Paralleled Hitachi devices would work rather well. I still have a number of both P and N. Got them in 1978 from Japan. They really were one of the best of the complementary devices ever made. They did have some limitations, however, and they would not work in my JC-1 pre-preamp topology that I originally got them for. They need some DC voltage across them to work properly.
Another device pair, the Fairchild PE 8050 and 8550 had lower beta but similar Rbb'. I used those for the Sota Head Amp in 1979, instead. They worked better with very low DC voltage across them.
Another device pair, the Fairchild PE 8050 and 8550 had lower beta but similar Rbb'. I used those for the Sota Head Amp in 1979, instead. They worked better with very low DC voltage across them.
I would sell them at a premium price to serious audiophiles TIAEK. They are worth much more than their original cost. They really are exceptional.
You parallel 20 .9nV op-amps thats .9nV/sqrt(20) add a little for a very small gain resistor and that's .22nV it's not rocket science, and possibly really only useful as a measurement amp (and relatively expensive).
BTW 2-IRF3600 JFET's in parallel run at about 80mA will easily do .2nV and not have Ib problems either. Head amps that only work with the most exotic 3 or 4 Ohm MC's also could be considered somewhat limited.
Last edited:
It is easier to parallel fewer 0.3nV devices like the Hitachi to get ultimate performance. The Hitachi are hi beta too!
For parallel use, hFE and VBE must be aligned --- that's why I got so many pcs.
It means you have to measure hFE and VBE for each and find the best pair, trio, quartet...
To find 20 pcs with the same hFE and VBE is beyond my thought.
You'll need much more than 1000 psc.
I guess 4 is the max number.
You have to consider you sum up caps if you put them in parallel.
It means you have to measure hFE and VBE for each and find the best pair, trio, quartet...
To find 20 pcs with the same hFE and VBE is beyond my thought.
You'll need much more than 1000 psc.
I guess 4 is the max number.
You have to consider you sum up caps if you put them in parallel.
Last edited:
Distortion at different rsource is not typically provided in specsheets, in which I always suspected manufacturers to select the strong points of their product. So, are the rest of the bunch like opa134 even worse?
😀
I was more getting at the 'no progress in the last few decades' quip. We could, at some point in the future, have systems around that would allow us to cool our circuits effectively and quietly, such that we can reduce thermal noise, without being subjected to the whine of a pump. Of course you could always place the pump in another room. Or cool your equipment from a cylinder of gas then process the recycle/recool, the used gas when you're not critically listening. That way you're not consuming the worlds helium supply either.
I've done this but not for a MC headamp 😀
To get more than 3dB advantage over SOTA today for MC headamp, you have to cool the MC cartridge too.
How about 2 devices in total and total consumption of 3mA ? .. which is my little MC headamp
The MicBuilders document points out that MCs today are all Golden Pinnae and LoZ.
Astro-photography?
Now that sounds perfect. One can envisage a supercooled, elevated magnetic bearing system. The only trouble is, it might interact with the cartridge itself.😱
Gerhard. But also from Germany.
I suppose you refer to my measurement preamp that I mentioned in
the blowthing thread some 3 weeks ago. I included all circuit diagrams
and at least a screen dump of the layout, photos of the real thing
and measurements done on it and with it. The essentials parts of the BOM
are all available at Digikey, nothing exotic. Bandwith from < 1Hz to 1 MHz,
no 2nF input capacitance per part unlike these fat FETs (not including Miller),
no transformers on "unobtainium" core material like N48 ( that Siemens/EPCOS
must have produced by tons a month btw) and that are just a BW limiter
at the lower end where it's most interesting.
Some people even got the leftover boards for experimenting.
It really measures more like 200pV/sqrt HZ, but including the feedback
network and Friis' formula that would require the AD products to be
somewhat better than their data sheet. But I'm used to that with AD.
If you do not want to search it:
< http://www.hoffmann-hochfrequenz.de/downloads/lono.pdf >
< http://www.diyaudio.com/forums/lounge/146693-john-curls-blowtorch-preamplifier-part-ii-5178.html >
< https://picasaweb.google.com/103357048842463945642/LowNoisePreamplifiers?authuser=0&feat=directlink >
regards, Gerhard
Last edited:
Thanks for this Gerhard.
