And doesn't this remind you of when you were in the boat, and then later than night, you were lying, looking up at the ceiling, and the water in your head was not dissimilar from the landscape, and you think to yourself, "Why is it that the landscape is moving, but the boat is still?"
You may be aware that Ed Meitner has a patent on a "Magnetic pickup preamplifier", US4547819 where "a magnetic pickup ... is terminated by a negative impedance ..." but this is for MMs and he "appears" to be manually calculating the necessary feedback compensation from the stray capacitance and resistance values of the cartridge and is not sensing it thru a resistor (here, Nelson Pass' aleph current source comes to mind).
The new Accuphase C27 phono preamplifier applies followers (JFETs in this case) at the input of the MM circuit (basically the same topology as the MC circuit but built separately on the same board) and obviously to allow for a higher input impedance. I cannot comment on what syn08 had in mind but that could be a possibility. Not sure though if you would consider the Accuphase a cool preamp, probably not
analog_sa,
I won't build it. As stated, the grounded base input topology is not negotiable and because of other disorders.
To put it mildly, since the opposite characteristic is desirable.
Life never ceases to surprise. A convincing reasoning is likely to be found in the glossy brochure included.
It could. At the same time, people confusedly talk about impedance matching, cartridge loading and negative input resistance.
He was exhibiting at the High End show in Munich. He has a new loudspeaker with Altec wide band, you know with the Manta Ray horn.
The funny thing was (actually not funny at all) that a tube amp he had on idle during the night put a wooden board on fire that was thermally too close to the tube amp. The sprinkler got on and the water and dirt was dripping down in the first floor flooding an open stand of a company from Italy, i do not remember the name. A total mess and as far i heard there was no enshurance paying the considerable amount of damage done.
The funny thing was (actually not funny at all) that a tube amp he had on idle during the night put a wooden board on fire that was thermally too close to the tube amp. The sprinkler got on and the water and dirt was dripping down in the first floor flooding an open stand of a company from Italy, i do not remember the name. A total mess and as far i heard there was no enshurance paying the considerable amount of damage done.
Out the window play the unicorns of ham, soothsayers on the wall catch light in silver jars, never trust a swamp dweller they tend to break unexpectedly.
When designung low noise circuits, the theory generally is that you have try to have the highest gain in the first stage (of course, taking good care not to cause overload etc) in order to minimize the noise contribution. So, my question about these common base 0db voltage gain designs at the INPUT of a high gain amplification chain is: do they not add more noise than they should? Why buffer the mc cartridge when it is quite capable of driving happily into 100 Ohm load impedance? I, correctly or not, would have tried to put as much gain into the head amp as possible and avoided the buffer. What am I missing? Help me out here!
A common base stage does not have an input impedance that is zero ohms, at least not in the open loop case. With shunt feedback it can be much reduced like the PhonoClone design but lets talk about the common base stage.
So the common base stage is also amplifying the voltage. I like common base input stages because they have a "natural" input impedance that matches a low impedance MC quite well and does not suffer from early effect so is high speed without the need of high feedback or series cascoding.
So the common base stage is also amplifying the voltage. I like common base input stages because they have a "natural" input impedance that matches a low impedance MC quite well and does not suffer from early effect so is high speed without the need of high feedback or series cascoding.
#A common base stage does not have an input impedance that is zero ohms, at least not in the open loop case. With shunt feedback it can be much reduced like the PhonoClone design but lets talk about the common base stage.
So the common base stage is also amplifying the voltage. I like common base input stages because they have a "natural" input impedance that matches a low impedance MC quite well and does not suffer from early effect so is high speed without the need of high feedback or series cascoding. "
Why is early effect an issue when the amplified output signal is still only a few mV? OK I can understand it if we were swinging a few volts, but a typical MM input is 3-5mV, which is what the head amp is supposed be converting the mc input to. If early effect is an issue, you can always cascode the 1st stage.
I see you mention the low input impedance as being good on the common base design . . . but why do you then load the mc cartidge with 100 Ohms or some other low value? surely the value of a low impedance input is of no significance if you then stick a 100 ohm resistor across it.
