diyAudio

diyAudio (https://www.diyaudio.com/forums/index.php)
-   Analog Line Level (https://www.diyaudio.com/forums/analog-line-level/)
-   -   Building UCD 250LP input buffer (https://www.diyaudio.com/forums/analog-line-level/323197-building-ucd-250lp-input-buffer.html)

kjb 27th May 2018 01:48 PM

Building UCD 250LP input buffer
 
4 Attachment(s)
Hi there!
Im trying to build a input buffer for UCD250LP.

This is my first attempt to design a PCB, I need to get feedback.
Its based on the recommendations in the ucd 250lp datasheet.
Anyone care to take a look at it and see if I'm doing some really bad layout/design? Please :)

I guess there is a lot that can be done better. I know it looks like s**t compared to other pcbs I have seen.
I can share the KICAD files if anyone is interested.

Schema:
Attachment 683430

Top layer:
Attachment 683429

Bottom layer:
Attachment 683431

Pin assignment
Attachment 683433

My plan is to do input buffers for the ucd400oem also.

kjb 28th May 2018 02:48 PM

Maybe I should specify some details.

This is for the Hypex UCD250LP oem module that sometimes can be bought from ebay and second hand. I bougt my 4 modules here at diyaudio (Swap meet). This pcb should work with the other hypex low profile modules as well I guess.

Some questions that comes to mind:

1. The op amp used is the LM4562NA DIP,
Is it better to use SMD version of lm4562 (any performance gains) or perhaps use any other op amp? OPA1612, OP2134?

2. Is 100nF + 47uF good values for the rails close to the LM 4562 in my design?

3. What kind of smd capacitors is good for this application, COG ?

4. Im going to hand solder the pcb. I have selected the 1206 size for all smd components, is it better (performance) to use smaller parts?

5. Is there ways to improve the layout regards to noise and performance?

Cheers!

kjb 31st May 2018 02:57 PM

Any thoughts at all? Would be great to know if I'm doing things wrong before ordering the pcbs. Its a long shipping time so I rather don't redo them if possible. Sure I will try to tidy some thing a bit before ordering.
I guess its a bit boring project because they are for oem amps that is not easily available, but I'm dong this for my private project, not commercial.

And Have just finished a pair of Lx521 open baffle speakers that needs a lot of amplifiers. So its kind of urgent :) hehe.

thizzgiant 31st May 2018 09:42 PM

Quote:

Originally Posted by kjb (https://www.diyaudio.com/forums/analog-line-level/323197-building-ucd-250lp-input-buffer-post5449529.html#post5449529)
Any thoughts at all? Would be great to know if I'm doing things wrong before ordering the pcbs. Its a long shipping time so I rather don't redo them if possible. Sure I will try to tidy some thing a bit before ordering.
I guess its a bit boring project because they are for oem amps that is not easily available, but I'm dong this for my private project, not commercial.

And Have just finished a pair of Lx521 open baffle speakers that needs a lot of amplifiers. So its kind of urgent :) hehe.

Overall the design looks fine. I've made a few different UCD400 oem boards over the last year and in one I used the same basic buffer design as you're using and it works great.

My only worry on your boards is if you've thought enough about how the UCD250LP will connect into the board, how the amplifier will then be heatsinked and with the amplifier connected are you going to be covering any of your other connectors. Those were issues I had and without knowing exactly what your plans are there might be a few "physical" things that could be better thought out. All in all looks pretty good for your first pcb.

kjb 1st June 2018 09:33 AM

Thanks for the response! The big connectors must be placed on the back of the board, away from the ucd. And the 18p connector and the rest of the components on the top side. Towards the ucd250lp. I will use a stright non angled connector to the ucd.
There might be a problem to keep the board in place only hold in place by the ucd connector.

kjb 4th June 2018 08:04 PM

4 Attachment(s)
Made a new version,
*Faston connectors for out + and -
*smaller 805 size on some components
Pin header on the back side, same as the GND pane.
*Other small fixes.

