Tone control for class D amp?

[RojDW48]

Just for fun I bought an amazingly cheap class D amp from China (less than £5 - inc P&P - how do they do that?) and popped it into a Dansette. It sounds quite normal with a CD source but with the ceramic cart it is very treble heavy. Is there anything simple I could do about that - doesn't need to be an adjustable control? I am not sure about what might upset these awfully modern things.

[Herald1360]

That'll be a problem with its input impedance being too low for a good match to the ceramic cartridge. If there's enough spare gain, you could improve matters by trying a 470k or 1M resistor in series with the input from the cartridge to the amplifier.


If that doesn't work, you'll need to knock up a buffer amplifier- a simple unity gain non inverting opamp circuit will do the job.

Straightforward analogue electronics in front of the class D amp won't cause any strange problems.

A simple source/emitter follower will do too.

[Synchrodyne]

A simple emitter follower might come up short in terms of input impedance, so an emitter follower with a bootstrapped input might be preferable on that count. Such circuits were used in commercial equipment to obtain 2 M input impedance, e.g. some of the Tripletone amplifiers of the later 1960s.


Cheers,

[PJL]

a JFET follower is simple to add?

a JFET follower is simple to add?

Hence source in my post #3, very easy if you can get a JFET, they seem to be going out of fashion (please prove me wrong).

[paulsherwin]

Just bung a 1M resistor in the signal path. There's likely to be easily enough gain given that the class D chip is likely to have an input impedance of around 100k. We aren't talking about audiophile hifi here.

[RojDW48]

Quite right Paul - it's a Dansette when all is said and done with a £4.50 Chinese amp! The 1M resistor had the desired effect I am happy to say ( I experimented with lesser values before posting), which raises the question I have been far too lazy to ask myself. The Chuo Denshi cart and CD player have more or less the same output (I think but am happy to be corrected) so why the large value resistor? Basically I am ashamed to admit I don't understand impedance matching! Many thanks Chris et al for the good advice.

[Synchrodyne]

This one seems to recycle periodically, and a literature search shows that it had been a source of confusion and myths going back to at least the 1960s – although why it should be so is unclear.

The full answer will probably be found within these earlier threads:

https://www.vintage-radio.net/forum/...ad.php?t=60957

https://www.vintage-radio.net/forum/...ad.php?t=65007

https://www.vintage-radio.net/forum/...251#post300251

https://www.vintage-radio.net/forum/...515#post444515


In brief, piezo cartridges, crystal and ceramic are essentially capacitive devices which means that their source impedance is essentially capacitive.

This capacitive source impedance forms, with the amplifier input resistance, a high-pass filter. For any given cartridge, the turnover frequency of this filter is thus dependent upon the amplifier input resistance. Below the turnover frequency, the audio response drops by 6 dB/octave. Ideally, one would want this turnover frequency to be below say 50 Hz to avoid bass attenuation. For most piezo cartridges, this would require an amplifier input resistance of 1 M or more, often 2 M. Anything lower than that will result in a higher turnover frequency, and consequent bass loss.

For example, if 2 M is required to have a flat response down to 50 Hz, then a load of 100 k would move the turnover point up to around 1 kHz, with noticeable loss of bass – a bit like turning the bass control right down.

The problem is that most solid-state equipment has a relatively moderate input resistance, with 100k probably near the higher end of the spread that would be found these days for “line” or “auxiliary” inputs. Hence the bass loss problem when piezo cartridges are connected to such inputs.

Some possible solutions to the problem include:

Use a 1 M or greater series resistance, as you have done, This way, the cartridge sees the desired high impedance required for a flat response, but the output is significantly attenuated, thus requiring more gain in the amplifier.

Use a buffer amplifier with a high input impedance. For example this was done by Tripletone as mentioned above, by Linsley Hood in his 1969 modular preamplifier, and by Ferrograph on its F307 amplifier.

Use a “normal” input impedance amplifier, but include equalization (this is not RIAA equalization (*)) to counteract the bass loss caused by low impedance loading. Essentially what is required is a 6 dB/octave bass lift circuit with the same turnover frequency as with the bass loss interaction between the cartridge self-capacitance and the input resistance. Ideally this would be adjusted to suit each cartridge and input resistance combination, but for the higher quality ceramics the self-capacitance was often assumed to be 600 pF for this purpose. This approach was used by Quad in its 33 control unit, advocated by Burrows and used by H.P. Walker in his 1971 design.

This third method is a bit like inserting a bass lift tone control, so with your original thought to use a tone control you had pretty much nailed the problem in terms of its net effect.

Other sources, such as CD players, radio tuners, tape decks, etc, usually have very low (say < 10k, often more like 1 k) output impedances that are essentially resistive in nature. Thus they are close to being immune to unsatisfactorily low loading when connected to the vast majority of domestic audio equipment (although there are oddities and exceptions).

Piezo cartridges often have outputs of similar magnitude to other domestic audio “line” sources, but it is their input impedance requirements that make them behave differently to the latter.

(*) With probably 99% of piezo cartridges, the RIAA equalization is done mechanically within the cartridge, so is not required in the amplifier to which it is connected,. Beware though of the myth that connecting a piezo cartridge to a low-impedance input will cancel its internal equalization and so require its reinstatement in the amplifier. It is not so.


Cheers,

[Radio Wrangler]

Very nicely put, and quite comprehensive.

Designing good inputs for piezo cartridges back in the early transistor days was quite difficult and likely was the motivation for the appearance of lower cost magnetic cartridges.

Essentially the piezo cartridge produces plenty of voltage, but its output behaves as if it were fed through a small series capacitor. Any resistive loading of the output turns it into a high pass filter. If you couldn't get the resistive component of the load above a megohm or two, you lost bass. Any capacitive loading, like from long cables on the output of the cartridge formed a potential divider, attenuating the signal.

Piezo cartridges were common in one-box players and radiograms where cabling was short and low capacitance. This was another factor in the swing to magnetic cartridges as hifi systems with separate amplifiers and turntables came into fashion.

Ceramic cartridges with lower output voltage have larger effective series capacitance and so tolerate lower resistance loads, but then they forced an extra valve's worth of gain on the record player amp, and that cost money. You got nothing for free!

David

[RojDW48]

Thank you gentlemen - I can understand the basic concept now. Thanks for taking the time to explain it so thoroughly and clearly.

I didn't know that the RIAA correction was done mechanically, I merrily assumed a single low pass slope was used to approximate it. You learn something everyday.

[Synchrodyne]

Of the many writings on this subject, I think that the attached article best encapsulates the history, situation, problems and potential solutions in not-too-many pages, and is also illustrated with pertinent graphs. Note that it includes the genesis of what became the Bailey-Burrows and then the Bailey-Burrows-Quilter preamplifier circuits.


Cheers,

[Radio Wrangler]

Much of the gain variation in the RIAA curve is a compensation for the inherent differentiation of the velocity to voltage conversion of a magnetic cartridge. Piezo transducers lack this effect and so correction of their frequency response does not need to be so dramatic and is within the scope of some mechanical compensation.

David