Vintage Mic Transformer for MC step-up - gain-phase and impedance measurements

[Gold]

That sounds a little like my old P-10 phono preamp topology.

It seems no one has actually quantified the benefit of a balanced interface. So it’s usually dismissed because it isn’t a whiz bang solution.

Also possibly because it will show the emperor has no clothes when it comes to noise figures.

 

[JohnRoberts]

It seems no one has actually quantified the benefit of a balanced interface. So it’s usually dismissed because it isn’t a whiz bang solution.

The benefit of balanced technology is pretty well established in professional audio interfaces. As I mentioned, maybe even in this thread, one small TX company got a patent for applying the concept of balanced inputs to a phono preamp. The patent wrapper (the back and forth arguments exchanged between inventor and the examiner) was almost humorous to read as the junior, wet behind the ears patent examiner, marveled at the new to him benefit of using balanced inputs for CM noise cancellation.

Also possibly because it will show the emperor has no clothes when it comes to noise figures.

It has been decades since I thought about this stuff seriously. My experience was mainly with consumer playback technology. IIRC the some (?) of consumer MM phono carts came with the low or the (-) lead on one channel of the stereo cart bonded to the head shell ground. Suggesting to me that perhaps single-ended is the expected front end topology.

JR

 

[Cqwet Dbdfte]

The European market, Thorens, Philips, Ortofon, B&O, used 5pin and 7 pin DIN connectors for TT, and other 2ch sound equipment, which had (have) a separate ground pin. I have not looked into the circuits used for phono inputs.
The fifth pin, (typ. center) was connected to shield/ground tonearm. No separate ground lead for tone arm ground.
The US market was not receptive to the European connector standard.
Obviously the transformers used for MC carts lend themselves to differential signaling.
Shortsighted marketing geniuses favored the RCA plug.

https://pro-jectusa.com/product/connect-it-phono-s-5p-din-mini-xlr/ https://ortofon.com/pages/ta-100-9-tonearm

 

[Gold]

The fifth pin, (typ. center) was connected to shield/ground tonearm. No separate ground lead for tone arm ground.
The US market was not receptive to the European connector standard.
Obviously the transformers used for MC carts lend themselves to differential signaling.
Shortsighted marketing geniuses favored the RCA plug.

The fifth pin is usually tonearm shield IME. I thought the tonearm connector would be the best connector to do a balanced interface with. People recognize it as a tonearm connector so putting the same opposite sex connector on the other end of the cable and plugging it in to the phono preamp would make sense to non technical people. The connector is large enough to run 2pr microphone cable. It's even plugging one less thing in than RCA's.

 

[Cqwet Dbdfte]

Well said, without a "signal" level, you cannot state the SNR.
I'm not "marketing" my phono preamp.

I measured 2.692V rms @ 1KHz 5cm/sec track. Noise, low freq dominated, 1.8mV rms.Two different DMM's brands, same numbers.
My listening level peaks out at 88 - 90 dB spl, room noise about 44dB, without A/C running.

 

[Gold]

I measured 2.692V rms @ 1KHz 5cm/sec track. Noise, low freq dominated, 1.8mV rms.Two different DMM's brands, same numbers.
My listening level peaks out at 88 - 90 dB spl, room noise about 44dB, without A/C running.

5cm/sec is 3dB down from 7 cm/sec. My calibration is 7cm/sec = +4dBu. To make our measurements match you should calibrate 5cm/sec to +1dBu or 0.87vac. Once 5cm/sec is calibrated to 0.87vac measure the SNR or noise floor.

 

[Cqwet Dbdfte]

5cm/sec is 3dB down from 7 cm/sec. My calibration is 7cm/sec = +4dBu. To make our measurements match you should calibrate 5cm/sec to +1dBu or 0.87vac. Once 5cm/sec is calibrated to 0.87vac measure the SNR or noise floor.

I will, tomorrow or so. 2.7 V at 1KHz is 27V rms at 20 Hz, also the low frequency noise is also getting a 20dB lift.
I have a passive relay 128 step attenuator followed by a (superfluous)10dB gain stage and cathode follower.
I'm planning to replace that board with my phono pre amp, after I figure out what I want to do with power supplies.
The acoustic noise from my environment greatly exceeds that of my amplifiers, but I will look into my voltage references for noise profile. I optimised the phono stage for linearity, with mu-followers, noise was a secondary concern, my MC transformers has a 27dB noiseless gain, if one keeps magnetic fields away, and tweak out any ground loops.
I may adopt the 5 pin connector, either XLR, or that Tuchel flat pin, used for power supplies.
This may involve pulling out tone arm wires, which is not fun.

