ROON ROCK with a linear power supply?

This issue has been discussed to some degree but wonder now the dust has settled and we had some time with ROON ROCK and our intel NUCs whether some have ventured forth and swapped out the SMPS for a linear supply and, if and by what manufacturer, did you do so?
Any sonic benefits that you’d care to share?
I know this topic is divisive so no need for posts stating categorically that it will have NO influence when we’re piping the data through ethernet. That one is understood but there are still some mysteries in audio.

I use a couple of 12v lead acid gel batteries and a trickle charger. It works really well when using the NUC connected direct to a DAC.
NP7 12L or equivalent X2 and a suitable trickle charger up to 20ah or thereabouts. Very cheap and very effective.

I thought the NUC needs 19V … I would say a 12V and a 6V battery would be a closer match unless you are running another level of regulation.

A NUC can run on anything between 12v and 19v using the external power connector and 12v to 24v using the internal connector. The information is there on the Intel Ark pages. What matters is that if, for instance, your board specifies 36 watts, the PSU can provide that regardless of voltage within the minimum and maximum.

Most NUCs are rated 19V, if a lower voltage is used the current will go up, if the power adapter is not designed to deliver the increased in current it will overheat and possible damage. It is advisable the rated voltage(19V) should not be more than 4V down, this is to avoid large current being drawn. An amp meter is a good tool to monitor the current. Note the current must be measured at full load, i.e full CPU utilisation.

You have to meet a minimum power requirement within the allowable voltage range. That means current varies as required. Can you reference your statement suggesting there may be damage caused please. If you go to for the 7i7BNH under essentials it states range of voltage inputs accepted as 12v to 19v. Clearly any PSU used needs to meet the power requirements of the NUC. Get that power rating right and the current draw will be fine.

I’ve been running mine with this:

Did you notice any sonic improvements when doing so?


Even when power requirements are met that doesn’t mean it is safe to use. For example, power adapters rated at 19V @2.5A gives a maximum power of 47.5W but if the voltage is down to 12V, the current will have to go up beyond @2.5A capable, in order to maintain the same power envelope. If the power adapter is not designed to operate beyond the 2.5A rating, it will either current limit and cause the voltage to down further, heat up or shutdown. Voltage and current matching to the NUC is strongly recommended.

We will have to agree to disagree then Guy. I’ll take my lead from Intel, they specify 12 to 19 volts. I’ll bow to their knowledge and continue to run my NUCs as they are, from 12v batteries when connected direct to the DAC. I do think what you are saying is alarmist and I do wonder if it is being said from a genuine safety view point or from that of a PSU sceptic putting an alternative spin on things. If you do want to dispense safety information you might want to brush up on your Ohms law.

There’s no need to use Ohm’s law here, it’s just a simple power formula, I x V. Say, if your power adaptor is 19V x 2.5A gives 47.5W, in order to give same power rating of 47.5W at 12V, you need 3.95A.

On that we agree. :slightly_smiling_face:

Yes, though to be fair, I made a couple other tweaks at the same time.

Personally, unless I could definitely hear a sonic improvement, I would not sacrifice efficiency–power bills, carbon footprint, etc.–in an always on server. A front end computer or player that I turn off when not listening, different story.

Aren’t most cheap trickle chargers switchers?

[quote=“Leporello, post:14, topic:29838”]
Aren’t most cheap trickle chargers switchers?

Yes and no. They actually sense the condition of the battery and are not in the purest sense SMPS’s. However despite the fact any noise they may generate is lost in the huge reservoir of lead acid batteries, you can just disconnect them and let the battery do its thing if you wish. In my use case it makes no difference but I have a clean supply running my USB-SPDIF converter.

Any consensus on this?

To clarify, my DAC is NOT connected to my NUC (running ROCK - 7th gen i5 BNH).

So NUC to router (Verizon FIOS G1100 - Quantum Gateway router / modem combo) via Ethernet. Router to streamer (SOtM SMs-200) via streamer.

Would I notice SQ improvement when / if powering my NUC (w/ ROCK in it) with a Linear Power Supply? How about doing the same thing on my Router?


The only mileage is in powering whatever is connected to your DAC with a linear supply. I have run my ROCK NUC direct into a DAC and via the network. Via the network I noticed no improvement but thought there was when direct into my DAC. But that isn’t consensus, that is just my experience! I presently have my router that my endpoint plugs into running off an LPS-1 and that is slightly less clear but perhaps that is because I use power plugs to get Ethernet upstairs and they do mess up the sound. For me the router does a good job of isolating some of the crap but it does that in its own right, I can’t say with any conviction that the LPS-1 helps. My end point is a single core low powered NUC running ROCK into a Hydra Z and ZPM so isolation is exceptional to begin with.

1 Like

This is correct, but it doesn’t follow that lowering the voltage through a constant resistance results in an increased current. The opposite is the case (I = V/R).

It does if the power requirement is the constant. In that respect he was correct and I was wrong because I said it was ohms law which is voltage, current and resistance and I meant voltage, current and power. If a NUC consumes 15 watts then a lower voltage requires a higher current for the same power consumption. A computer is a complex active component so it’s resistance is not a constant.

It might “require” a higher current to reach a particular power envelope, but it’s not going to get it by lowering voltage unless the resistance is lowered by a greater proportional amount. That doesn’t happen. The resistance of a computer might fluctuate thermally, but it doesn’t magically lower enough to result in increased current when a lower voltage is applied. If we could do that trick then our energy concerns would be over.