[bhjazz]

While I wait for some (more) parts for my Cavalli, I am working on hotrodding my ProJect phono preamp. The stock parts seem pretty generic, so I thought a major overhaul would get the performance up a bit. Plus, it'll be fun, right?

Here's the board.

The questions I have are regarding some of the parts.

1) All of the caps have a specific notation on them: The red caps, for instance, read, "330nJ 160V". Does this mean 330pF? I know the "J" means 5%. (geez, are they even caps.

Are they inductors???)

2) The flat, square caps near the top are even more cryptic: 68n Nq". Any guess what this means?

3) The small disc caps near the bottom read 10, 27, and 100. This is ohms, yes?

4) Would there be any reason to change the transistors? I have a 78L15 and a 79L15 TSL transistor.

Many thanks!

 

[error401]

Quote:

Originally Posted by bhjazz /img/forum/go_quote.gif 1) All of the caps have a specific notation on them: The red caps, for instance, read, "330nJ 160V". Does this mean 330pF? I know the "J" means 5%. (geez, are they even caps. Are they inductors???)

They definitely look like caps to me, and the value makes sense. These would be 330nF (0.33uF) caps rated for 160V. Probably MKT. The J does seem to indicate 5% tolerance.
Quote:

2) The flat, square caps near the top are even more cryptic: 68n Nq". Any guess what this means?

I'd guess 68nF, +/-30% tolerance, ceramic. Not sure what the q means. These seem to be small value filter caps for the voltage regulators. Value and type would not be critical.

Quote:

3) The small disc caps near the bottom read 10, 27, and 100. This is ohms, yes?

Unlikely, capacitance is measured in farads. These could be anything, small caps usually use a 3-digit code with the third digit being an exponent, value taken in pF, eg 100 = 10 x (10^0) = 10pF, which may be the value of your last cap. The two digit ones I'm not sure. Could be read literally as nF as these values would probably make sense, but really you'd have to measure them.
Quote:

4) Would there be any reason to change the transistors? I have a 78L15 and a 79L15 TSL transistor.

These are +15v and -15V voltage regulators, respectively. Probably not worth changing unless you choose to completely replace the power supply with a better design.

I seem some TL071 opamps in that circuit that would probably have a larger detrimental effect on the sound than any of the capacitors would. These are cheap chips that don't have the greatest sound, you might experiment with better quality parts in those positions. You might do the same with the JRC 4580's as well. I don't have a schematic, but I suspect these mods will make a bigger difference than swapping caps, it looks like they've used film caps in the filters so they shouldn't be colouring the sound too badly.

The white boxes are also caps, apparently MKT film, but I can't read the values from the photo...

 

[ericj]

Quote:

Originally Posted by bhjazz /img/forum/go_quote.gif While I wait for some (more) parts for my Cavalli, I am working on hotrodding my ProJect phono preamp. The stock parts seem pretty generic, so I thought a major overhaul would get the performance up a bit. Plus, it'll be fun, right?

Just want to say, Alex Cavalli has contributed to the designs of several headphone amps - so it could be confusing to refer to any one amp by his surname.

Quote:

Originally Posted by bhjazz /img/forum/go_quote.gif 3) The small disc caps near the bottom read 10, 27, and 100. This is ohms, yes?

Like error401 said, these are ceramic capacitors, measured in picofarads.

Ceramic disc caps are generally frowned upon in the signal path, but way down in the small dozens of picofarads pretty much everything wrong with ceramic caps isn't a problem anymore.

MLCCs or film caps are "better" but the difference would be tiny to nothing.

Sure are a lot of cheap crappy opamps in there for just a phono preamp.

TL071 = single channel fet-input opamp, very cheap.

JRC4580D = dual-channel bipolar-input opamp, bargain basement cheap.

You could replace these with any of the opamps we talk about here except for the super cranky parts - no ad8397, lm6172, etc.

Replacing these chips will have the most profound impact.

Install sockets so that you can try a variety of chips.

It's probably best to replace the tl071's with other FET parts, replace the 4580d with whatever - even a jrc2068 would be noticibly better.

Also, you could replace the electrolytic caps with good quality new caps - Panasonic FC or FM, or Nichicon UPW or UHE. Electrolytic capacitors have come a very long way over the last 20 years so replacing these caps, even if they haven't dried out, could represent a performance improvement.

Just get the same voltage and capacitance rating for each spot.

 

[bhjazz]

Many thanks, you two! It all makes more sense now.

Quote:

Originally Posted by error401 /img/forum/go_quote.gif Unlikely, capacitance is measured in farads.

Of course! Geez, I really rushed my post. Thanks!

Quote:

Originally Posted by error401 /img/forum/go_quote.gif I seem some TL071 opamps in that circuit that would probably have a larger detrimental effect on the sound than any of the capacitors would.

I've read a few things about the TL071, but not a lot. A few posts in other forums stated that it was an okay chip for this circuit. However, after reading your info, I may look for an alternative. There is definitely an alternative for the JRC 4580D: the OPA2134, which appears to be an easy, direct replacement.

