BTW, Wayne identifies all the good LN 'instrumentation amp' type topologies as originating from Demrow in Jurassic times, including his own, John Roberts' and Graeme Cohen's supa LN mike preamp. SSI/TI/THAT too if you allow some license.See the thread for details.
There must be modern FET instrumentation amps with low (<3.5nV/rt(Hz) ) Ein which would be even better for MM. Any suggestions Guru Wurcer?
__________________________
It would be easier with a new THAT IC and a few external parts. I designed a similar one for the Naval Ordinance Labs in Bethesda MD I'll post it here (with parts you can get now) later this weekend.
An interesting question for Doug. What about new IC's that come only in leadless packages, are they worth using in stuff aimed at DIY?
BTW Bob Demrow was our apps guy and that design was put in a module by Analog before they made IC's. I inherited it and made the first all in one IC version. Unfortunately Bob has passed away and I could never ask him where the original idea came from and Walt and I were the ones that made sure he got some credit.
Last edited:
An op amp example - http://twistedpearaudio.com/docs/linestages/retro_hiz_schematic.pdf
A tube example - http://www.tubecad.com/2007/09/29/balanced input & output phono 2.png
From - Still More Phono Preamps
Last edited:
Understood. Thank you very much for the info.
Might be better to apply the RIAA HF-cut in the first stage ( cap across each feedback resistor) as that will give much better overload margin at HF.
The final stage looks a bit complicated; is it a quasi-floating balanced output?
Don't see why not if you can get suitable adapter PCBs.
But I'll have nothing to do with BGAs, thank you.
Yes, with common mode suppression by using the OPA1632. The input devices are OPA1642.
The suppression circuit can be deleted by grounding the Vocm pin.
Last edited:
Try the links in post #74
Yes please Guru WurcerScott Wurcer said:
_________________
and Doug, I'm sure there's loads of gurus (and pseudo gurus) here that want to know how you got unmeasurable hum in a phono playback system with unbalanced leads & preamp.
OK, so 16 lead QFN is OK there are multi-sourced QFN to DIP adapters. Here is the part which replaces the first stage on that Twistedpear schematic, keep in mind the input, being an in-amp is only a diff-pair at the input and because Les and company know audio there is NO Ib comp. So you have a 1nV noise input with as little current noise as is practical. Enough internal nodes are available to do the old trick that worked on the 5534 where you disable the input pair and substitute an external JFET pair should work too.
Attachments
Trying to compare specs on those. Headline for the OPA1642 is 0.8fA/rtHz , but THAT very helpfully give a handy graph of noise against gain and source impedance which is rather more useful. But then again looking at the obligatory phono stage design TI give there is a noise floor plot.
I think they are neck and neck unless I have missed something ( and ignoring 1 chip vs 3 etc).
I think they are neck and neck unless I have missed something ( and ignoring 1 chip vs 3 etc).
Sorry I misread the DS there is Ib comp but there is also a 1583 with lower Ib noise.
But "mechanical stuff" is a necessary key point.
Just to pick one : channel separation.
You don't feel the magical breath of some classical music when the cartridge makes just 20 db channel separation and if this is moreover frequency and amplitude dependent
While some slight deviation from RIAA is not a big thing as the listening room acoustical properties will determine the frequency characteristics to a higher extend than even a bad phonostage could.
Ok. headphones is always better than speakers but then you will easily hear
the impacts of "mechanical stuff"
In terms of electronics aspects : power supply. The least effort : PS is battery.
Is discharge of a rechargeable battery "noisy"?
It should be as discharge is "statistical".
In medical diagnosis devices such as EEG battery supply of instrumentation amps is or was state of technology. How was that done? I was informed just to short the battery poles with some capacitors would to the job to eliminate discharge nose.
Is discharge of a rechargeable battery "noisy"?
It should be as discharge is "statistical".
In medical diagnosis devices such as EEG battery supply of instrumentation amps is or was state of technology. How was that done? I was informed just to short the battery poles with some capacitors would to the job to eliminate discharge nose.
There is a nice study often linked here, there are batteries at the .1nV level of noise. Even in the presence of shot noise the key is the ESR of some batteries is extremely low. The study shows extreme shot noise suppression in some cases which for some is a difficult concept.
http://www.mikrocontroller.net/attachment/100819/Measurement_of_Chemical_Battery_noise.pdf
http://www.mikrocontroller.net/attachment/100819/Measurement_of_Chemical_Battery_noise.pdf
Last edited:
So maybe the designers of sensitive test equipment like null detectors (Fluke 845AB or HP 419A) knew something after all. IMO, battery power is most useful for eliminating ground loops. OTOH, rechargable batteries are a PITA and the old ones invariably need to be replaced, usually by shoehorning in something not quite original. It would be interesting to do those same noise tests on the current best practices with cap multipliers or even optically isolated power supplies.
