A couple of quick circuits

[Douglas Blake]

Originally Posted by rsavas

Thanks for sharing the Laplace directive usage too.

You're welcome. I'm always happy to share the (non secret) stuff.

[Mark Tillotson]

Originally Posted by Douglas Blake

It drove me nuts getting that equation right (which might explain quite a lot, actually )

Here are the results from my circuit in the first post. Gotta love that nice straight green output line. It tells me that the recording and playback curves are near perfectly complimentary.

My RIAA simulation is based on this approach, which separates out all the zeroes and poles into separate RC sections, so in theory its a perfect match to the inverse transform (ignoring opamp non-idealities and compensation components):


Followed by a four-pole Butterworth rumble filter using Sallen Key stages.

[Douglas Blake]

Originally Posted by Mark Tillotson

My RIAA simulation is based on this approach, which separates out all the zeroes and poles into separate RC sections, so in theory its a perfect match to the inverse transform (ignoring opamp non-idealities and compensation components):


Followed by a four-pole Butterworth rumble filter using Sallen Key stages.

Nice! I've seen similar configurations, including all passive versions using FETs for gain. It's one of the most designed circuits I know of.

Me, I'm old school. The design I posted started out with a ua741 way back in 1980 or so. I realize it needs some touches, which is why I asked for comments, but it was a working circuit at one time.

To be honest, I've been thinking about buying a turntable with the pre-amp built in. There are advantages to keeping the gain close to the transducer.

My personal thing is doing the most with the least. When studying electronics in school our teacher always hammered at us... "First you get it to work at all, then you get it to work right, and finally you take out everything that isn't needed to make it work." ... Guy hasn't been wrong yet. LOL.

[sgrossklass]

Originally Posted by Douglas Blake

Me, I'm old school. The design I posted started out with a ua741 way back in 1980 or so. I realize it needs some touches, which is why I asked for comments, but it was a working circuit at one time.

It did look like something that started life with a 4558 or similar...

µA741, eh? Had to look for a bit until I found some clues to the noise level of that one. The dual version MC1458 as made by Philips was given with 30 nV/√(Hz). Better than an LM358, but still quite a bit worse than a 4558.

If it's any consolation, P06 is considerably noisier for what I suspect is a similar kind of history. (The entire feedback network of the first stage is about an order of magnitude too high in impedance.)

Your design is about on par with an average 4558-based typical integrated amp circuit of the 1980s in terms of noise and no doubt better in distortion.

Originally Posted by Douglas Blake

To be honest, I've been thinking about buying a turntable with the pre-amp built in. There are advantages to keeping the gain close to the transducer.

Certainly, especially when you are a fan of Audio Technica MM cartridges. Not a few of those want 47 kOhm || 100-200 pF, cable included (100-150 pF, typ).

Integrated phonopres weren't usually very fancy though, at least back in the '70s.

Originally Posted by Douglas Blake

My personal thing is doing the most with the least. When studying electronics in school our teacher always hammered at us... "First you get it to work at all, then you get it to work right, and finally you take out everything that isn't needed to make it work." ... Guy hasn't been wrong yet. LOL.

Basically it isn't wrong - but you have to figure in decades worth of experience, refinement and competition. When "good enough" is good enough, don't be surprised if you get steamrolled. You only need to look at how the Japanese turned several industries upside down.

Nowadays there are helpful websites, helpful forums such as the Analogue Source department next door, and you can access schematics for numerous commercial devices easily (*cough* HiFiEngine *cough*). Phono preamps have been a solved problem for 3 decades.

If you fancy a single-chip design, this is what has been gracing Yamaha's more down to earth integrated amps for years:


This is likely to be as low noise as they typically come for MM, short of going for a low-noise FET input stage. (This opamp seems to be at least on par with an NE5534A in the voltage noise department, and possibly current noise as well given the moderate input bias current spec.) Knowing the NJM2068's somewhat limited output driving abilities (less robust than the NE5532's), I would imagine that distortion performance could be improved by including a discrete Class A buffer stage in the loop, headphone amp style. With 280 ohms worst-case, it doesn't have to be super-duper powerful, a pair of BC639/640 (or 2N5551/5401) at ~5 mA or so seems easily sufficient. (You can go AB with some high-beta transistors if you insist, maybe BC550/560 or KSC1845/KSA992.) One might even consider going single-ended for simplicity's sake. I could draw up a schematic later if you're interested.

[Douglas Blake]

Originally Posted by sgrossklass

It did look like something that started life with a 4558 or similar...

µA741, eh? Had to look for a bit until I found some clues to the noise level of that one. The dual version MC1458 as made by Philips was given with 30 nV/√(Hz). Better than an LM358, but still quite a bit worse than a 4558.

If it's any consolation, P06 is considerably noisier for what I suspect is a similar kind of history. (The entire feedback network of the first stage is about an order of magnitude too high in impedance.)

Your design is about on par with an average 4558-based typical integrated amp circuit of the 1980s in terms of noise and no doubt better in distortion.

Certainly, especially when you are a fan of Audio Technica MM cartridges. Not a few of those want 47 kOhm || 100-200 pF, cable included (100-150 pF, typ).

Integrated phonopres weren't usually very fancy though, at least back in the '70s.

Basically it isn't wrong - but you have to figure in decades worth of experience, refinement and competition. When "good enough" is good enough, don't be surprised if you get steamrolled. You only need to look at how the Japanese turned several industries upside down.

Nowadays there are helpful websites, helpful forums such as the Analogue Source department next door, and you can access schematics for numerous commercial devices easily (*cough* HiFiEngine *cough*). Phono preamps have been a solved problem for 3 decades.

If you fancy a single-chip design, this is what has been gracing Yamaha's more down to earth integrated amps for years:


This is likely to be as low noise as they typically come for MM, short of going for a low-noise FET input stage. (This opamp seems to be at least on par with an NE5534A in the voltage noise department, and possibly current noise as well given the moderate input bias current spec.) Knowing the NJM2068's somewhat limited output driving abilities (less robust than the NE5532's), I would imagine that distortion performance could be improved by including a discrete Class A buffer stage in the loop, headphone amp style. With 280 ohms worst-case, it doesn't have to be super-duper powerful, a pair of BC639/640 (or 2N5551/5401) at ~5 mA or so seems easily sufficient. (You can go AB with some high-beta transistors if you insist, maybe BC550/560 or KSC1845/KSA992.) One might even consider going single-ended for simplicity's sake. I could draw up a schematic later if you're interested.

Ummm ... Okay.

[rsavas]

Many possibilities, I posted this jfet composite in another thread, it is from the LSK489 application note. I did discover that I think the 75uS comps are incorrect, C4 should be 22n. If you find lsk489 too expensive you could sub low cost jfets. A DC servo could be used to eliminate C5 and o/p ac coupling cap.
Even the new TI OPA1656 can be used. I called it the new NE5532

[rsavas]

Looking further,
I discovered that TI sells a DIP-ADAPTER-EVM, Digi-key has them, I just added one to my order.
If you read further in the OPA1656 DS, it has great application information for you, wonderful stuff.