12V 60W PSU

badwhitevision

badwhitevision

Forerunner
This might go either here or in the computer hardware forum, but since a 12v adapter is commonly used for a lot of things, I am posting it here.

This might be a first, a whole dicussion for a simple 12V adapter, but I overthink things and I am thinking a lot about this one to the point that I am obsessing over it.

The requirement:
I need a 12V 5A adapter to charge a 2S Li-ion Battery Pack that acts as a UPS to run an external enclosure for a 3.5 inch HDD.

Why 5A? The start-up current of the HDD is around 2.85 A and it then idles at 0.85A with writes/reads when torture testing going up to 1.2A. Now the adapter supplied with the external enclosure is rated for 1.5A but is able to provide the start up current. This adapter is at it's max (and above it with the startup current) only with powering the drive, hence asking for a 12V 5A adapter makes sure I have a lot of headroom for the charger too.

The options:

A standard chinese black plastic adapter. 2 pin AC input and a barrel jack.

A "server" SMPS (the ones that come in a metal casing) with 3 pin AC inputs and multiple DC outputs (all of the screw type variety).

The concerns:

1. The HDD enclosure has the ground for 5V and 12V tied. This means that any noise/ripple on the 12V input is going to find it's way into the computer.

Why should I care about the noise/ripple?

Good question, and usually I don't. But in this specific instance, the computer also has a USB SDR (Software defined radio) attached. This means that such noise is going to bring up the noise floor of the entire spectrum and makes the SDR unusable.

The solutions:

1. Isolate the 5V input to the SDR via a LC choke and hope it deals with the problem.

2. Get an 12V PSU that ensures that no noise/very less noise reaches the USB input.

I am looking to solve this via number 2.

To this end, my basic understanding of electronics tells me that plastic adapters are no good here, since there is no earthing and hence any noise is directly sent to the output. (But contrary to this limited knowledge of mine, an old Pace adapter that I had actually had a low noise floor).

I would like to buy the metal ones, but they seem to be in a range of prices and the ones from Meanwell and the like seem to be ₹1000+ for a 60W PSU. (On the low end, there are ones that are ₹300)

Similarly, the black plastic adapters also are in a range of prices. I already have an old Pace 60W adapter (which is a 2 pin plastic adpater) and it was one of the better ones with a low noise floor. However, it had to be relegated to other duties, hence the need for a new one.

I know I sound very confused. But basically I am looking to buy a bang for the buck low noise/ripple 12V 60W PSU and am confused with whether I should buy an "adapter" or "Server".

Also, I am not a major in electronics/electrical/anything. This is a hobby of mine and while over the years, I have gained some little knowledge trawling the internet, I have no formal knowledge of the subject. So I might have made mistakes, called out terms wrong etc.

Calling on people with more experience to help.

@napsterquest is someone I have seen dealing with low voltage electronics, so tagging you here. Apologies if you are not the right person.



Thanks,
badwhitevision
 
The pragmatic option would be to use a laptop power adapter — they're high quality, cheap and easily available, both new and used. And then step down that voltage using low noise buck converters, to 12v and 5v. Robojax on youtube has a collection of videos of commonly available modules, you'll need to go through to see which ones are low noise:

www.youtube.com

Robojax Reviewed Buck Converters (Step down converter)

I am presenting all step down or Buck converters modules I have reviewed so you decide to which item is better for your applicaiton. I have no affiliation an...
www.youtube.com www.youtube.com

Alternatively you could get a USB isolator, they're meant to filter out electrical noise/ripple for sensitive equipment like your software defined radio module. Here's an example of one: https://evelta.com/industrial-usb-isolator/
 
You might be overthinking this, usually the max current consumption on typical 3.5inch HDD is 1.5A, just to be safe you can go with a power supply of 2A, this is most of the time the current capacity of wall wart adapters for a external 3.5inch HDD drive.

So your assumption that your HDD while starting absolutely needs 2.8A is wrong, any motor starting current will be significantly higher that it's normal operating current and it varies a lot with what voltage you apply to it.

Your HDD will work completely fine with 2A power supply, also giving you some headroom.

badwhitevision said:
need a 12V 5A adapter to charge a 2S Li-ion Battery Pack
Click to expand...
Are you sure about that? You should typically charge your batteries at 0.5C, which means if your single cell is rated for 2500mAh, you charge it at 2500 x 0.5C = 1250mAh = 1.25A.

I suspect your UPS will be charging your 2S pack in series, so it should not be taking a lot of current to charge them.

Did you DIY your UPS? If not then it should be directly supplying power to the AC adapter of your HDD wall wart adapter, and in case of power cut it should immediately switch to battery power.

Or your UPS just takes DC power only? Need more explanations.
 
  • Like
Reactions: rsaeon and StygianClaw
Seconding Heisen on the overthinking part.

