree Clear Wide (Spread) Angle kind of LED. Anyway, maybe I will measure the voltage across it to determine the Vdrop?
Reply
post #12 of 12
9/29/09 at 2:15am
Thread Starter 
Quote:
Originally Posted by jcx View Post
arcording to Walt Jung's measurements the 317 CCS has >1 KOhm Z at 100 KHz - not at all a problem for the audio performance of this circuit

http://waltjung.org/PDFs/Sources_101_P2.pdf
Thanks jcx, I am trying hard to locate this info on "317 CCS has >1 KOhm Z at 100 KHz" in the document. Please advise further if possible. Thanks and appreciated.
Reply
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › LM317 or discrete transistor?