The REAL John Linsley Hood Headphone Amplifier

I’ve had some progress with the original version of the JLH headamp. The first channel is almost done, but I noticed a small problem with populating the Q1/Q2. The holes on the pcb are too small for the legs of TO-92 devices. What would be the easiest way to solder them? I was thinking of cutting the legs short and start with one leg and the continue with the others. Maybe it’s not wise to try to drill the holes bigger?
 

Attachments

  • 8FC46FCE-8455-46E2-93C6-6750E36F298C.jpg
    8FC46FCE-8455-46E2-93C6-6750E36F298C.jpg
    346.5 KB · Views: 1,211
  • E2AE14A0-9768-4EAF-AF32-F85C59BA6758.jpeg
    E2AE14A0-9768-4EAF-AF32-F85C59BA6758.jpeg
    784.9 KB · Views: 1,246
Last edited:
I rechecked the Gerber files and there was a conversion error.
So those holes were 0.4mm instead of 0.6mm.
The solder pads on the topis 1.5mm diameter in copper, and somewhat smaller in solder resist.
So you should have no problems drilling out to 0.7mm.
If necessary, a little bit of scrapping will certainly give you enough pad all round to solder.

I shall change Gerber and repost once you have built and tested OK.
No effect whatsoever on the XEN version using SOT23 devices.


Thx,
Patrick
 
I have finally done some listening tests with the original version of the JLH headamp circuit. I think it’s quite similar to Xen version of the JLH, but unfortunately I was not able to do direct comparison between them. Instead I compared to Salas DCG-3 with my Sennheiser HD-800 phones. In my opinion the JLH was better in terms of presence and vocals. Also the bass was surprisingly deep and controlled. All in all a very good headamp and easy to listen to.

I had quite bad oscillation problems during the build, but luckily Patrick helped me out with his vast knowledge. Oscillations were tamed by placing a simple JFET buffer before the amp's input (the little green piece of proto board). Also small ceramic caps of 100pF were soldered between base and collector of the output devices. Too bad it required the buffer in front. Would have been nice to listen it directly in a more original form, but I suppose the buffer is quite transparent sonically.

I listened the amp in my newly finished case with the motherboard provided by forum member Algar_emi with Xen regs and all. I attached a couple of pics of the assembly. Still missing some parts for the SHPP protection board.

Next I will test the Xen version of JLH headamp by using this motherboard.
 

Attachments

  • IMG_2846.jpg
    IMG_2846.jpg
    567.5 KB · Views: 1,090
  • IMG_2848.jpg
    IMG_2848.jpg
    728.6 KB · Views: 1,097
  • IMG_2850.jpg
    IMG_2850.jpg
    824 KB · Views: 1,060
  • IMG_2852.jpg
    IMG_2852.jpg
    641.3 KB · Views: 514
Last edited:
Oh wow...I just installed the Xen passive crossfeed board in front of the Xen JLH headphone amp board. The sound is stunning and the soundstage really opens up nicely away from your head. Now driving my Senn HD800s with ease...:hphones:

Sorry about the poor pics but hopefully other builders get some ideas of how to mount the Xfeed board as an example.
 

Attachments

  • F87DD62C-8C35-4446-9122-243ECEAB4552.jpeg
    F87DD62C-8C35-4446-9122-243ECEAB4552.jpeg
    902.2 KB · Views: 1,062
  • A087AC3B-9618-4520-8B83-5816E03B54E8.jpeg
    A087AC3B-9618-4520-8B83-5816E03B54E8.jpeg
    833.6 KB · Views: 1,042
  • BE76804F-82CE-44FC-81CE-D67D5D4EF930.jpeg
    BE76804F-82CE-44FC-81CE-D67D5D4EF930.jpeg
    752.1 KB · Views: 1,003
  • D33AFAFF-3659-4B6E-AC9C-08A1356E1A04.jpeg
    D33AFAFF-3659-4B6E-AC9C-08A1356E1A04.jpeg
    712.2 KB · Views: 983
a) I guess you know that if you are not using a pot with the SMD XFD, you need to change the 12k resistors to 5.49k (= 12k // 10k pot).

b) For that particular layout, it is actually better to use the through hole version of the XFD, as seen here :
Current Drive for Headphones & The Super Linear Transconductance Amplifier
Just leave the JFET buffer unpopulated, or cut it off altogether, as I did.

But glad you like this one too, amongst all the others.


Patrick
 
In the link you posted for the F5-HA, R4,8 are 5.5k, as per the original design.
This was changed later in the passive version to 12k // 10k (of the pot) in the passive version.
The SMD PCB assumes you are using a pot, and hence 12k was specified.
If you are not using a pot, you should use 5.49k instead.
I'll let you find out the difference in Spice between the two.

The through hole version gives you a wider choice of resistors and capacitor types.


Patrick
 
After a while I’m listening to the Xen version of the JLH headamp with my LCD2s. It sure feels like Zack de la Rocha is spitting to my face and I can here every detail of Tim Commerford’s bass while listening to Rage Against The Machine - Evil Empire. Hell of a headamp!

I attached a little family portrait of my rig in a brand new AV rack. In bottom row on the left is the guilty one this time. :hphones:
 

Attachments

  • E55AC46D-0EC8-4328-A971-AAB0EED9E44A.jpeg
    E55AC46D-0EC8-4328-A971-AAB0EED9E44A.jpeg
    712.1 KB · Views: 853
Passive Danyuk Crossfeed

Someone wants to build the JLH HPA with our passive Danyuk crossfeed.
Since we stop all GB activities because of Corona, we cannot provide PCB.
As a compromise, here is a layout that you can easily build with standard Vero boards.

Component values :
R1 = 11k
R2 = 5.6k when used alone; 12k when in parallel with a 10k pot.
R3 = 100k
R4 = 18k
R5 = 27k
C1 = 2.2n
C4 = 22n

We recommend Vishay Dale RN55 and WIMA FKP2 or MKP2.
Blue colour indicates jumper.


Patrick

.
 

Attachments

  • XEN Danyuk Passive Crossfeed Vero.png
    XEN Danyuk Passive Crossfeed Vero.png
    47.4 KB · Views: 565