Archive | 13:04

You can’t correct a bad start

4 Aug

As part of this, rather extended, look into decoding the matrix systems i thought i’d touch on an area where most are unaware that can seriously affect what is to be decoded.

And, when it comes to transcribing the LP, that is the cartridge and phono pre-amp.

Poor performance in this area can seriously hinder the ability of any decoding process, whether it be hardware or software. The main issues are channel seperation and, our old friend, phase.

Over the (too) many years i’ve been a Hi-Fi-olic i’ve been overly fussy over the quality of cartridge i used to play my album collection, even before i started as a trainee in the electronic business, which actually made me even more fussy. The main area that used to really bug was how some cartridges weren’t capable of producing a reasonable stereo stage, often producing something that fitted inbetween the speaker, but never quite making it all the way.

Of course, this is down to the cartridge having poor seperation between the two channels. What is shocking is that this issue continues to this day, which i guess if your just playing standard stereo albums, and your happy with you hear, then it doesn’t matter.

But what if your playing matrix encoded albums that are being put through the relevant decoder? There are many cartridges that fail to approach what is generally considered to be the minimum acceptable level of 20db, some i’ve measured actually fall to no more than 8db as the frequency reaches it’s resonance point.

If you think back to the encode equations i included a post or two ago, it’s obvious that, being mathematically based, the decoder expects to see an unadulterated signal to work on, but the effects of reduced seperation causing various levels of signal cancellation between the two channels can but only affect the decoders ability to do its job.

This effect becomes even worse if the channel seperation reduces (as it does with two catridges i have here) with increasing frequency then the effects become worse, especially when it comes to harmonics. Could this beone reason why one person suffers more issues with decoders containing gain (pumping) logic circuitry than other?

Having had a little experience in software decoding these albums i can firmly state that it definitely affects the quality of decode possible.

So, what’s the best to use?

Back in the day, and after trying many cartridges, i finally found one manufacturer who was capable of producing cartridges capable of producing a fully wide, stable stereo image. And that was Audio Technica. And the cartridge that changed my world was the (not so humble) AT-15SLa. Yes, a CD-4 cartridge.

At the time i noticed an improvement in the width of the stereo stage, decoding of matrix albums, as well as overcoming some issues i had playing CD-4 albums.

Coming up to date, and having tried aa few other manufacturers cartridges, i’m back using the great, great grandson of my original AT, that being the AT-440ML. This cartridge is capable of superb >20db seperation and, ‘when correctly loaded’, a remarkably flat phase response.

Many people have had issues with AT cartridges, complaining they sound bright, etc. But that’s because just plonking them into an arm and expecting them to work perfectly just won’t work.

Firstly, the arms height needs to be adjusted so it’s level with the LP being played, as well as the cartridge needing to be correctly aligned in the arm. And next, THE most important part…..

The Pick-up pre-amp. This area is responsible for interfacing and amplifying the low level signals from the cartridge so that it can be used. And it’s in this area where many a crime is committed.

This pre-amp (often called RIAA pre-amp) is there to correct the frequency response back to ‘normal’ after it had been altered during the master cutting stage.

And here, i must correct myself because it’s not just the frequency that requires correcting, but (you knew it was coming, didn’t you) also its phase.

A poorly designed phono pre-amp that only had its frequency in mind can cause havoc with the phase of the signal. A properly designed pre-amp will ensure that both its frequency and phase response comply with the RIAA specifications, ensuring an accurate (as possible) conversion.

Back in the day there was much talk of correctly terminating the cartridge to the pre-amp. The standard input impedance for a moving-magnet cartridge is set at 47k ohms. But in reality that can be the wrong impedance to get the best out of a particular cartridge. Take the Shure range from the 70’s and 80’s. We were told they worked best being loaded by ‘a total’ of 470pf to get them to sound better. Nowadays i think it might have been ‘the easy way out’ because a reduction to almost zero capacitance and a reduction in the impedance might have given results.

So, back to the AT-440ML. This family of catridges were originally designed for use in JVC’s CD-4 system, and they (basically) remain very true to the original design, so much so in that the AT-440ML is today used by those with CD-4 albums today. It works very well into the 100k recommened input impedance of the demodulator.

But what of it’s use as a stereo cartridge? You’d say, well it states it works well into 47k, so what’s the problem? Well, they WOULD say that wouldn’t they. Whose going to buy a cartridge that needs a non-standard input?

Whilst modifying my RIAA pre-amp i decided to look into what the AT-440ML needed the ‘see’ to work at its best. Cutting a long story short i ended up at an impedance of 39k with zero capacitance. With that i was able to get a nice, respectable, frequency responce along with a phase responce of 0 degrees, leading up to -6  degrees at resonance, with channel seperation never falling below 20db, even at resonance. A very respectable result.


For those thinking about delving into their pre-amps to do a bit of modifying, note that i used see word ‘see’ when describing the input impedance. That figure that is ‘seen’ does not necessarily mean the the resistor used at the very input of the pre-amp needs to be changed to 39k because that resistor can be anything from 47k upwards, depending on its configuration.

Ideally, when designing an RIAA pre-amp, it would be ideal if its input impudence is >1M ohms, so that it has no affect on the value of resistor used to set it’s impedance. But some designs can have an impedance of anything from 100k up to 1M, so the value of resistor chosen to set the circuits impedance needs to be calculated to ensure the circuit supplies the required impedance.



%d bloggers like this: