RN3 power supply upgrade details.
So here is the write up of my latest science experiment and bit of Audiophoolishness…
I removed the stock SMPS power supply and replaced it with a collection of parts to make up a LPS of suitable capacity and capability.
It is overkill to a certain extent but then that is a usual and expected trait for these sorts of things.
The power supply I added to the RN3 consists of a Telema 25VA transformer
http://www.ebay.com/itm/131363018720
mounted to a circuit board
http://www.ebay.com/itm/131863613806
feeding an LT3042 ‘precision’ +5Vdc regulator board
http://www.ebay.com/itm/142115498378
All of which feeds directly (hardwired) to those pair of wires I previously soldered to the main board.
What follows isn’t a ‘strict’ electronic analysis but is meant for clarity and ‘worst case’ figur’n, which will suffice to make the main point.
The transformer I used has dual 9Vac output coils and together are rated at 25VA/9V = 2.77 Amps.
And for our purposes of supplying +5Vdc to power this RN3
2.77Amps x 5Vdc = 13.88 Watts of available power.
The RN3 is rated to use 30 watts but I measured its current draw at ≈1amp actual power usage, which means,
1Amp x 5Vdc = 5Watts.
So this PSU supplies 13.88Watts and we use 5Watts, which is ≈ 36% of the rated capacity of this PSU as configured.
Which isn’t an ‘ideal’ percentage of this PSU, if all I cared about is % efficiency, but is perfectly fine since I have ≈ twice again the amount of current delivery capability in reserve.
IOW this amount of draw is sufficient to fully ‘engage’ the regulator board (so it has a sufficient load to regulate) but it, relatively speaking, is ‘loafing along’.
Which means it generates low heat, more that the SMPS to be sure but even if the entire PSU were only 50% efficient (unlikely) it would only generate ≈5Watts of heat, and it would still have almost 3 watts of reserve power remaining.
So in my original figur’n of how I’d layout this modification I was going to mount the transformer in the vacant area on the left and the regulator board near the connector block I had already added
Like this…
But I verified that the transformer board would fit in the space vacated by the stock SMPS.
Since I wanted to use the on/off switch and the existing IEC power connector, this arrangement would make the AC power routing much simpler, and shorter.
Be aware that the available space is a bit tight since the machine screw ’Zerk’ type fittings used to secure the sides of the top cover, protrude into the available mounting space, which in turn forces the board closer to the mainboard,
It’s fortunate that using the standoffs lifts the board above these ‘Zerk’ fittings, even so the ac power wires are positioned in this space so the extra room is a ‘good thing’.
And so I removed that connector block and hard wired the 2) +5Vdc wires directly to the regulator board with its 2) +5Vdc output pins, and I also hardwired the ground wire as well.
Wiring the Telema transformer board was the most tedious part in that I tried 2 different approaches with the 2nd technique being FAR easier.
So for those who would venture down this road I’ll save you the trouble of doing this the hard way and just explain what worked best for me.
The 115Vac inputs (primary windings) and 9Vac outputs (secondary windings) use 2 coils each.
They must be wired together in order to use the full output of this transformer.
Essentially the 2 coils are wired in parallel for both the input (primary windings) and output (secondary windings).
The input solder pads are a pair of ’square’ solder pads, and a pair of oval solder pads that are ’staggered’, which means the interconnection between them has to jump over one pad to get to the other mated solder pad.
They look like this,
O-X-O-X
where the O’s are wired together, as are the X’s.
What worked best was to divide the 16gauge ac power wire I used into two equal sized wires, into a ‘Y’ configuration.
These ‘Y’s then connect to both of the X’s, and to both of the O’s, where the X would be the ‘hot’ lead (Red wire) and the O would be the neutral lead (black wire) from the front panel switch.
And to possibly add just a touch of confusion, the red and black wires could be swapped between the X/O connection pairs since the load is a coil and it doesn’t ‘know’ ‘hot from neutral, nor care.
And this applies to both the primary and secondary windings.
IOW as long as both of the X’s and both of the O’s are used each by a ‘Y’ connection, the ‘polarity’ in not an issue.
I used 3 pieces of heat shrink to make the ‘Y’ so they both were insulated from each other and connected the hot on the top of the board and the neutral to the bottom.
I used this technique for the outputs, but the output coils are not ‘hot’/‘neutral’ but they are ‘polarized’, so the square/oval solder pads remain, which need to be wired in the same way, with the square pads (X’s) together and the oval pads (O’s) together.
That is if you are using 115Vac as the source voltage (like in the US).
If you need to wire the primary for 230-240Vac, the input wiring is much simpler, in that you bridge the middle X and O together and send the hot to the X and the neutral to the O at opposite ends of this connection, like this;
O-XO-X
This wires the input coils in series. (Sorry no Picture Available).
I also added a 5x20mm fuse holder ($2) and wired the Brown wire (‘hot’) directly from the IEC connector thru the fuse, and then along with the Blue wire (‘neutral’) sent them on to the front panel switch, unchanged.
And I left the IEC connector’s ground wire alone.
The output from the on/off switch went directly to the input side of the transformer board.
And the 9Vac output from the transformer board went directly to the regulator board’s input.
Mounting the boards was fairly straight forward in that I used 1/4” standoffs to raise the boards above the chassis.
This mounting scheme necessitated that I use nuts on the outside of the chassis which do stick down further than the tiny stick on feet, which in turn necessitated the use of more robust and ‘taller ‘ feet.
But other than the new nuts on the bottom of the chassis and the bigger feet, there is no other clue any modifications have been made to this RN3.
Truly a ‘sano’ upgrade.
And after things play for a while, settle in and I hear what the results of all of this fussing yields,
I’ll take some measurements of the power supply noise,
and the current draw in several places,
and take a few thermal readings,
and take a few more pics…
But in the mean time my 2 channel PSU will be built and perhaps even wired to the Mutec 3+, and I have an ac power distribution project to install as well.
Just WAY to much fun.
JJ