Electronically, computers are a noisy place. An external 12V power supply, however bad, is unlikely to make things worse. Unless you notice a difference in output quality, there is no need to worry.

A note about using a higher rated power supply. Most cheap power supply have poor load regulation. The rating of 12V is at full load. If you are working at 10%, the output voltage could be much higher. Just a thing to keep in mind when buying one.
 
Last edited:
  • Like
Reactions: greenhorn
Achieving low noise is expensive. What rsaeon posted are the cheaper options. 1k for a mean well is among the cheaper options i can see.

The cheap USB isolators are usb 2.0 and the 3.0 ones are way too expensive.

The other thing is that, if something is designed to run on computer usb 5v supply, it should be pretty noise tolerant. The noise on a computer 5v supply is usually far worse compared to the noise a single hdd can inject, so you may be over thinking the issue here. This is unless it's already known that the SDR dongle is very noise sensitive
 
I would not agree on the plastic vs metal thing, thats a common misconception and something that aims to 'make' something premium. Prime example in this in consumerism is Apple with metal (now glass) vs plastic. Look at spec sheets if they share it, else see the components they use, these will give a better idea. Another option is to see if you can get hold of an oscilloscope and check the output yourself, and add some extra filtering if you want

Secondly, the Li-Ion UPS, will also introduce noise since it will be using buck boost or something similar to give you 12V. Am assuming the UPS is a direct 12V output which will feed the HDD.

Thirdly, the HDD itself will may have some noise. May it be better to move to SSD? Or use something like RAM storage and then dump to HDD?

Alternatively, leave all this and use USB isolators on both the SDR and the HDD too if you want double protection.
 
  • Like
Reactions: nRiTeCh
Thank you everyone for the input and response.

Like @Heisen asked, here are more details. It was stupid of me to assume that y'all would know the exact specifics of my project without providing them :/

The objective here is to make sure my "homelab" continues to work, even when there is a power outage.

The "homelab" is a laptop, so it has it's own battery which ensures that it works even with AC power cut off.

To this laptop, I have an SDR and an external HDD attached. SSDs/2.5inch HDDs are not possible since this contains my media library which runs into 4-5TB, and I don't see myself being able to spend on SSDs while with 2.5inch HDDs, the same issue creeps up since the laptop's 5V USB will not be able to power so many of them on it's own and will necessitate powered USB hubs, which will then have to be supplied by UPS battery packs and we are again at the first step. And another thing, I can't bring myself to spend 12-13K on 5TB HDDs for just my media library. Important documents/photos/memories that I shouldn't lose are stored in a SATA SSD inside the laptop.

As for the SDR, I use it to listen in on VHF/UHF HAM band conversations. It is itself a cheap chinese DVB-T RTL-SDR dongle, and so doesn't have much "shielding"/protection/noise tolerance and such. Or as @greenhorn stated, the noise floor would have been much higher, if not for the fact that the design allows for noise tolerances.

Like @dpandey pointed out, computers are inherently noisy and the same is reflected on the noise floor in applications, but when I add this external HDD enclosure, the noise level increases. I say this with much conviction because the noise floor is at -80 when the enclosure isn't connected, versus at -60 when it is connected. This leads me to believe that it is the enclosure (or rather the adapter powering the enclosure) that is causing this noise.

The UPS is entirely DIY.

Specs are as follows - 2S2P Li-ion 18650 pack with a HX-2AS BMS (4A max) with a total capacity of 4000mAh.

The pack is charged by a TP5100 module with charging current set to 2A (0.5C)

The UPS determines power failure and switches over to the battery pack with the help of a MOSFET (IRF540N).

The 12V input is fed via a diode to a boost converter (XL6009E) which provides the final stable 12V output.

It is also fed to the MOSFET and the TP5100 module.

To smooth the transition between pack and mains, a 2000uF cap sits at the output of the boost converter.

You might have seen this exact schematic somewhere else because I have beasically copied this Instructable (https://www.instructables.com/DIY-Mini-UPS-for-WiFi-Router-V50/) replacing the 3S pack with a 2S one and instead of the buck converter (to charge the 3S pack), the TP 5100 module.

This elaborate setup is to ensure that the battery pack is used only when neccesary instead of always powering the load with the pack. (I have read that this leads to problems with the charging setup and might very well end in a fire).
rsaeon said:
The pragmatic option would be to use a laptop power adapter — they're high quality, cheap and easily available, both new and used. And then step down that voltage using low noise buck converters, to 12v and 5v. Robojax on youtube has a collection of videos of commonly available modules, you'll need to go through to see which ones are low noise:

www.youtube.com

Robojax Reviewed Buck Converters (Step down converter)

I am presenting all step down or Buck converters modules I have reviewed so you decide to which item is better for your applicaiton. I have no affiliation an...
www.youtube.com www.youtube.com

Alternatively you could get a USB isolator, they're meant to filter out electrical noise/ripple for sensitive equipment like your software defined radio module. Here's an example of one: https://evelta.com/industrial-usb-isolator/
Click to expand...
@rsaeon Thank you for introducing me to Robojax. Somehow in all these years of trawling YT, I haven't found this gem.