Quote:

Originally Posted by error401 /img/forum/go_quote.gif The white boxes are also caps, apparently MKT film

They are, indeed. Mundorf rates them as "upper middle class" versions.



Quote:

Originally Posted by ericj /img/forum/go_quote.gif Just want to say, Alex Cavalli has contributed to the designs of several headphone amps - so it could be confusing to refer to any one amp by his surname.

You are correct, and I was not fair. I am working on the CK2-III. Credit where credit is due!

Quote:

Originally Posted by ericj /img/forum/go_quote.gif Also, you could replace the electrolytic caps with good quality new caps - Panasonic FC or FM, or Nichicon UPW or UHE.

Thank you, and a great idea!

For starters, then, I will look into changing the opamps, then move on to upgrading the electrolytic caps with some nice Panasonics. Thank you for providing me with a simple approach to my upgrade!

 

[bhjazz]

And...it looks like a simple chip upgrade would change the TL071 for the OPA134, and the 4580 for the OPA2134. Cool. Thanks!

 

[ericj]

Right, but, although there are many rebuild enthusiasts who think the opa134 series is the best thing ever . . . It's good but there are other options.

Like i said, socket 'em. So, some day, when you're feeling adventurous, you can plug in something else.

Nearly all DIP-8 single and dual opamps use the same single or dual pinout.

This design has a +/-15v supply, so, that excludes most of the Analog Devices opamps unless you feel like replacing the 7815 and 7915 with 7812 and 7912.

Other than voltage restrictions and the possibility of a DC offset issue if you replace the tl071 with a bipolar-input opamp, there are dozens of possibilities. There are a whole lot of audio-quality opamps out there, each with it's own subtly different sound signature, and you can have a lot of fun experimenting with the variations.

Most of the time, a phono preamp has a DC blocking capacitor on the output, even though it may have a FET opamp and a bipolar supply. Even so, before you hook up your modified preamp to anything, put a volt meter on the output and check for DC.

 

[bhjazz]

Excellent information. You da man! Thank you.
I'll definitely socket the parts, and will do a bit of research for other chips. What would you say about the OPA627? I have read some good things about it here, for sure.

Should I dig through the parts bin and replace the resistors? I have a huge bank of Vishays just sitting here...

Thanks again.

 

[ericj]

The opa627 is good, opa637 is better but you'd best trace out the circuit first and figure out what gain the tl071 is running at before trying to use one. the opa637 is the uncompensated version, needs a gain higher than 5. Since this is a phono preamp, the gain is probably way higher than 5, but, it's best to know for sure.

Thing is, those are $20 opamps, and i have a hard time justifying the cost. Even though i have the 627/637 combo in my M^3. I occasionally consider pulling them out and selling them, and going back to opa606 in all positions or something.

The opa604 and 2604 are used in a lot of good cd players. You might enjoy how they sound. it's a jfet input opamp so you can use them in all positions.

opa2107 is good for a dual op-amp. about $11. there isn't a single version, so you'd have to go for something else in the tl071 positions.

The lm4562 and it's single channel version the LME49710 are Really Good for the money and have a very neutral and detailed sound.

Edit: Oh, and i'm not a believer in upscale resistors. It looks like they used metal films throughout. Unless something looked burnt I'd leave them be.

 

[balou]

Replacing the TL071 with a OPA134 might not be a good idea. Of course I don't know the exact circuit of this preamp, but I don't think that they really put the TL071 inside the audio chain.
The TL071 most likely just serves as a DC servo. The OPA134 may have worse DC performance (it's a opamp specially designed for music after all). Nice reading about dc servos: http://www.head-fi.org/forums/showthread.php?t=169016

Quote:

and the 4580 for the OPA2134.

I really don't like the opa2134. limited soundstage... ericj made some good suggestions. I personally like the ad8066. the ad797 is a very low noise opamp with a high power supply noise rejection ratio, something that make him interesting in a high gain circuit like a phono preamp - but the lm4562 has similar specs, and it's sound is described as more pleasing (lm4562 'neutral', ad797 'cold'). But you shouldn't believe what some freaks on a freak forum say to you about opamps - test them out yourself

Quote:

Most of the time, a phono preamp has a DC blocking capacitor on the output, even though it may have a FET opamp and a bipolar supply.