OK here is my final take, after consulting Mr Vogel, on whether balanced MM inputs are of any use or not.
1) Electrical fields into cartridge. Any sensible cartridge is electrically shielded, so balancing not required. For dumb unscreened cartridges, the coupling will not be identical for the 2 channels so balancing of very doubtful benefit.
2) Magnetic fields into cartridge. These will cause a voltage across the floating cart coil and will not be rejected by a balanced input.
3) Coupling into cable. Negligible with usual cable lengths and half-sensible cable layout. I can't see any source for a common-mode signal. Balanced input therefore not required.
4) A balanced input need not be more than 1.1 dB noisier than an unbalanced output. May be possible to reduce that difference by using in-amps.
Any objections?
1) Electrical fields into cartridge. Any sensible cartridge is electrically shielded, so balancing not required. For dumb unscreened cartridges, the coupling will not be identical for the 2 channels so balancing of very doubtful benefit.
2) Magnetic fields into cartridge. These will cause a voltage across the floating cart coil and will not be rejected by a balanced input.
3) Coupling into cable. Negligible with usual cable lengths and half-sensible cable layout. I can't see any source for a common-mode signal. Balanced input therefore not required.
4) A balanced input need not be more than 1.1 dB noisier than an unbalanced output. May be possible to reduce that difference by using in-amps.
Any objections?
It's your call, for lack of some hard data/measurements from real world equipment I don't see you changing your mind. It would be nice to have the information out there, no industrial instrument (classic geophones for oil search) operating on the same principles would ever be done any other way.
I hope it is not time to hang up your hat when your favorite device is only available in BGA or DFN. It's a coming, as volume production dictates the package selection.
An adapter for a DFN, so much for a low inductance package. How are you going to properly decouple the supplies?
I see Elektor has come out with yet again another phono pre-amp,Supra, 4 of lt028 in parallel, using some old lm833.
I guess it is hard to comment on a design that we can not show without permission.
Does Doug think these MLCC COG grade parts belong in audio? Any tests to compare MLCC COG vs polystyrene?
I still have to get/read the LA article on the cap measurement bridge. Glad to see that Jan is selling specific articles online. Maybe Santa will drop it off to me.
An adapter for a DFN, so much for a low inductance package. How are you going to properly decouple the supplies?
I see Elektor has come out with yet again another phono pre-amp,Supra, 4 of lt028 in parallel, using some old lm833.
I guess it is hard to comment on a design that we can not show without permission.
Still dumping on MLCC, although no mention of cheap/available, tight tolerance NPO/COG grade devices.
Does Doug think these MLCC COG grade parts belong in audio? Any tests to compare MLCC COG vs polystyrene?
I still have to get/read the LA article on the cap measurement bridge. Glad to see that Jan is selling specific articles online. Maybe Santa will drop it off to me.
Last edited:
I have a concern about the cartridge screening in this model. most MM are wrapped in some form of metal (Grados aside). They are then generally glued into a plastic case which is bolted to the (generally metal) headshell and tonearm. The tonearm is generally grounded by a seperate wire to the preamp chassis (except Rega where its one of the screens of the cables to the preamp. The cartridge screen ground is connected via a tag to one of the signal grounds (no standard on that).
Given the average turntable has two sources of E-fields (transformer and motor) it does seem on initial inspection that the standard wiring practice is suboptimal for getting hum free performance*. Even if you can get things low enough that an IA input is not audibly necessary it could require some work on the stock wiring. A discussion of wiring and grounding practice is certainly vital to a discussion of high performance vinyl.
*To put it another way, you would be unlikely to get sign off for that sort of arrangement in most fields.
Given the average turntable has two sources of E-fields (transformer and motor) it does seem on initial inspection that the standard wiring practice is suboptimal for getting hum free performance*. Even if you can get things low enough that an IA input is not audibly necessary it could require some work on the stock wiring. A discussion of wiring and grounding practice is certainly vital to a discussion of high performance vinyl.
*To put it another way, you would be unlikely to get sign off for that sort of arrangement in most fields.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.