Now, I can imagine from the noise point of view, you can possible argue that at LOW frequncies where the cartridge impedance is very low, the transistor noise might be an issue, but what happens at say 10KHz when the cart z is much higher? The current noise contribution then becomes an issue.
So the common base stage is also amplifying the voltage. I like common base input stages because they have a "natural" input impedance that matches a low impedance MC quite well and does not suffer from early effect so is high speed without the need of high feedback or series cascoding. "
Why is early effect an issue when the amplified output signal is still only a few mV? OK I can understand it if we were swinging a few volts, but a typical MM input is 3-5mV, which is what the head amp is supposed be converting the mc input to. If early effect is an issue, you can always cascode the 1st stage.
I see you mention the low input impedance as being good on the common base design . . . but why do you then load the mc cartidge with 100 Ohms or some other low value? surely the value of a low impedance input is of no significance if you then stick a 100 ohm resistor across it.
Now, I can imagine from the noise point of view, you can possible argue that at LOW frequncies where the cartridge impedance is very low, the transistor noise might be an issue, but what happens at say 10KHz when the cart z is much higher? The current noise contribution then becomes an issue.
Exactly, and more:
- It is not obvious that MC cartridges should be loaded with a low impedance. Manufacturer specs are within a wild range.
- A high impedance input can always be lowered, not the other way around. Only low impedance is useless for (e.g.) high output MCs.
- Non degenerated common emitter and common base have exactly the same input impedance (Beta/gm) and transconductance gain (gm).
- My main problem with Hiraga or any other similar approach is the required power supply. To place the devices in the minimum noise area, you need to run them pretty hot (some 10mA) so batteries are not really an option. The requirements (hum, noise, etc...) for a power supply to feed the Hiraga thing are huge. As usual, it is very bad engineering to move the stringent requirements from the signal stage to the power supply, it's kinda sweeping the dirt under the rug.
Now, if input current is really an issue, using a bias current cancellation technique is pretty straightforward. This is just another servo, and I was able to bring down the input current to the nA levels, close to a JFET input... Unfortunately, it is thermodynamically impossible to cancel the input bias current without impacting the noise. How large the penalty is, depends on the cancellation technique. My best was about 0.3nV/rtHz (by using low noise opamps, discretes could do better, at the price of complexity).
It is really funny to watch how opinionated the GEB team can be, in particular without showing any practical implementation or results. It reminds me of simulation afficionados, but at least those did some efforts to support their opinions.
I have probably heard and measured more phono stages then you can imagine. Some of them sounding great with a high impedance loading. So i do not like to be put into the GEB team. I could also accuse you to be opinionated because you simply can not beat me in terms of experience with other peoples creations. For example I owned and measured several versions of the Connaisseur that sounded great with a high impedance fet input stage. I own several versions of the Essex equalizer and was a Hiraga dealer so again heard and measured several incarnations of Le Pre-Pre. I own Phonostages by Trigon, Bladelius, Hoer-Wege, Ultee, Convergent, Lorycraft, LC Audio you name it. On shows i worked with Mark Levinson, Krell, Rowland, Aesthetics, Aire, Burmester, Pass to only name a few. I have usually a good relationship to the designers and know a lot about the technologies they are using. I have experience with tubes, fets, bipolars, opamps and a mix of it with currentfeedback, voltagefeedback, LCR RIAA etc. etc. That is exactly the reason because i like your designs without even hearing them. They breeze some fresh air into a szeene that is basically stuck in the past. I only want to help if i can and think your designs are great. I really would like to listen to one of them on my system. I also did some simulation concerning loading and will present them here in some days.
Sorry, I have no problems with your position and experience. It's the Spambot team that's slightly irritating in it's sufficient ignorance.
Exactly this observations I get by some evaluations and comparing of different circuit topologies.
But the exactly reason I haven't find out, because I don't find p-spice parameters of moving coil cartridges, that have internal DC resistance <20 ohms
Here two schematics, that outperform the most others regarded sonic quality (not necessarily regarded signal to noise ratio - for me purely cosmetic parameter).
Hiraga MC Preamp
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.