Artificial floating ground

Hi All,

This is my first post, so I hope it works ok!

While reading this thread:

much less crazy idea

I noticed the talk of generating floating grounds with big resistors, which reminded of the scheme that Quad used for their 306 amplifer (amongst others, I'm sure). By putting 2 smoothing caps in series, the voltage at their midpoint is roughly equal to half the total supply voltage. To ensure that it remains close to this point, two transistors are used to steer the point to the required voltage. This then forms the signal ground for the whole amplifier.

This arrangement offers DC protection for the loudspeaker - if one of the output devices saturates. the midpoint will be allowed to rise to the appropriate tail, thus avoiding damage to the speaker. The only drawback of this is both capacitors must be rated to the total supply voltage (unlike a conventional split supply, of course)

Hopefully, this might be useful if you happen to have a large mains transformer with a single winding...? Apologies if everyone already knows this!


BTW, click here for the
complete Quad 306 schematic :)
 

Attachments

  • quadpsu.jpg
    quadpsu.jpg
    18.8 KB · Views: 1,166
Digging up this rather old thread! :)

I want to use this schematic to run a chip amp that needs a dual-rail supply from a single-rail laptop type SMPS.

1) Will it likely be okay hanging off the end of an AMPS, or does this only work with a standard linear PSU?

2) Those transistors are not around these days, what would be a good (affordable) alternative? My amp is only 25watts so maybe less power capable transistors could be used than in the Quad. I thought maybe NXP BCP53-10 to replace ZTX750. BCP56-10 to replace ZTX650? I don't know much about choosing transistors but they have similar voltage rating and HFE and 1A compared with the ZTX 2A current handling.

Thanks!
 

Ian Finch

Member
Paid Member
2010-04-11 4:22 am
Coffs Harbour, NSW
......2) Those transistors are not around these days, what would be a good (affordable) alternative? My amp is only 25watts so maybe less power capable transistors could be used than in the Quad. I thought maybe NXP BCP53-10 to replace ZTX750. BCP56-10 to replace ZTX650? I don't know much about choosing transistors but they have similar voltage rating and HFE and 1A compared with the ZTX 2A current handling.

Thanks!
My 306 amp is fitted with ZTX653/753 and these can replace any in the ZTX65X/ZTX75X range of parts which only vary by Vceo/Vcbo. Note other models in the 306,606,707,909 series use such other parts too. e.g: Quad 909 Manual - Stereo Power Amplifier - HiFi Engine
 
Digging up this rather old thread! :)

I want to use this schematic to run a chip amp that needs a dual-rail supply from a single-rail laptop type SMPS.
You do know that this is likely to result in rather considerable capacitive coupling to mains, inviting ground loop issues?
1) Will it likely be okay hanging off the end of an AMPS, or does this only work with a standard linear PSU?
I would expect a SMPS to be somewhat unhappy with capacitor values as shown. The circuit also implicitly assumes non-zero output impedance.

A splitter like that with two capacitors of value C is about equivalent to having output coupling capacitors of somewhat less than 1/4 C in a stereo amp. I.e. here you'd have the effective equivalent of about 1000 µF output coupling caps.

It's generally a better idea to properly convert the circuit to singly-supply operation. If you already have a PCB intended for split supply operation, tough luck. Either get another for singly supply or obtain a proper split secondary transformer. Stacking two switching supplies may be an option, but it would depend on the specifics of the supplies used, and I am not sure whether it's a good idea.
 
You do know that this is likely to result in rather considerable capacitive coupling to mains, inviting ground loop issues?

I would expect a SMPS to be somewhat unhappy with capacitor values as shown. The circuit also implicitly assumes non-zero output impedance.

A splitter like that with two capacitors of value C is about equivalent to having output coupling capacitors of somewhat less than 1/4 C in a stereo amp. I.e. here you'd have the effective equivalent of about 1000 µF output coupling caps.

It's generally a better idea to properly convert the circuit to singly-supply operation. If you already have a PCB intended for split supply operation, tough luck. Either get another for singly supply or obtain a proper split secondary transformer. Stacking two switching supplies may be an option, but it would depend on the specifics of the supplies used, and I am not sure whether it's a good idea.

Hi,

Can you explain more about the capacitive coupling to mains, please? I thought the SMPS is totally isolated from the mains side by the switching TX.

I can design a single rail circuit, however it would IMO be easier to split the supply to dual-rail because I have other line level bits to run as well as the amp. If I converted the amp to a single rail design it would need output AC coupling caps anyway, so whether they are on the output or in the PSU doesn't matter a lot to me.

I'd like to try it and see how it does. If it doesn't test well I'll change my tune.
 
You could build a 'full bridge' amplifier, that is, an amplifier with balanced output. Think of it as two bridged mono amplifiers at 180 degrees output, or better yet, a balanced amplifier circuit. Then there would be no need for an output coupling cap.:) You could use DC servos for each phase input set at 1/2Vcc, so as the speaker sees 0VDC between the two output nodes. Could easily make a DC detector between the outputs to trigger protection logic. Perhaps a tiny bit more complex:cheeky:, but the concept is sound.;)

(pun intended:D)
 
Hi,

Can you explain more about the capacitive coupling to mains, please? I thought the SMPS is totally isolated from the mains side by the switching TX.
For EMI reasons, SMPS' without a protective earth connection (e.g. figure-8 connector) usually implement a Y-rated capacitor in the order of 1 nF between secondary-side ground and mains return, or two of them to mains phase and return. Otherwise common-mode radiation would be an even bigger problem, I guess.

Supplies that do come with a protective earth connection (which would allow draining mains filter currents primary-side) may connect secondary-side ground to PE, or may opt to use a resistor of about a 1kOhm, presumably in parallel with some capacitance (found for some Thinkpad PSU, IIRC).

Those are the dirty little secrets of SMPS' in audio. If you want minimal capacitive coupling to mains, an old-fashioned xmfr with a shield winding still is the way to go.

As for the above suggestion, actually a lot of single-supply amplifier designs are BTL (loads of Class Ds and AB car amplifiers). The only drawback is that each amplifier has to drive half the load impedance, and as gain effectively increases by 6 dB, so does noise.
 
Last edited:
Just to make sure we are on the same page,you are saying this diagram is a no no.
This is actually not the same as the OP's problem. Your SINE sources are independent transformer secondaries, right? This is a split power supply with a different ground reference node, so it should basically be working fine and give the same performance under the same loading. However, unless you're voltage-limited, you may find that complexity can be reduced substantially by just stacking the secondaries on top of each other, leaving you with only one rectifier bridge and one set of filters.