Thank you for the tip about the laptop power adapter. I am actively considering one.

The USB isolator is very pricey and I am not inclined to spend that much.
vivek.krishnan said:
Secondly, the Li-Ion UPS, will also introduce noise since it will be using buck boost or something similar to give you 12V
Click to expand...
@vivek.krishnan you are right, using the XL6009, introduces noise into the system, to this end, I was advised to use a LC choke (100uH and 100nF cap) to reduce the noise.

The goal of this entire post was to ensure that I buy an adapter/PSU that doesn't supply any more noise than that is already being injected by the laptop itself (which exists even now) and the UPS' boost converter (which I hope to solve with the help of the LC choke).

Unfortunately, I do not have access to an oscilloscope to determine the exact noise frequencies and design the choke for that frequency. However, this is where @Robojax' channel comes into use as he does have a few videos on these buck/boost converters.

@Heisen My calculations for the 5A requirement were based on the 2A charging current of the charger. In normal use cases, the 5A limit will never be reached, since when the HDD is operating, it never consumes more than 1.2A and with the 2 A charging current required for a fully discharged battery, the maximum current used will be around 3.2A.

However, in an edge use case and one that I see becoming more common with the increasing duration of power cuts as the summer approaches is one where the battery is discharged, the HDD has spun down due to lack of power and once power returns, the charging adapter takes up 2A and the spin up current requires another 2.8A.

When you say the start up current is always higher, wouldn't this lead to activation of over-load protection of the SMPS if it is not rated for it? I know that this doesn't happen in reality, but I am wondering as to why. Is this because the higher current draw automatically pulls down the voltage and as a result the entire power draw remains the same?

In any case, the 5A adapter may prove useful in that if I were to add an extra disk tomorrow, I might be able to power both off the same adapter via a splitter.

@jai1611 Thank you for the Meanwell SMPS recommendations. They don't seem to be available locally and would have to be ordered online. I am more inclined towards rsaeon's suggestion of using the laptop adapter since it's readily available, but this is also something I am considering.

If it isn't obvious by now, I'm trying to keep this to as minimum cost as possible, while acheiving the highest quality that can be attained at that price point.

Once again, thank you for all the inputs and helping me here.
 
  • Like
Reactions: Heisen
The right solution for this application will be a linear power supply, which offers least interference, the Korad Ka3005d is a good option, but that won't come cheap.

If your use case demands 5A you have to get a switching power supply which offers even higher current rating.

But I think any switch mode power supply will introduce noise, specially coupled with your DIY solution.

Your budget is also very low. @rsaeon laptop power supply suggestion seems to be a good one, but you have to consider using a buck converter with it because most probably it will output 20V, and the buck converter will also add it's own noise.

I did find this 12V 5A LED power supply which you can use for you application.

The datasheet says it has 120mVp-p ripple/noise. But good thing is you won't have to use a buck converter with this one. It's meanwell power supply, has all the certifications, so datasheet can be trusted.

1712400453234.png


You have to decide between this and the laptop power supply. If you are planning to buy those metal casing switching power supply, make sure to get the mean well ones.
 
  • Like
Reactions: badwhitevision
Another cheap suggestion i could think of was to create a star ground point nearby ( i had steel windows in my house and i rubbed off a bit of paint and tied a wire around it. Emi will usually come thru supply rails. In your case the hdd uses a different supply.

That leaves the chance of a ground and a ground loop between the laptop power supply ground and the HDD power supply ground. The 100% guaranteed solution would have been a galvanic isolator, but since that's out of your budget, take all your grounds (I'd say hdd 0v, laptop chassis gnd and maybe the 0v of the port the dongle is connected to)

I'm not sure it will work, but it's cheap. I've done this to ground my turntable preamp which had too much him and this worked.

Also a bit risky. It assumes all these points are all 0v. May not be true. Ie your laptop power 0v and usb 0v may not be 0v relative to each other and attempting to connect them together may be explosive or expensive. Adding a capacitor in series mitigates a bit of that by blocking DC. Nothing much you can do if it's ac
 
  • Like
Reactions: badwhitevision
Update - Finished the build successfully. Used an old 12v 5A "plastic" supply that was in another build and did not face any drop in signal quality (I found this out when testing the UPS). Replaced the PSU in the other build with a local cheapo supply since noise didn't matter there.

Tried using the 100nF cap and 100uH inductor to reduce noise, but that only made it worse and not better. Overall, satisfied with the output.

Will try posting a build post for the UPS itself. Learnt a few things along the way.

Thank you all.
 
  • Like
  • Haha
Reactions: vivek.krishnan, MoviesMusicMayhem and rsaeon
Top Bottom