You want a circuit for low output moving coil? Here you go
http://www.diyaudio.com/forums/analogue-source/289428-phobos-balanced-mc-phono-stage.html
I have built one and with my Monster Alpha Genesis (0.24mV output) it has plenty of gain. And by plenty of gain, I mean that output gets to around 2V to match the output of CD player. 2 stages only, not a single wasted component.
And I think that you are are on the right track. If you have a MC cartridge then the right thing to do is build a phono stage optimised for a MC cartridge. I don't see the point of building a phono stage optimised for MM, then adding a transformer to make it do something else. So, I already have an MC pre-amp, and a strain gauge pre-amp (DIY design which is elsewhere on DIYAudio) and soon I will have a separate MM pre-amp. It will have a j-fet (2SK369) input but follow a similar topology. The right tool for the right job.
It's all in the thread including measurements.
The prime reason is distance ... the same reason you use balanced outputs on mikes and proper balanced inputs on mike preamps.
If you are using your cartridge into an unbalanced preamp (like most people), you WILL get hum .. only the level is open to discussion. With a proper balanced input fed by the usual floating cartridge, you can achieve NO hum.
See Wayne's thread for detail. Also big advantages in RFI.
Caveats re magnetic pickup in the coils as per Self bla bla
Has anyone measured the effect of dc on the MM coil. Like direct coupling a NE5532 with 200nA bias current ( 800nA max!). FFT at say 1Khz with a very low bias current ( FET input opamp) and with a NE5532 direct coupled.
Lots of opinions on this one.
If the FFT's are OK, then they should also sound the same ?
Lots of opinions on this one.
If the FFT's are OK, then they should also sound the same ?
Floating input
In his Wireless World article "Distorson in low-noise amplifiers" (August 1977, page 30, Fig. 5 and September 1977, page 59, Fig. 9) E.F. Taylor proposed an MM RIAA equalized preamplifer having a floating input inserted in the feedback path. Pictures below, comments welcome. Does anybody know if this intriguing circuit has been used somewhere ?
In his Wireless World article "Distorson in low-noise amplifiers" (August 1977, page 30, Fig. 5 and September 1977, page 59, Fig. 9) E.F. Taylor proposed an MM RIAA equalized preamplifer having a floating input inserted in the feedback path. Pictures below, comments welcome. Does anybody know if this intriguing circuit has been used somewhere ?
Attachments
No I'm afraid not, but some time ago I stumpled across an application note from National Semi called "Audio Applications of Linear Integrated Circuits" AN299 where they used a LM318 in an unusual way (see attatchment).
Unfortunately the LM394 is made of unobtainium as far as I know, but I guess you could use something like 2SA970 instead.
Unfortunately the LM394 is made of unobtainium as far as I know, but I guess you could use something like 2SA970 instead.
Attachments
I know all that, just wondering why Doug does not agree. Searching the web I also find a plethora of directly coupled phono pre-amps. Example from Nakamichi below, in fact I would think the coupling cap would be an exception once dual rails were common.
Attachments
Last edited:
This seems to lead to the forum but not the thread.
I would be interested to know why people feel a balanced MM input is superior.
Right at the moment, I don't see a good reason.
Mic cables tend to be metres long. MM cables not so much. It is true that balanced connections are almost universal in pro audio, but this is because you are typically connecting equipment that is connected to ground, and 50Hz ground currents can flow. Balanced connections deal with this superbly. But this does not seem to me relevant when you are connected to a cartridge coil that is wholly floating.
Not true in my experience. Unmeasurable hum is possible if the ambient magnetic fields are low enough.
Would love to, but as noted the link seems to be wrong.
No. Not unless you can arrange for precisely symmetrical demodulation of RF, which you can't.
a link to a discussion of pros and cons of flat vs low pass pre?
sorry, but from EE perspective the claim that the flat preamp is "better" is simply wrong re click/pop dynamic range performance - low pass filtering is much the better route to avoid clipping the pre out with the hf edges of clicks, pop, mistracking
if the "problem" is analog preamp low pass exact corner frequency mismatch with digital the rest of the RIAA implementation then there is the obvious option of calibrating your preamp filter and adjusting a single DSP Biquad's coefficients accordingly
is simply wrong in all terms relevant to RIAA correction for ultimate human listening to recorded music playback - no recording/mastering/production/playback/room/speaker chain is within orders of magnitude of the flatness/accuracy where these "impossible to remove wriggles" could be seen - and they can be corrected to much better than implied - way down into the noise of the system at human hearing integration times
with just a PC motherboard soundcard calibrating a low/mid audio single pole lowpass should be better than any purchasable or even hand trimmed C tolerances - if you don't get very much better than 0.1 % repeatability then you aren't trying - and you do get to use long averaging in the calibration
what makes this easy, accurate is that we want to measure a frequency domain single pole filter's combined RC time constant, not accurately measure either R or C
and hi resolution of relative amplitude and precise time are the strong points of digital audio, even the cheapest PC motherboard chipsets, much less the "prosumer" audio soundcard you might use for digitizing vinyl at the quality level where you start to care about the analog pole tolerances
the soundcard Xtal accuracy is the only real limitation – and 100 ppm absolute accuracy is expected from cheap parts
I'd even expect to be able to see the polystyrene 120 ppm/C tempco with a soundcard RC mid audio time constant measurement and my indoor temp range – want to put a thermistor or bandgap T sensor in the preamp too?
the DSP correction isn't "loads more processing" - it is a single DSP Biquad section per analog pole to be corrected and the use case is correcting a single analog pole in a low pass phono preamp
all of that before even considering music production practice limits to RIAA real accuracy, or human hearing thresholds for sub 0.1 dB frequency “wiggles”
sorry, but from EE perspective the claim that the flat preamp is "better" is simply wrong re click/pop dynamic range performance - low pass filtering is much the better route to avoid clipping the pre out with the hf edges of clicks, pop, mistracking
if the "problem" is analog preamp low pass exact corner frequency mismatch with digital the rest of the RIAA implementation then there is the obvious option of calibrating your preamp filter and adjusting a single DSP Biquad's coefficients accordingly
is simply wrong in all terms relevant to RIAA correction for ultimate human listening to recorded music playback - no recording/mastering/production/playback/room/speaker chain is within orders of magnitude of the flatness/accuracy where these "impossible to remove wriggles" could be seen - and they can be corrected to much better than implied - way down into the noise of the system at human hearing integration times
with just a PC motherboard soundcard calibrating a low/mid audio single pole lowpass should be better than any purchasable or even hand trimmed C tolerances - if you don't get very much better than 0.1 % repeatability then you aren't trying - and you do get to use long averaging in the calibration
what makes this easy, accurate is that we want to measure a frequency domain single pole filter's combined RC time constant, not accurately measure either R or C
and hi resolution of relative amplitude and precise time are the strong points of digital audio, even the cheapest PC motherboard chipsets, much less the "prosumer" audio soundcard you might use for digitizing vinyl at the quality level where you start to care about the analog pole tolerances
the soundcard Xtal accuracy is the only real limitation – and 100 ppm absolute accuracy is expected from cheap parts
I'd even expect to be able to see the polystyrene 120 ppm/C tempco with a soundcard RC mid audio time constant measurement and my indoor temp range – want to put a thermistor or bandgap T sensor in the preamp too?
the DSP correction isn't "loads more processing" - it is a single DSP Biquad section per analog pole to be corrected and the use case is correcting a single analog pole in a low pass phono preamp
all of that before even considering music production practice limits to RIAA real accuracy, or human hearing thresholds for sub 0.1 dB frequency “wiggles”
Last edited:
Don't get too hung up on term. It's commonly used with microphones in the pro world. Dynamic mics also have a floating coil, and yet the circuit is commonly called balanced in the pro world. Ditto for mics with a transformer output. If one is writing about it, it could be mentioned that it is actually a differential circuit but also commonly refereed to as balanced, and explain the difference.
I'll echo what others have said in the thread. Talk about cartridge loading, including step up transformers for MC carts. Keeping noise low, both from the inputs and from the tonearm wiring. Different EQ methods, active, passive, split. Basic stuff that is important.
Mini-bible here on the restoration of damaged audio. http://www2.ece.ohio-state.edu/~schniter/ee597/handouts/restoration_chapter.pdf
The tick and pop problem was solved a while ago. I do have to disagree, autoregressive interpolation does appear to do a better job when each pop does not look like the impulse response of the RIAA network. Finding the end of the pop becomes more difficult too.
BTW Doug that would be a good citation for any book on vinyl. If you want to be up to date a thoughtful discussion of the analog/digital possibilities should be there.
Last edited:
> National Semi ... used a LM318 in an unusual way
That is dead-conventional. LM318 has a speedy output section but poor (for audio) input stage. The Offset pins 1 5 are also all-purpose inputs to the internal gain. The differential pair becomes the new "input section"; the stock inputs disabled by pulling them to V-. The result IS an op-amp, with '394 input characteristics and '318 output.
You could do 99% the same with a '5534. The '318 output may still be faster but how fast can a needle slew? The '394 can potentially be bias-adjusted for best OSI noise-fit, though between '5534 and '394 at 100uA the difference is below vinyl-hiss.
This would also work with many low-noise and even jellybean transistors, though the DC gain of 550 makes 2mV offset a significant 1V output DC. (IMHO you really want a 20Hz pole.)
