balanced to unbalanced convertor

[basite]

Hi all,

I was thinking lately about the advantages of balanced signals. I was looking for a pohono preamp a while ago, and balanced versions (balanced inputs) came up too. This made me wondering, the signal from a cartridge is, in effect, a (floating) balanced signal. Why don't we use them?

As I don't have the budget for a balanced phono pre, I want to try an experiment with a balanced to unbalanced input convertor. in theory, I would gain 6db SNR, and a stronger signal too. My cartridge is very very low output, so the 6dB would be a good start too

the plan is to connect the tt output to the "unbalancer", and then use a very very short (unbalanced) cable to the phono pre.

I made up a circuit, based on a TL071 op amp per channel,


I'd like some feedback on this design, can I improve this, would it work to begin with? I want to do this as an experiment, and improve later on if I find that it offers me advantages, and maybe build myself a balanced phono preamp alltogether.

I still need to design a power supply, will be using a transformer with center tap. I also think I made a mistake with the mass connections on the inputs. I connected them to the ground/center tap, but I think it would have been better if they were a chassis ground.

Anyhow, Thanks in advance,

Keep them spinning,
Regards,
Bert.

[basite]

anyone?

[kevinkr]

You haven't mentioned what sort of cartridge you have, but my gut feeling is that regardless of cartridge the TL-071 is not nearly quiet enough. The LM4562 and related devices would be a much better choice.

If you are using an LOMC then a transformer could well be ideal and provides balanced to unbalanced conversion if wanted as part of the deal.

[basite]

Thanks for the reply!

Cartridge is a Denon DL304, I am considering a better op amp too.
Transformers would be ideal, but they're fairly expensive, I think, so I wanted to try this as a cheaper alternative...

Thanks,
Bert.

[kevinkr]

Hi Bert,
I run an SPU and a DL-103D and IMHO there is no currently available op-amp that is quiet enough for use at these signal levels. Output of the DL-304 is just 180uV @ 5cm/sec lateral recorded velocity which places it very squarely in the very low output camp.

A pair of step up transformers or a fully discrete head amp are the way to go. If you want to go balanced without a noise penalty you need to use a transformer. A balanced active solution is going to be at least 3dB noisier than an unbalanced solution providing the same level of overall gain.

Traditionally head amps for these sorts of signal levels have comprised multitudes of well matched reasonably low noise transistors in parallel. Oddly enough the 2N440X family as well as a number of BC series devices are quiet enough for this application and cheap.

Some people have had success paralleling low noise op-amps as well. You'll need to read up on how to do this as it is not quite as simple as you might think. (At minimum a small series resistor is required on the output of each paralleled op-amp to assure that they do not fight over slight OLG and input related offset mismatches..) And the big caveat is I do not have a sense of how good this might sound.

I'd recommend roughly 22 - 26dB of gain with this cartridge, and if active should be unbalanced. I'd get transformers - something from Lundahl or look for partridge 972 on eBay.

[basite]

Hi Kevin,

what would be the transformer of your choice?

[basite]

Or, what would happen if I used a better opamp, and preamplify the signal already, to, MM level (say, 5mV). If possible, I might also lower the input impedance, to say, 5k total, so 2,5K resistors. If, however, that would work. I will need bigger input capacitors then though...

Just an idea...

Looking at transformers now too...

Thanks,
Bert.

[cyclecamper]

Are the Deane Jensen discrete op-amps still available? They are constructed of discrete components, including ICs which are massively parallel transistors internally. They also make about the best small-signal audio transformers you will ever find. But nothing cheap.

[AndrewT]

Have a look and seriously consider SALAS RIAA.
The thread is now quite long and includes all the build suggestions.
Support from SALAS is very good.

[PRR]

> the signal from a cartridge is, in effect, a (floating) balanced signal.

Traditionally one pin is strapped to cable or arm shield, forcing unbalanced. This can sometimes be removed.

> in theory, I would gain 6db SNR

How? It is the same signal, now split across two input devices. The natural result is 3db poorer S/N. Good design can give the same S/N.

You may be thinking of active amplifiers where "balanced" often means one unbalanced output plus an inverter to give additional output and balance. The cart total output is what it is.

That's for thermal hiss. Buzzz picked-up in the cable may be reduced 20dB or more. However hum picked-up in the coils generally is not changed.

> feedback on this design

The one op-amp unbal has large input resistors adding thermal hiss. The cartridge is 5K-50K of signal; you have added 20K of resistor hiss.

(Also the diff-amp input impedance is not "47K" nominal load.)

Counting very roughly on thumbs:

cart, 20K: 2uV hiss
20K diff-amp: 2uV hiss
TL072: 1.4uV hiss
NE5532: 0.7uV hiss

cart and plain '072 preamp: 2.5uV hiss
cart and plain 5532 preamp: 2.1uV hiss
cart and 10K+10K+'072 diff-amp: 3.2uV hiss

Changing diff-amp input resistors to a few K gets diff-amp hiss below cart hiss, but puts cart load far below nominal 47K.

You must go to 2-opamp or 3-opamp (or instrumentation amp) to get high input resistance with low noise resistance. Microphone amplifiers re-scaled for the higher impedance of a phono cart.

Mostly, there is little advantage to taking phono carts "floating" or "balanced". Hiss is not improved, and can be worse. Hum/buzz can be reduced, but cable length MUST be short (to limit capacitance loading) and the area "should" be "clean" (no arc-welders, fluorescent lamps, heavy power lines).