A Fun MC to MM gain stage

[Colin James Wonfor]

By using LTC SPICE you can check this odd circuit out. I does work well but the PSU must be very low noise and very well regulated.

 

[amirm]

I must say I am not into LP cartridge differences . That key thing said, this looks like a class AB amplifier. Is there something specific to amplification of MC to MM?

 

[Colin James Wonfor]

100uV to X , normally 1 - 5mV . Very Low Noise, low input impedance and very little phase change if possible, and a good bandwidth.
It is a Common Base Push Pull Circuit, but it is odd, as the transistors are set to be constant current mode so the gain appears to be done by the resistors in the Emitter to Emitter and the resistors in the Collector to power rails.

 

[RayDunzl]

I must say I can't remember seeing the input on that side of the junctions before.

 

[Blumlein 88]

100uV to X , normally 1 - 5mV . Very Low Noise, low input impedance and very little phase change if possible, and a good bandwidth.
It is a Common Base Push Pull Circuit, but it is odd, as the transistors are set to be constant current mode so the gain appears to be done by the resistors in the Emitter to Emitter and the resistors in the Collector to power rails.

Has been awhile, but it looks similar to my memory of the John Curl circuit in the SOTA headamp. That used class A push pull transistors. Two pair in series as I recall. I doubt it was common base input however. It was one clean, quiet superb headamp.

 

[Colin James Wonfor]

I first used it in the A.C.Magnum pre-amps in 1977 so it has been around for while, but now with matching complementary transistor on the same substrate, much better performances are possible.

 

[DonH56]

Neat idea. I designed a similar one long ago, probably picked it up from somewhere. But, with a few significant changes. I found the basic circuit's gain and biasing was not very stable. The bias diodes added an extra tempco term (possible to null out but took some tweaking and I never got it as stable as I would like -- note I am talking stable bias for gain and such, not risk of oscillation) and the gain was sensitive to both the resistor ratio and the transistor current and temp. And as you said it is very sensitive to power supply noise and stability. At least where the thing was biased and at audio frequencies the emitters didn't start to look inductive until well above the frequency of interest. I piddled with putting diodes and transistors on a heat sink, putting it in a tight box with a well-regulated supply, and so forth and that helped a bit. I ultimately switched to current sources from a bandgap cell to control the bias and provide a hi-Z load, then used feedback to set the gain and help control the input impedance.

 

[Colin James Wonfor]

That is good, I used in mine two tranny per channel and common the opposite Base to base and used a green led which seemed to reduce the noise, and standard 78/79 series regs and fed them in a low noise OP -amp as a shunt regulator. Yes the input impedance was fine and ideal for the old low Z MC carts I found.
I also did a variation that I used as a RF front end on a FM tuner, it worked well and did go in small run of Inca Tech FMT tuners (50)
I also tried it with FET,s but never completed it as I was in hospital again (1987) this also meant I missed my son Scott's birth.

 

[Colin James Wonfor]

I have just noticed the Bible references in the text above the number 7, now let fine some with 42??

 

[DonH56]

Sorry to hear of your health issues; missing a birth is tough!

You'd probably not liked the FET version unless it was JFETs... Harder to control input impedance and noise is generally much higher with MOSFETs (LF noise corner, flicker noise, and all that jazz). Good in their place... I've not thought about it, however; does common-gate obviate some of the usual noise issues with MOSFETs and audio frequencies? Doesn't seem like it would...

p.s. I didn't get the Bible reference but it's been a while...

 

[Colin James Wonfor]

I Think I used 3K117 and 3J254 RF Fets, I have long since lost all the info. Also tried Fetlingtons but they were crap. Oh and a few tubes ECC88 I think.

Magic numbers 7,10,12 in the Bible, and 42 for "Hitch-hikers Guide to the Galaxy" it was to see if there was any fans out there, oh found none.

 

[DonH56]

Interesting. Any RF FETs (MOSFETs, GaAs, InP, etc.) I have seen had 1/f corners well above the audio band so were noisy. That said, I have not looked at a MC or MM phono pre design in ages.

And no, didn't see that show, sorry.

I liked tubes for their extremely high voltage overload point, less so for their noise in a MC stage. Good ones were fine, but hard to find. I used regulated filament and B+ supplies to reduce noise and keep them more stable. I tried current sources/sinks to fix biasing but they added noise. Nuvistors did OK, but most of mine used 12AX7/ECC83 just like everybody else because they were widely available and fairly cheap (back then). 6DJ8's seemed to have less noise but I've long since lost my old tube manual (buried in the black hole of old audio files in the basement or out in the shed).