Split rail PS

When building a split rail low current PS with discrete components - like say +/- 15v - does it matter if the transformer doesn't have a centre tap on the secondary? It's just that I've seen circuits for both around. Some say to omit the centre tap connection to ground even if it has one.

Reply to
Dave Plowman (News)
Loading thread data ...

It depends on what you are trying to power. If the device simply wants a

30v supply, then the centre tap isn't needed. If the requirement is for +15, 0, -15 then you are going to have to get the 0 reference from somewhere. It can be done in the load device with suitable circuitry, but it's much easier to use the centre tap.
Reply to
charles

I can't think how one would work without a centre tap - and presuming it doesn't just mean two secondaries with a connection where they're joined. Is it some setup where one builds a (say) 0-15-30v PSU where the

15v terminal is used as ground and can sink current as well as supply it?
Reply to
Scott M

Depends on the nature of the load.

If the load draws exactly same current on both rails, no current will flow in the 0V connection, and in theory it's not needed. In practice, this is very unlikely.

If the load can draw different currents on the two rails, then a current will flow in the 0V connection. If there's no centre-tap on the transformer, this current in the 0V rail will need to be passed to the opposite rail, wasting power in the power supply. If the difference in rail currents is small, this probably doesn't matter. If it's big, the the power supply will have to start dissipating/wasting larger amounts of power, and in that case a design which can use a centre- tapped transformer will be more efficient and run cooler.

Reply to
Andrew Gabriel

If your transformer has a centre tap then the best way is to use it. (bridge rectifier to give 2 full-wave rectified supplies, earth the centre tap).

If it has two separate windings then you can either

1) bridge rectify them individually, making 2 isolated supplies and then connect - on one to + on the other to form a centre tap. 2) connect the windings in series (and in phase) to produce a centre tap.

If it has a single winding then the best way is probably to use a power op-amp to produce a "virtual" centre tap. Output current is very limited though, and you have an extra device to heatsink. This is sometimes done for low level audio stuff as the virtual tap can be very low noise if done properly. It is also highly accurate and will always be relative to both supply rails.

Reply to
mick

I> say +/- 15v - does it matter if the transformer doesn't have a centre ta p I> on the secondary? It's just that I've seen circuits for both around. Som e I> say to omit the centre tap connection to ground even if it has one.

There are ways to build +- psus with no transformer CT. One of the simplest uses a single secondary winding and 2 diodes, one giving a + rail, one a - . The downside is larger reservoirs needed. There are other schemes around too. I think you'd need to give more info about your transformer, what rail s you want, and how symmetrical the load is.

NT

Reply to
meow2222

you tend to need it if the current draw is asymmetric.

Reply to
The Natural Philosopher

It's the PS for a fairly esoteric moving coil pickup pre-amp which uses balanced inputs. The quiescent current at least is very similar on each rail. Basically was chasing the last residue of hum, and discovered (by accident) if I disconnected the transformer centre tap it improved. The scope showed the hum to be more on the negative rail. Changing the neg reg to a different make improved things to the point I doubt I'll better.

Reply to
Dave Plowman (News)

That's exactly the sort of job where a virtual ground is useful. You can get a single, nice smooth regulated supply then split it using an op-amp and 2 close tolerance resistors. Your ground is then bang on centre-rail no matter what the loadings are (within the output capability of the op- amp).

Reply to
mick

Sounds like a symmetrical load. Adding resistors can help deal with small i mbalance. For low current loads like that, passive power rail filtering is often enough, and can often help. IME though poor grounding layout and lack of IC decoupling are as often the problem with homemade preamps.

NT

Reply to
meow2222

Seems the small amount of hum left is coming from the actual cartridge or wiring to it within the arm. (It's an SME arm) Unplug the connector at the base of the arm and the hum goes - only a reasonable amount of hiss left, with the gain full up. Fine at any listening level.

The arm assembly is mounted on wood so it's not a ground loop. Moving the arm over the deck makes no difference to the hum - so I'd guess it isn't coming from the motor. The wiring within the arm is single wires - but then the arm itself should screen them? I've checked it is grounded to my star ground point.

Reply to
Dave Plowman (News)

Where the wires leave the arm to enter the cartridge?

Reply to
Tim Streater

Try connecting just the ground side of the cartridge, and see if that causes any hum. Could also try fully connecting it, but with the cartridge output shorted at source.

Reply to
Andrew Gabriel

I'm not sure where SME grounds the arm. There are five wires going to the connector on the base of the assembly (other end from the cartridge, as it were) - one of which is the arm ground. They all disappear up the middle of the pivot.

Reply to
Dave Plowman (News)

There's no actual ground side as the pre-amp has a balanced input. The pre-amp is mounted as close to the arm as possible. The cart has an output impedance of 2 ohms so as near a short as possible. ;-) But I'll have a play.

Reply to
Dave Plowman (News)

Ah, that's whay the screening doesn't matter much. However, it opens up the possibility for some other faults around the balanced inputs having gone wrong. Check none of them have got one side shorted to ground in the wiring or connectors. It could also be that one side of a balanced input circuit has failed.

Still worth testing with the outputs shorted - that might identify a faulty balanced input stage. (BTW, I expect it's the resistance which is 2 ohms - I'm surprised it's that low, but I hope the impedance is somewhat higher.)

Reply to
Andrew Gabriel

what you SHOULD do is to ground everything at the amplifier INPUT. traditionally that means that the arm carries a separate earth for the metalwork back to the amp chassis, which should be the ONLY part of the hifi stack that is actually attached to earth.

And that ONLY at the preamp input itself.

in order to reserve the safety niceties, what the means is that other equipment is only grounded electronically to cases by a 100 ohm resistor if the cases are to be earthed, and that the outer sheath of the phono/DIN plugs are NOT directly connected to mais earth on peripherals.

Reply to
The Natural Philosopher

In article , snipped-for-privacy@care2.com scribeth thus

Get yourself an older Audiolab 8000A amplifier the MC inputs in that were excellent performers:)...

Reply to
tony sayer

As I said, the arm is electrically isolated from the actual deck due to being mounted on a wood board. It appears to make no audible difference if the arm body is grounded or not. I've tried it direct to the main amp ground terminal both on its own and also grounding the PS O volts rail

It's the pick-up to on board pre-amp where the hum occurs. Pre-amp to main amp control unit is fine.

Reply to
Dave Plowman (News)

The question is a bit too genralised. What's the purpose of the PSU?

What design[s] did you have in mind for the circuitry?

A CT transformer may not be needed, but why not ask for some equally relevant information such as the type of mains switch or colour of the box you want to put it in?

AB

Reply to
Archibald

HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.