 

[Gold]

The noise is the cable run. If you doubt that then substitute the cartridge with a resistor at the phono input and gaze at those beautiful numbers.

 

[Cqwet Dbdfte]

In my case the the phono cable runs are very short, the secondary of the MC transformers have termination resistors to give the correct reflected impedance to the cart. The cables to and from preamp are 30cm twisted pair double wrap shield microphone cables. Only one end connects the shield to the case of RCA connector.
The cables from TT are ~ 1m included with the RB300 tone arm.
The line stage included, besides a tape-wound R-core transformer, also a small EI-lamination 5VA transformer, on a PCB, which I replaced with a toroidal transformer with much less magnetic fields on a new PCB I designed.
Shielding from magnetic fields is problematic, non-ferrous alloys are practically transparent, but easy to machine and looks OK. Steel would be better, copper plated even more so.
I have started to replace typical 120VAC power cables with their shielded counter part.
If had long cable runs in a hostile noisy environment I would consider using EMT for AC mains lines, and even separate ones for signal lines, and use "greenfield" to cover the ends, if needed. Physically separating power supplies with power transformers and remoting these from sensitive signal lines and MC carts should help. -66dB from Ortofon spec 0.5mV is only 23 nano volts.
Reducing my phono stage output to +1 dBu from the present
+10dBu will not change the ratio of signal to noise.

 

[Gold]

If you don’t have a phono preamp with variable gain that can be calibrated to a Standard Reference Level it’s impossible to do the noise floor measurement correctly.

A professional phono preamp should have variable gain so it can be calibrated to a Standard Reference Level.

 

[Cqwet Dbdfte]

I have the provision for adding an interstage attenuator, which I have sitting in a drawer, remote and all, but have not yet installed.
I could make the measurement on the line stage output, but one more stage would be added to the soup, but the expected level of -66dB would get a bit small for my DMM to measure reliably, as it does not have an averaging function . I would collect some other gear to measure 455uV rms.

 

[Gold]

If you attenuate after the initial gain stage the SNR is already locked from the first gain stage. The first gain stage is where the noise is.

 

[ccaudle]

The noise is the cable run.

I'm a big fan of balanced interfaces, but I'm having trouble understanding where the noise is flowing on a floating cable. The biggest source of noise on single ended connections between different pieces of mains powered gear is the parasitic power supply currents which flow on the shield between chassis, and since the shield is also the signal reference that necessarily adds the noise to the signal at that point.

A phono cartridge obviously has no mains connection, so there shouldn't be 50/60/100/120 Hz currents flowing on the cartridge cable even if it is single ended instead of balanced. Where is the noise coming from?

 

[JohnRoberts]

If you attenuate after the initial gain stage the SNR is already locked from the first gain stage. The first gain stage is where the noise is.

Generally the initial gain stage dominate combined system S/N but there can be noise contributions from the EQ stage due to the significant response range within RIAA EQ (approaching 40 dB more LF gain than HF gain.

Over the decades there have been multiple design topologies combining flat front end gain stages with separate following RIAA EQ stages (including from me). There are headroom tradeoff considerations about using too much front end gain.

Then there is some (free?) S/N benefit from wrapping RIAA eq around the total system gain stage. As I have already shared RIAA is effectively a roughly 2kHz LPF. Historically we have seen RIAA stages that measure better THD+N than the HF linearity they actually deliver when stressed by HF content.

I considered marketing a professional phono preamp because I hoped they would be less phoolishness than from the consumer audiophile market. I was sorely disappointed so backed away from that market segment.

JR

PS: I even designed a 19" rack mount phono preamp under Phoenix audio Lab (Loft/loftech) brand.

 

[Gold]

The noise is at -55dBu. It swamps circuit noise. That’s why the noise floor of the phono preamp doesn’t matter. If you use a simulated source impedance you can match the numbers on the spec sheet. It will look beautiful.

Try it before you dismiss it. I’ve done the work. I’m not being cagey about anything. If you don’t believe me try it.

 

[Gold]

You don’t even need to calibrate the phono pre. Just observe the difference between having a cartridge attached and substituting a resistor. There will be a huge noise difference.

