Split rail from two positive regulators

[karatestu]

Hello. Trying to get my little brain round something and need a little help.

I have four Avondale Audio TPR4 regulators sitting around doing nowt. There are two separate positive voltage regulators on each board. Its basically a tracking pre regulator with voltage reference and capacitance multiplier tagged on the front.



I want to try these with on my nva phono board. But that is split rail and I don't have any negative versions (Avondale only do +ve). I suppose i could swap cap orientation and change the three pin devices for negative versions but i don't want to hack them incase I end up selling them

The following is a post by Les W on pfm about this very thing


LesW
Retired at user request
chiily said: ↑
Can the TPR4 be configured for +18/-18v working...Maybe to power my B4?



You can, very simply. The caveat is that you will need either two quite separate transformers OR a transformer with two independent secondaries.

The idea is that you treat the outputs from the TPR4s as though they were batteries in a multi-cell flashlight. The connections are: +V...0V....+V....0V

The join between the supplies becomes the 0V - zero volt point.

Only ONE TPR4 is needed to make a symmetrical supply.

Hope that's made it clear although I've spent ages on the phone sometimes trying to explain this principle.

I have some graphics for those still confused. Let me have an email and I'll send a copy - les(at)avondaleaudio.com


So I need two independent positive power supplies using separate transformer windings. There are three connection points on each regulator - Vin, Vout and OV. On the first regulator the positive voltage from one of the power supplies is connected to Vin. Vout from this reg is connected to OV of the second positive regulator and it is this link that becomes 0V.

So do I connect the second independent positive raw supply to Vin of the second regulator as usual ?
Where do I get the positive and negative outputs from ?

I'm being a bit thick here

[Nick]

Two entirely separate transformer secondaries to their own bridge rectifier and cap. Positive on cap to vin on regulator. Negative on cap to 0v. That then gives you two entirely separate 18v supplies call them a and b. Connect vout of supply b to 0v of supply a. Vout of supply a is +18v. Point where a and b are connected becomes 0v. 0v of supply b becomes -18v.

[karatestu]

Two entirely separate transformer secondaries to their own bridge rectifier and cap. Positive on cap to vin on regulator. Negative on cap to 0v. That then gives you two entirely separate 18v supplies call them a and b. Connect vout of supply b to 0v of supply a. Vout of supply a is +18v. Point where a and b are connected becomes 0v. 0v of supply b becomes -18v.

Thanks Nick. It makes sense now. It is clear that the OV of the two positive voltage power supplies should not connect at any point, otherwise I will be needing a change of pants

I have drawn a primitive diagram just so that I am sure I have got this right. Hope you can see the detail. Just noticed I drew the transformers wrong oops

[karatestu]

There was some talk on pfm that using two positive regulators with LM317 in this way can be better than using the LM337 because the 317 regulates better. I wouldn't know about that.

The nva phono uses 317 and 337 in the traditional way. It has small film caps on the output of the regs which I thought were a bit of a no no due to the ringing they can cause when there are film decoupling caps in parallel at the opamps. I always thought the 317/337 output decoupling caps should have a bit of esr for damping (like a tantalum).

It does require another transformer though so not so good when enclosure space is limited.

[Nick]

As always the first place to start is with the documentation of the part:

In general, the best type of capacitors to use is solid tantalum. Solid tantalum capacitors have low impedance
even at high frequencies. Depending upon capacitor construction, it takes about 25 μF in aluminum electrolytic to
equal 1-μF solid tantalum at high frequencies. Ceramic capacitors are also good at high frequencies. However,
some types have a large decrease in capacitance at frequencies around 0.5 MHz. For this reason, 0.01-μF disc
may seem to work better than a 0.1-μF disc as a bypass.
Although the LM317-N is stable with no output capacitors, like any feedback circuit, certain values of external
capacitance can cause excessive ringing. This occurs with values between 500 pF and 5000 pF. A 1-μF solid
tantalum (or 25-μF aluminum electrolytic) on the output swamps this effect and insures stability. Any increase of
the load capacitance larger than 10 μF will merely improve the loop stability and output impedance.

It does require another transformer

Or two secondaries and bridge rectifiers.

[karatestu]

As always the first place to start is with the documentation of the part:

In general, the best type of capacitors to use is solid tantalum. Solid tantalum capacitors have low impedance
even at high frequencies. Depending upon capacitor construction, it takes about 25 μF in aluminum electrolytic to
equal 1-μF solid tantalum at high frequencies. Ceramic capacitors are also good at high frequencies. However,
some types have a large decrease in capacitance at frequencies around 0.5 MHz. For this reason, 0.01-μF disc
may seem to work better than a 0.1-μF disc as a bypass.
Although the LM317-N is stable with no output capacitors, like any feedback circuit, certain values of external
capacitance can cause excessive ringing. This occurs with values between 500 pF and 5000 pF. A 1-μF solid
tantalum (or 25-μF aluminum electrolytic) on the output swamps this effect and insures stability. Any increase of
the load capacitance larger than 10 μF will merely improve the loop stability and output impedance.

It does require another transformer

Or two secondaries and bridge rectifiers.

Thank you Nick. Something to think about.

Does my diagram above look correct?

Stu

[Nick]

Does my diagram above look correct

Yes, looks perfect.

[karatestu]

Many thanks. Will be doing it in the next few weeks hopefully.

Stu

[karatestu]

I suppose this would work with two secondaries that are not centre tapped ?

[Nick]

I suppose this would work with two secondaries that are not centre tapped ?

Yes, or two secondaries that are center tapped and ignore the tap.

[Daniel Quinn]

Why not just get a transformer that will give you a working DC voltage ?

Are we saying voltage regulation is better ?

My DIY phono stage as 2 positive rails of 24v , I use 18v trafo and caps with no regulation . Would regulation improve the sound ?

[Nick]

Are we saying voltage regulation is better

Don't know about we, but I would say that in many cases yes. However in many cases it might not be worth the extra cost and complexity.

Would regulation improve the sound

Again I don't know, but the fact that it seems to be oft repeated that increasing the size or number of transformers in that phono stage does improve things would lead me to believe that regulation would improve things. But there are huge numbers of assumptions in that. Its possible (for example) that a bad power supply leads to a subjectively better result.

[karatestu]

The NVA phono stage comes as standard with regulators. They are very simple . LM317 & 337 with the necessary resistors and 0.1uf film caps on input and output ( from memory).

I remember RD saying that the stage did benefit from regulators but very simple ones.

I am trying something which is quite a bit more complicated and if I can't improve subjectively on the stock regulators then they will be sold. I don't know if I don't try. I have some opamp based super regs knocking about too but I don't think i will be trying them .

[karatestu]

DQ - is your phono stage based on the Naim Stageline or similar. They use two +24V rails as standard. Naim put the regs in the power supply at the end of one metre of wire. I would suggest that the regulator would be better close to the load if you were going to do it.

[Daniel Quinn]

Yes it is a clone of the stage line.

I've got about 8 phono stages including the nva , personally I prefer the naim over the nva
.
.you can buy regulators for it , but I just use two 350va transformers ,2 33,000uf caps per rail and 1 bridge