Attachment 684997

Attachment 684998

And learned there was a 3D viewer in KiCad :)

Attachment 684999

Attachment 685000

Nisbeth 4th June 2018 08:43 PM

A couple of suggestions:

- Move C4 so it is parallel to R7 - then you can shorten the supply trace to the opamp.
- The routing on pins 1 + 2 on J6 could be shortened considerably if you route part of it in the ground plane layer (which shouldn't matter). You can exit the pin 2 trace on the correct side of J6 etc.
- If you swap R5 and R6 you don't have to route a trace under R6.
- I would think you can clean up the routing around R1-R4 (and the two caps) by experimenting with turning the components at right angles to each other. You want to go for shorter traces, not the best-looking component placement :-)
- Etc.

And in general I think your traces look really thin, especially the ones than run through the connector pins. I would do what I could to get rid of those :)

kjb 5th June 2018 11:09 AM

1 Attachment(s)
Thanks for the feedback!
Since the buffer is differential/balanced I'd like to get the non-inverted and the inverted traces and components close to each other and in parallel as much as possible. I tried to look at how Bruno made his pre-amp input section.
But I guess It's room for improvement for sure. I'll try to put the components closer to each other for starters.
Picture from Brunos preamp: https://www.hypex.nl/img/upload/doc/...The_G_word.pdf
Attachment 685084

I'ts hard to route the traces to the dual pin header without going between the pins and still keep the traces close to each other. So the traces gets really thin.. I don't really know how to solve that one..

However. I'll be back with another version.

kjb 5th June 2018 09:52 PM

4 Attachment(s)
New version with some improvements. I hope :)
Including a resistor from + 15V supply to pull up the dc protection pin.

Attachment 685175

Attachment 685178

Attachment 685177

Attachment 685176

rmaudio 7th June 2018 10:07 AM

Quote:

Originally Posted by kjb (https://www.diyaudio.com/forums/analog-line-level/323197-building-ucd-250lp-input-buffer-post5446811.html#post5446811)
Maybe I should specify some details.

This is for the Hypex UCD250LP oem module that sometimes can be bought from ebay and second hand. I bougt my 4 modules here at diyaudio (Swap meet). This pcb should work with the other hypex low profile modules as well I guess.

Some questions that comes to mind:

1. The op amp used is the LM4562NA DIP,
Is it better to use SMD version of lm4562 (any performance gains) or perhaps use any other op amp? OPA1612, OP2134?

2. Is 100nF + 47uF good values for the rails close to the LM 4562 in my design?

3. What kind of smd capacitors is good for this application, COG ?

4. Im going to hand solder the pcb. I have selected the 1206 size for all smd components, is it better (performance) to use smaller parts?

5. Is there ways to improve the layout regards to noise and performance?

Cheers!

LM4562 is about as good as it gets overall. Assuming you want 'wire with gain' and not 'colour'.
SMD device itself offers no advantage for the opamp itself - same just diffrent packaging. It does possibly allow a better layout - esp more continuous 0V 'plane' ; smaller loop areas - but it's marginal. DIP part allows use of socket.
Technically a socket distances the device from ground plane / decoupling etc and increases risk of connection failure. But really not an issue.
The big advantage is , of course, ease of switching op amps. If only to show how good yours is :)

Decoupling looks fine.
COG / NP0 ceramic caps - yes. Most stable and negligible voltage coefficient.

1206 components. No problem. Smaller components - I go down to 0603 - allow marginally tighter geometries.
But against that I think it's true (as in I've read it somewhere reputable - not measured it myself and not just someone saying it on an internet forum:rolleyes: ) that larger resistors exhibit less noise - ie noise over and above the theoretical minimum noise.
On resistors - use Metal Film (MELF) / Thin Film types. Avoid Thick Film types. Linearity.


All times are GMT. The time now is 12:36 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 17.65%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio

Wiki