 

[Cqwet Dbdfte]

I had almost forgotten how common the 5-pin DIN was, and turntables using it, over 43 years ago. I do not know how the shielded DIN wire was constructed, but I was baffled by that RCA crap with a separate ground lead. Amazn and bay are full off adapters between the two connector systems.
The cart needs shielded wires, and using coax from tone arm base (hopefully the tonearm is a metal tube) which would force signal into to shield. Not ideal for noise reduction. A microphone cable would be better, where the shield is connected at the TT end only, and the two twisted wires carry the signal to the phono preamp, not the shield. Then you woubetter shielding. The ground lead would still be needed. The microphone cable may have more capacitance that a typical RCA phono coax, which would be a problem for MM carts.
I do not have a huge hum problem, so a cabling switch may not have much value. I have changed location of the 60Hz iron as much as possible.
Using transformers for MC carts has been the standard since the beginning, and could perform a bal to single ended function. I still only ground one end of the shield, at the source, using typical one pair microphone cable.
The cart itself is a pickup point for 60Hz fields, and non ferrous tone arms won't help much here. Keeping the line frequency fields away from carts is imperative, which would include synch motors, that better be well shielded and/or as far away from the cart as possible. Some TT mfg has placed those quite far away driving platter with long silicone O-rings.
Maybe for that reason.
I'm sure someone has dug more into this subject.

 

[Gold]

I have an Ortofon RM309 that I was under the impression was from the 1960’s. It has the 5 pin connector. It is a DIN connector but so is a MIDI connector and an Amphenol 5 pin. Calling it a 5pin DIN doesn’t get you very far. I call it the tonearm connector. I’m sure it has a real name.

The best I ever did with an unbalanced run was with RG59 and BNC connectors. I got a 3dB noise improvement. Cable capacitance with RG59 is also reduced. Before I switched to a balanced interface all unbalanced phono connections were done with RG59 and BNC connectors. The longest run I ever diid was about 20ft. No noise difference between 2ft and 20ft.

If using a balanced interface with the tonearm connector, the shield should be connected on both ends. The 5th pin is tonearm shield. If the tonearm shield is connected to turntable ground it could cause hum. The shield is only useful for RF/EMI so it could be connected to preamp chassis through a small value capacitor. That way it is only connected at RF/EMI frequencies. I never found that cable length mattered much.

 

[Cqwet Dbdfte]

Yep, the 5-pin DIN is also used for MIDI, Tape machines etc. That was (is?) the main audio connector in Europe, where I lived 42 years ago. I built amps using DIN connectors, always 5-pin for 2-channel, did not think much about it, it worked. Typical DIN-5 are not as robust as XLR or Tuchel, but it worked in the consumer sphere. BNC is my least favorite RF connector, but it is OK for audio. There are gobs of RF connectors to choose from depending power level and frequency and passive intermod specs. New types invented continuously.
Right, RFI/EMI is what the shield is for, 60Hz from magnetic fields is very hard to suppress. EMT conduit may work, have not tried.
Ground, or more correctly "bonding" shields to cases can is some instances be part of a ground loop problem. My suggestion would require twisted pairs for signal, and the shield connected to one of the phases at one end. Will not work with coax, where bonding at both ends is the only way for the signal to get thru. The tonearm connector is a bit small, not so robust.

www.calex.com/pdf/4ground_shield.pdf
Above dismissed by Cable Shield Grounded At One End Only - EMC Standards who did not consider differential signalling, the only subject from the Calex paper. Every Ethernet CAT-5 are differential with transformers at each end, that Ethernet cable could well work with audio signals too. The old RS-485 is differential with terminations, and there are many digital signalling systems using differential with very small signals, to improve noise immunity.
Maybe the cart signal could have a defined potential to shield, like some of the above.
Either way, RCA plugs for phono is not a "good thing".

 

[Gold]

Yep, the 5-pin DIN is also used for MIDI, Tape machines etc. That was (is?) the main audio connector in Europe

The MIDI connector is not the same as the tonearm connector. I’ve never seen the tonearm connector used for any other purpose. XLR was used for audio in Europe and the US from the 1960’s on.

Previous to that in Germany the big round Tuchel like on vintage German microphones was the standard audio connector. Currently made by Binder.

In the US before the xlr became standard most equipment used barrier strips.

where I lived 42 years ago. I built amps using DIN connectors, always 5-pin for 2-channel, did not think much about it, it worked. Typical DIN-5 are not as robust as XLR or Tuchel, but it worked in the consumer sphere. BNC is my least favorite RF connector, but it is OK for audio.

The Tuchel Amphenol round C91 series connector was and still is used extensively. It is also often also called a DIN connector.

Every Ethernet CAT-5 are differential with transformers at each end, that Ethernet cable could well work with audio signals too. The old RS-485 is differential with terminations, and there are many digital signalling systems using differential with very small signals, to improve noise immunity.

I use Belden Meditwist for multichannel AES runs.