supplying multiple telecoms items from a single battery - earth-loops?

Status
Not open for further replies.
M

mab2

Member
Hi I have a small solar setup (24v battery) running a few lights & devices (including my laptop (19v)), and I would like to run the router (12v), the wireless phone base station (7.5v) off the battery too.

Do I need to have electrically isolated supplies for the router, telephone base and laptop to avoid earth loops? (as they are all wired to each other)? Or can I make a switching reg supply with +7.5v, +12v and +19v supplies on a common earth?

thx

mab2
 
I don't see a problem with earth or ground loops. Come off the 24 VDC using regulation to feed your loads. I don't see where ground loops would even figure nto the scheme?

Ron
 
It would be recommended to add a series resistance on the + lead (a few ohms) and as much capacitance in parallel as you can fit/afford for each device to help reduce transferred noise. You lose a little bit of power with the resistor (often negligible) but the combination of the resistor in series with as much parallel capacitance as you can fit will swallow high frequency switching noise from devices.
 
Last edited:
Yes, I strongly agree with Sceadwian! When I said just come off the 24 VDC with regulation I should have added using good regulation practices. For example don't be stingy with capacitors and as Sceadwian mentions toss in a few low resistance resistors. The DC from your source is about as pure as the driven show, but you don't want devices tossing noise into your party.

Ron
 
Thanks.

I know you have to to be careful about earth loops on audio systems but I wasn't sure if telecoms stuff was referenced to earth or not.

I was going to add plenty of low ESR capacitance. The laptop draws quite a bit of power though so I was going to use a series inductor rather than resistor.

mab2
 
mab2... It's not really ground loops with higher frequency circuits though the effect is similar. In higher frequency circuits where the feedback methods can be more complicated it's often called 'cross talk'

LC filters are great, but to get effective response you need very high inductance which aren't practical unless the frequency is quiet high, so an LC circuit will filter out switching noise from something like mhz level I/O switching noise from say a delta sigma DAC. Just a few ohms of resistance though will only drop the effective line regulation another couple percent but can effect noise filtering drastically.

You're not protecting the load from the power supply or the power supply from the load, you're trying to protect each load from every other load on the same power supply.
 
Hmm... I see - I think.

With the laptop drawing 1.5A at 19v I could get away with a few 1/10ths of an ohm, but I'm not sure about ohms.

Thanks

mab2
 
As long as you don't mind the size/weight/cost just pick a decent sized inductor. It may even be a non-issue, different makers use different pre-filtering methods for their devices, so it's really random chance what you'll end up with.
 
Thanks - I've got a few spare small mains transformers - I could use the secondary of one - that ought to be enough uH s

mab
 
Mains transformer cores would be good for low frequency noise but not necessarily high frequency the core isn't made of the right material high frequency can go right through them. Try using generic ferrite cores from the noise suppressors of monitors/keyboards/basically any digital device that's well designed will have a 'bump' right at the point where it's connector plugs in, I rip them out of anything I recycle into my bits bins.

The more turns you can get through the core the better. They eat high frequency noise for breakfast
 
Last edited:
The more I learn about inductance, filtering and RF circuits the less I seem to know

OK, I've got plenty of junk - including psu cables with ferrite suppressors - I'll use one of them - though I've never understood how they work with both supply & return going through them.

thx

mab
 
BTW, Ethernet is isolated by design. Each end is transformer coupled.

Telephone has a "protector" to ground from both sides of the line, otherwise it's mostly isolated. Protector failures give interesting results. You can find them on ebay sometimes.
 
Last edited:
Sceadwian said:
Why? A ferrite core doesn't care which way the electricity is going.
Click to expand...

Well I was thinking if you have supply and return going through the core together (carrying equal & opposite current) then the two cancel out - no net magnetic field should mean no effect from the core.
 
Status
Not open for further replies.

Similar threads

V
  • Question
Wanting To Build Solar Street Light with Multiple Features
Replies
11
Views
5K
shokjok
S
N
  • Question
Controlling multiple outlets with SSR Relay module and Arduino
2
Replies
31
Views
6K
nicofiaba
N
F
  • Question
Making a sinusoidal power waveform from two halves?
Replies
9
Views
2K
ZipZapOuch
  • Question
Powering RPI5 8gb from 5v supply
Replies
3
Views
2K
aliarifat794
A
P
  • Question
3v High Capacity Rechargeable Battery?
Replies
4
Views
2K
rjenkinsgb
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…