A TREAD sized regulator - the r1
Apr 19, 2011 at 9:02 AM Post #106 of 189
I received prototype boards yesterday and in a fit of "I can do that other stuff later" built a 5V output one.
 
I would say that the build is straight-forward and easy to complete.  I didn't get the sense while assembling the regulator that it was a tight build, but when you finish and look at it you realize there is a fair amount of parts in a small space.  You definitely want to take your time to ensure that you get the parts down in the right place the first time, as it is challenging to change parts on the board afterwards. 
 
And change parts I did.  There is always a slight issue when you use one part footprint (FET) for another (depletion-mode MOSFET). That means that for the prototype boards Q2, the LND150, needs to be installed turned around 180 degrees compared to the footprint on the board.
 
As for performance, all I got to do last nite was having a working reg that output 4.99V as shown on my meter as I needed to go to sleep.  I did see that the start-up behavior was that, for the 5V version, it quickly started to ~4V and then over about 5 seconds the voltage came up and stabilized at the 4.99V final output voltage.  This is expected as the low pass filter on the reference voltage charges.
 
So what now?  Well, I want to personally build two more regulators, one each with the other reference options, to ensure that any prototyper gets a functional board when they build from the BOM.  I may get one done this week, but I'm not sure about having both done.  After I get the two built, I figure they'll be ready for the rest of you.
 
Apr 19, 2011 at 9:52 AM Post #107 of 189
This looks like a nice little regulator! It's small enough to be built inside a large wallwart with a trafo for small power applications, should be perfect for a Y2 or small desktop amp.

Should be looking for a few of these boards when available.

A kit from beezar or glassjaraudio would be a great thing!!
 
Apr 19, 2011 at 10:42 PM Post #108 of 189
Some pics of the built unit:
 

 

 
And yes, before someone says something, the 1000uF filter cap is not a Panasonic.  Turns out that when I went to build, it seems that I forgot to order that part, so I pulled something out of my parts box.  And these pictures were taken before I turned Q2 around.
 
Apr 19, 2011 at 11:27 PM Post #109 of 189
You were not kidding about it looking tight on the board. 
 
Was max voltage and current discussed for this.  I thought I read it somewhere but can't seem to find it now.  It seems a couple of these could be used to make a nice multiple voltage bench power supply.
 
Apr 20, 2011 at 7:56 AM Post #113 of 189
Thanks for the compliments.  Tangent should get the original kudos for the size of the TREAD - I just tried to shove a bunch of parts in the same size 
biggrin.gif

 
Choosing some boards that a whole bunch of people have built as reference, the grubDAC/SkeletonDAC/BantamDAC are ~51x27mm rounding up.  The r1 is 53mmx33mm or for the less metrically inclined ~2"x1.25".  So yes it is tiny.  It is not the tiniest thing I've built, but for some reason everything I design is small.
 
Apr 20, 2011 at 8:37 AM Post #115 of 189


Quote:
Was max voltage and current discussed for this.  I thought I read it somewhere but can't seem to find it now.  It seems a couple of these could be used to make a nice multiple voltage bench power supply.


I may have answered this, but I'm going to again, as I can't bother be bothered looking back through the thread.
 
First to answer this in terms of the schematic and the PCB.  Looking at it generically, the limiting factors on the regulator are:
  1. The dissipation in R3, but this can be jumpered so it really is not an issue
  2. The voltage rating of C3 versus the capacitance rating to provide a good ripple filter.
  3. The value of R4 which acts as a current limit to the base of the pass transistor
  4. The dissipation (and therefore heatsinking) of Q1, the pass transistor
  5. The voltage rating of the other parts (like the op-amp and the capacitors0
  6. The current capacity of the traces - if you are really trying to push the boundaries.
 
The critical factors that I see are the dissipation of Q1 and the ratings on C3.  Q1 can be heatsinked off board, so that is not so much of an issue as C3.  Higher current draw requires higher capacitance which also requires a larger footprint for a given voltage rating.  "A given voltage rating" is critical - if you feed the board AC, you want a good safety margin, but if you were to feed the board DC, you can get away with a lot higher input voltage with the same parts.  With a very quick check at Digikey, I'd say if you want to stay within the footprints of the board, 33-34V is about the maximum output voltage of the PCB.
 
Now to answer this in regards to the BOM that I have posted.  Capacitors on the output side are rated for 25V, so I set the maximum on the output to be 24V.  I've also set R4 so that the approximate maximum output current is 250ma. This limit was set completely arbitrarily, so you could push up the current output limit as required.  While some may say having a current limit in the supply is a bad thing, if you've ever had your supply go into limit due to a build mistake, you can appreciate its value.
 
Apr 20, 2011 at 8:53 PM Post #119 of 189


Quote:
No issue.  It was only put over the board to prove that it would fit.



Although it technically fits it's probably not a reasonable build option since there is a good chance of direct thermal contact to the capacitors on either side.   This could be rectified in the production version by moving R3 down, turning it 90 degrees clockwise and moving C3 as close to the board edge as possible.  This should gain you at least 3mm that can be used to provide reasonable space on either side of the heat sink.
 
Apr 20, 2011 at 9:25 PM Post #120 of 189
Nice!! It will be a perfect fit for many of my projects!!!
 
 

Users who are viewing this thread

Back
Top