There are alternatives to MAT and '394 pairs. THAT Corp will sell you some very fine audio-aimed low-hiss matched pairs.
That is dead-conventional. LM318 has a speedy output section but poor (for audio) input stage. The Offset pins 1 5 are also all-purpose inputs to the internal gain. The differential pair becomes the new "input section"; the stock inputs disabled by pulling them to V-. The result IS an op-amp, with '394 input characteristics and '318 output.
You could do 99% the same with a '5534. The '318 output may still be faster but how fast can a needle slew? The '394 can potentially be bias-adjusted for best OSI noise-fit, though between '5534 and '394 at 100uA the difference is below vinyl-hiss.
This would also work with many low-noise and even jellybean transistors, though the DC gain of 550 makes 2mV offset a significant 1V output DC. (IMHO you really want a 20Hz pole.)
There are alternatives to MAT and '394 pairs. THAT Corp will sell you some very fine audio-aimed low-hiss matched pairs.
Perhaps you should re-read what I wrote about HF clipping and its consequences.
All granted. I'm a big fan of Bi-quads if you check out what I've published this Millenium.
And I'm such a cheapskate that I'd use 1% caps instead of 0.1%
Err.rh! These 'wriggles' are in the TIME domain. We are talking about removing clicks & pops. I've some experience trying to do this by analogue means in da last Millenium.
For the RIAA EQ, the obvious way is to do it analogue so you can claim your bits are hand carved from Unobtainium by virgins. Difficult to apply a similar cache to Bi-quads.
Thank you Scottscott wurcer said:
___________________
Pro Audio Design Forum • View topic - Flat Phono Preamp Based on John's P10 and 2SK389kgrlee said:
This is the actual beast. The de-clicking stuff appears towards the end.
Pro Audio Design Forum • View topic - A Low Noise Balanced In Moving Coil Preamp Using the ZTX851
This one has loadsa stuff on MCs including a long winded discussion with a wannabe Golden Pinnae. We are awaiting his DBLT results.
___________________
Doug, what about a simple (2 x BJT) circuit to emulate the THD profile of vinyl? I dreamt this up in da previous Millenium while playing with the first of the available good digital recorders, the Sony PCM-F1.
Makes EVIL digital sound like vinyl
Last edited:
Why would you want that...oh wait, some might, including some DJs' I know......Makes EVIL digital sound like vinyl
I already have this capability at will in my filters, thing is I don't like it, adds a bass lump and subdues the highs, brings out the mids too.
Dan.
You mean EQ which will give a vinyl sound?
What I discovered is quite different. It replicates the THD profile of vinyl via a good cartridge which goes UP at 18dB/8ve (mostly 2nd) and also with level. Freq. resp. etc is unchanged.
Adds vinyl hiss too
The added (mostly HF) THD makes it sound surprisingly warm, cuddly, sexy bla bla
What it doesn't do is allow for the change between inner & outer grooves.
For the RIAA EQ, the obvious way is to do it analogue so you can claim your bits are hand carved from Unobtainium by virgins. Difficult to apply a similar cache to Bi-quads.
And you apply your hand carved analogue RIAA EQ AFTER de-clicking digitally of course
[url]http://www.diyaudio.com/forums/analogue-source/298896-digitizing-vinyl.html[/URL] on this forum
covers the 'same' ground. It's a huge thread and all our heavy DSP hitters are present including Guru Wurcer.
I've only glanced through but it appears to converge on Wayne's solution ... at least as far as the strategy goes ... but Wayne does his EQ post de-clicking in EVIL digital
Last edited:
Thanks both of you gentlemen.
---
hazard500,
Actually I am interested in all aspects of turntables. Moving Coil was just part of various aspects. As of now I have MM DIY phono kit (CNC Phonostage) gifted by a friend (Sachin) which needs to be completed. Pardon me for littlle offtopic. But thanks again
Regards.
Getting back to balanced MM inputs, I would like to hear what people think about the noise situation. Seems to me it must be worse.
In a balanced input the 47k load will be split into 2 x 23.5k, which will give less Johnson and less effect from current noise, but the voltage noise will be unaffected and there are now two uncorrelated sources of it.
Using cart parameters of 600R and 500mH, and 5534 opamps, I calculate that the balanced noise will be 1.1 dB higher.
That would be quite acceptable if balanced operation solved other problems, but I've yet to hear any convincing argument that it does.
In a balanced input the 47k load will be split into 2 x 23.5k, which will give less Johnson and less effect from current noise, but the voltage noise will be unaffected and there are now two uncorrelated sources of it.
Using cart parameters of 600R and 500mH, and 5534 opamps, I calculate that the balanced noise will be 1.1 dB higher.
That would be quite acceptable if balanced operation solved other problems, but I've yet to hear any convincing argument that it does.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.