Jung regulator layout
Sep 23, 2004 at 9:49 PM Thread Starter Post #1 of 8

00940

Headphoneus Supremus
Joined
Nov 6, 2002
Posts
4,493
Likes
47
jung.jpg


schematic is based on the one disponible on AOS website :
jung_pos.gif


The groundplane only connect the "normal" ground as explained on tangent's website. Pads for ground sense-vreg sense. Smd resistors are 0805.

Any obvious mistakes ? Any ideas on how to improve on that first try ? I know I could simply get Aos's boards since they're cheap but I'd prefer to implement those directly on the dac pcb.

And to get 5V, I suppose I've no choice but to use LM336 ?
 
Sep 23, 2004 at 10:36 PM Post #2 of 8
Where is that 'top' trace at the bottom going to? Otherwise, looks fine. Definately have to prototype it first though, it's impossible to tell if it will work properly just looking at it.
 
Sep 23, 2004 at 10:40 PM Post #3 of 8
oh, just leftover of the V+ regulated. Should be replaced by a pad. Still missing a ground in pad.

Yep, it should definitly be tested as stand alone unit before being put on any bigger board
 
Sep 24, 2004 at 6:28 AM Post #4 of 8
Quote:

And to get 5V, I suppose I've no choice but to use LM336 ?


There are lots of ~1.25V and ~2.5V references out there to choose from. Do a Digi-Key search for "reference" and be amazed at the choices.
 
Sep 24, 2004 at 10:49 AM Post #6 of 8
The first step is to eliminate the obviously wrong choices. Eliminate the ones that aren't the right voltage. Then drop all the ones that aren't in the right package. All the surface-mount devices could probably be removed from consideration, for instance. You might even tell it you'll only consider devices in, say, TO-92 packages. Another thing that differentiates references is their current requirements. There are a lot of low-power references that you can ignore because efficiency is soooo not the point with a Jung regulator. That should reduce the number of choices dramatically.

Beyond that, references vary mainly in their stability parameters. Tempco, stability over time, voltage coefficient, etc. Just like with passives, you simply have to decide how much accuracy you are willing to pay for. The generic LM385 will run you about 60 cents. Some high-spec references will run several dollars. Where do you want to be on that curve?
 
Sep 24, 2004 at 11:07 AM Post #7 of 8
thanks, I'll do my homeworks. TO92 is definitely a plus (would it only be for the facility of reworking the layout). Something around 2$ seems reasonable. The canonical LM329 runs at 1.5$.
 
Sep 28, 2004 at 2:11 PM Post #8 of 8
A new part from TI might be useful:

Dual Op Amp with Integrated 2.5-V Reference: The TL103W and TL103WA combine the building blocks of a dual op amp and a fixed voltage reference, both of which are used in the control circuitry of switch-mode and linear power supplies.
For the A-grade, especially tight voltage regulation can be achieved through low offset voltages for both op amps (typically 0.5 mV) and tight tolerances for the voltage reference (0.4% at 25ºC and 0.8% over operating temperature range).
Characterized for operation from -40ºC to 105ºC
Applications: Battery charger and data acquisition systems
TL103W: Price (1 ku) $0.29
TL103WA: Price (1 ku) $0.32
 

Users who are viewing this thread

Back
Top