Attachments
@EUVL perhaps that circuit sketches the most basic transimpedance stage. Bias is mucked up, but the gist is there. It reminds me of the first prototypes I lashed up along transimpedance lines to explore these principles.
Raises another iconoclastic question which I discovered by accident: what happens if the cartridge is not ac coupled - ie the coil is in the emitter/cathode circuit and intentionally carries the dc quiescent current ?
In spite of the commonly held directive that 'thou shalt not pass dc through the cartridge', in practice I have found typically no change in performance for MM types up to as much as a few mA. Varies from cart to cart and gen type, but no detectable audible difference nor measurable difference IME.
Amongst other things, this then opens seldom opened doors for placing MM cartridge coils directly in cathode/emitter circuits, making use of coil R for biasing, in conjunction with natural transimpedance (low impedance) operation and loading.
Interesting, or should I close the door on my way out ?
Raises another iconoclastic question which I discovered by accident: what happens if the cartridge is not ac coupled - ie the coil is in the emitter/cathode circuit and intentionally carries the dc quiescent current ?
In spite of the commonly held directive that 'thou shalt not pass dc through the cartridge', in practice I have found typically no change in performance for MM types up to as much as a few mA. Varies from cart to cart and gen type, but no detectable audible difference nor measurable difference IME.
Amongst other things, this then opens seldom opened doors for placing MM cartridge coils directly in cathode/emitter circuits, making use of coil R for biasing, in conjunction with natural transimpedance (low impedance) operation and loading.
Interesting, or should I close the door on my way out ?
mA through?
But it's easy to do without the bias. We have these wonderful JFET devices that allow us to design preamps with negligible input bias current. They don't require a.c. coupling. And there are designs with no apparent capacitors for coupling, although they tend to conceal the capacitors in another portion of the circuit. With them we can use bipolars without the brunt of a substantial average input bias current, although the inductance of MM cartridges makes bipolar inputs noisier than necessary.
A few mA? For a typical MM cartridge this would require a few volts across. Not on my watch.
But it's easy to do without the bias. We have these wonderful JFET devices that allow us to design preamps with negligible input bias current. They don't require a.c. coupling. And there are designs with no apparent capacitors for coupling, although they tend to conceal the capacitors in another portion of the circuit. With them we can use bipolars without the brunt of a substantial average input bias current, although the inductance of MM cartridges makes bipolar inputs noisier than necessary.
To illustrate the principle, here's a minimalist preamp I devised in 2011. It's a sketch for a strain gauge preamp input stage - just one valve and one resistor.
Bias current for the strain gauge cartridge is the standing cathode current.
If one can tolerate the natural discrepancy between SG and RIAA eq, it has enough gain to be complete if the following amp is ac coupled and high input impedance !! Just one valve, and one resistor.
It works well, BTW.
So what happens if a standard MM is substituted for the SG cartridge ? Will the universe collapse to a singularity ?
Bias current for the strain gauge cartridge is the standing cathode current.
If one can tolerate the natural discrepancy between SG and RIAA eq, it has enough gain to be complete if the following amp is ac coupled and high input impedance !! Just one valve, and one resistor.
It works well, BTW.
So what happens if a standard MM is substituted for the SG cartridge ? Will the universe collapse to a singularity ?
Attachments
Last edited:
I'm not a believer, ribbon mike, or maybe some cantilever-less MC but I would like to see a compelling argument for MM in general. Has anyone done an experiment to actually measure the coefficient of electro-mechanical reciprocity in a MM or MI?
EDIT - I see you mention SG now that is a good example of 0 reciprocity.
What are the easiest changes to hear? Those of frequency response. What are the primary consequences of changes in cartridge loading? Frequency response.
How many reviewers know what the preamp input capacitance and the interconnect capacitance and tonearm wiring capacitance is? Perhaps a couple.
How many reviewers use test records to determine frequency response? Maybe a couple as well, although their data is curiously absent from the reviews.
How many reviewers have normal hearing?
Of course the strain gauge transducer has to have the bias current to operate at all. The MM cartridge doesn't need it and it may well be deleterious. In any case the cartridge was developed assuming zero d.c., and unless we wish to suppose the designers don't know what they are doing (a common diy aberration), it's prudent to avoid it.
I recall, I think some ten thousand posts or so back, Cordell said he could see virtually no reciprocity (not his words exactly). Of course there has to be a little or we wouldn't get a signal. But the importance of loading on mechanical damping is apparently small, although the thought of it contributes perilously to expectation bias.
An experiment reminiscent of burying the mic capsule in a lot of sand: immobilize the stylus and see how the impedance changes.
Last edited:
Oh no, not the 300lb of sand! I would think on and off the surface of an LP would show at least something. Maybe playing a vertical test tone and shorting one channel would show at least something.
Lucky, of course the change in FR is audible, but what about getting the same FR with equalization. B&K makes an ultra low noise mike with a deliberately slack peaky diaphragm and equalizes it flat. In fact operating a condenser mike (a nearly perfectly reciprocal transducer) into a short maximizes the distortion.
Last edited:
Well a few mA is a limit, not a target ! One might minimise it out of caution, but even a few mA (a volt or so drop) represents perhaps only a few mW of power total into the coil. IME it is typically tolerated well and I have not damaged a MM cart during testing this way. Neither have I found it to be typically deleterious to performance in any measurable or audible way.
Yes, but in practice any such assumption typically doesn't appear to matter, and in the case of some generators such as MI one can readily see why..........bcarso said:
Point is there's an upside opportunity by way of fresh circuit topology, especially in low impedance configurations, if dc is tolerated well by the cartridge. Even in conventional circuits, saving input coupling C is an immediate opportunity.
In MI carts, which many MM generators are, one can readily see why there is no dc effect on static position. I have also tested true MM carts and not noted static offset even at a few mA of coil dc.scott wurcer said:
As to dynamic reciprocity, in the past I've made calculation that shows electro-mechanical reciprocity to be very small in terms of force acting on the stylus due to normal electric load versus mechanical load of cartridge suspension mech impedance. By energy considerations, perhaps 4 or 5 orders of magnitude difference IIRC. Then even +40dB of load current might be considered negligible in mechanical effect, which tallies with what I have seen by way of no notable change to mechanical performance with transimpedance MM loading. The changes are more likely due to different non-ideals in the generator when operating at +40dB current in the case of transimpedance loading, IMO.
So I loosely agree with those posting only weak coupling between electric load and mechanical impedance as seen by the stylus, probably.
While I'm all for fresh circuit topologies, endorsing ones that pass bias currents through the cartridges seems just headed in the wrong direction. And we don't need coupling capacitors to begin with.
I see no basis for that, either in theory or in my experience of testing typical cartridges. I accept it is iconoclastic, and contrary to common lore, but that hardly seems a sound basis not to consider it on merit. It stands scrutiny - really. What tangible downside can there be in a MI cartridge, for example ?
Passing dc is not mandatory for transimpedance loading, but it certainly opens fresh topologies..........
- Status
- Not open for further replies.