My poor effort is nearly 3dB worse 😡
Have you ever used it as a MC headamp?
My poor effort is nearly 3dB worse 😡
Have you ever used it as a MC headamp?
Any more left-over boards, or would you be kind enough to share Gerber in private ?
Thanks for sharing your excellent work in any case.
Patrick
Thanks for sharing your excellent work in any case.
Patrick
Gerhard, I'm sorry if you took offense at my comments. They were intended as technical discussion, in hindsight could have worded my examples more carefully. If it matters I intended to appologize and clarify a little, but got busy and forgot.
I had intended to point out that when considering the source impedance, design choices/tradeoffs will differ.
I had intended to point out that when considering the source impedance, design choices/tradeoffs will differ.
Scott, I did not take any offense, no, not at all.
In fact, I have searched my pot core box if I could
find sth. that fits, but those that would be ok were
glued together already, no real help.
The selection of threads I read here is mostly
based on the fact that you participate, much
like Phil Hobbs or Larkins in sci.electronic.design. 🙂
I'm interested particularly in low frequencies because
in my daytime job I'm currently assigned to build a phase
comparator to regulate some atomic clocks, and they
should not drift fast from each other.
Euvl, there are no more leftover boards but you can get the
gerbers. I will do a re-design to get rid of the 1 dB droop
at 1 MHz, with 2 channels for cross correlation and RS-170
relays to automate the setup somewhat. And self-resetting
polyfuses. I have nearly burnt a 10-pack on that board. The
fat electrolytics are also pointless with the NiMH cells.
But the redesign may take some time. The weather says
that the motorbike season is opened, finally.
Proposals for the coupling C are most welcome. Must be easily
available, no snake oil please.
PMA, that picture shows clearly that I did not use the
appropriate tool to fasten the LEMO connector. 😱
Look at the fabrication date of the MAT02s, they must have
slept 25 years in my drawer, always too precious to use them.
BTW this amp is built on Vero powerplane board with a ground
mesh on the top side, one can use it even at VHF/UHF.
I'll run out of supply soon, I think Vero is no more alive?
There is an alternative from Roth, but not so nicely
plated-through, not pretinned and without stop mask.
KGRLEE, I have never used that for audio. A 5 Ohm MC
probably would have more thermal noise, so there is
no point to do it.
regards, Gerhard
In fact, I have searched my pot core box if I could
find sth. that fits, but those that would be ok were
glued together already, no real help.
The selection of threads I read here is mostly
based on the fact that you participate, much
like Phil Hobbs or Larkins in sci.electronic.design. 🙂
I'm interested particularly in low frequencies because
in my daytime job I'm currently assigned to build a phase
comparator to regulate some atomic clocks, and they
should not drift fast from each other.
Euvl, there are no more leftover boards but you can get the
gerbers. I will do a re-design to get rid of the 1 dB droop
at 1 MHz, with 2 channels for cross correlation and RS-170
relays to automate the setup somewhat. And self-resetting
polyfuses. I have nearly burnt a 10-pack on that board. The
fat electrolytics are also pointless with the NiMH cells.
But the redesign may take some time. The weather says
that the motorbike season is opened, finally.
Proposals for the coupling C are most welcome. Must be easily
available, no snake oil please.
PMA, that picture shows clearly that I did not use the
appropriate tool to fasten the LEMO connector. 😱
Look at the fabrication date of the MAT02s, they must have
slept 25 years in my drawer, always too precious to use them.
BTW this amp is built on Vero powerplane board with a ground
mesh on the top side, one can use it even at VHF/UHF.
I'll run out of supply soon, I think Vero is no more alive?
There is an alternative from Roth, but not so nicely
plated-through, not pretinned and without stop mask.
KGRLEE, I have never used that for audio. A 5 Ohm MC
probably would have more thermal noise, so there is
no point to do it.
regards, Gerhard
Last edited:
First of all, we can get pretty low noise with just paralleling bipolar transistors or jfets. Except for research projects, we can get to the equivalent of a 5 ohm resistor, fairly easily. Even lower, if we want. The problem with liquid helium or even liquid nitrogen is the complexity and cost. Why bother? What do YOU need less than 5 ohms noise for?
- Status
- Not open for further replies.