The DC servo may have saved them from being forced to degrade sound quality by adding a dc blocking cap.


about the power supply: the 78lXX series of regulators aren't really that good - a standard lm317 provides way better regulation. and power supply quality is quite important in high gain amplifiers.


edit: Quote:

Edit: Oh, and i'm not a believer in upscale resistors. It looks like they used metal films throughout. Unless something looked burnt I'd leave them be.

how dare you not believing in the advantage of magic pixie dust resistors! something costing 3$ a piece MUST be better than those evil 'standard' products

 

[ericj]

Quote:

Originally Posted by balou /img/forum/go_quote.gif Replacing the TL071 with a OPA134 might not be a good idea. Of course I don't know the exact circuit of this preamp, but I don't think that they really put the TL071 inside the audio chain. The TL071 most likely just serves as a DC servo. The OPA134 may have worse DC performance (it's a opamp specially designed for music after all). Nice reading about dc servos: http://www.head-fi.org/forums/showthread.php?t=169016

True, it could be. It's odd to see six channels of opamps in a phono preamp - my dinky little NAD just has just one dual channel opamp.

fwiw the CKKIII has an opa2134 doing DC servo duty, as did the original Kumisa III design.

The 134 series is 2nd bannana to the 132 series, which has better DC specs than the 134, so, there is that, but people seem to find the 2134 does good work in a DC servo.

Googling for a schematic might help us make recommendations here.

Quote:

I really don't like the opa2134. limited soundstage... ericj made some good suggestions. I personally like the ad8066. the ad797 is a very low noise opamp with a high power supply noise rejection ratio, something that make him interesting in a high gain circuit like a phono preamp - but the lm4562 has similar specs, and it's sound is described as more pleasing (lm4562 'neutral', ad797 'cold'). But you shouldn't believe what some freaks on a freak forum say to you about opamps - test them out yourself

My little NAD preamp came with a Ti branded ne5532, which probably mops the floor with the JRC chips in this one. The opa2107 is my favorite opamp in most circuits but i have to say i liked the ad8066 better in my phono preamp.

In this circuit, though, you'd have to swap out the regulators for 12v parts to run an ad8066 or most Analog Devices parts - the ad8066 maxes out at +/-12v, the ad8620 at iirc +/-13.5v.

I don't know why the voltage tolerances for the AD parts are typically lower - they just are. fwiw the ad8620 sounds thin and strained at 26 volts and better at 24 or 22 volts.

Conversely, the opa2107 loves voltage and sings at 30v (maxes out at 36v), so might sound real good in this circuit.

Quote:

about the power supply: the 78lXX series of regulators aren't really that good - a standard lm317 provides way better regulation. and power supply quality is quite important in high gain amplifiers.

Properly bypassed, they really aint bad, but an ambitious rebuilder might replace the 7815 and 7915 with an lm317 and lm337 on a daughterboard and 'air wire' them into the old regulator spots.

But it wouldn't be the first second or third thing I'd do to this board.

I might just make sure there's a .01uf cap right before each regulator and a .1uf cap right after, just for noise filtering.

Or just not worry about it.

fwiw the power supply in my NAD PP-1 is an lm317 configured as a current source rather than a voltage regulator followed by a cmoy-esq rail splitter. It really works fairly well at these tiny power levels.

 

[balou]

Quote:

fwiw the CKKIII has an opa2134 doing DC servo duty, as did the original Kumisa III design.

Interesting. Well, at least I now know how to recycle my OPA2132 and *34 chips

(must resist temptation designing new amp before even finishing last one...)
There's a ton of infos about which opamp to use as voltage amplifier, but almost none on dc servos. well, maybe it's really not that important.

Quote:

I don't know why the voltage tolerances for the AD parts are typically lower

I think because for audio, nobody really needs +-15V (voltage-swing wise). But low voltage is needed for portable applications. So it looks like they optimized it for low to middle voltages.

Quote:

fwiw the power supply in my NAD PP-1 is an lm317 configured as a current source rather than a voltage regulator followed by a cmoy-esq rail splitter

that's rather odd. I know from the JFET rail isolation that this can dampen the power supply noise a bit. ok, a preamp should be a much easier load than a power amplifier (be it headphone or speaker). but still... strange (meaning I don't understand the why and how

)

 

[ericj]

Quote:

Originally Posted by balou /img/forum/go_quote.gif that's rather odd. I know from the JFET rail isolation that this can dampen the power supply noise a bit. ok, a preamp should be a much easier load than a power amplifier (be it headphone or speaker). but still... strange (meaning I don't understand the why and how )

Yeah, it confused a lot of people, because some magazine printed the schematic without permission and misinterpreted it.

So if you want an easy/cheap RIAA preamp schematic, google for "nad pp-1 review" and look at the online pdf.

At any rate, NAD said that they did it that way because it allowed them flexibility with the wallwart in several markets.

It's still regulating, because it's still regulating the 1.25v between out and adjust. It's just not regulating the voltage.

They also have a pair of 17v zeners in parallel with the resistor divider, probably to ensure that the ne5532 never sees more than +/-17v.

It's a fairly clever design.

One of these days i need to trace out the Yaesu DC filtering circuit i've got. For their hand-held radios, they have an active filter that drops about 1.5v, conveniently stuck to the end of a cigarette lighter plug. smoothing and filtering w/o actually regulating has some merit.

 

[balou]

um.. it uses electrolytic input and output caps.. and the reviewer measured 7mV ripple on the power rails. Well, I'd rather build an El Cheapo or VSPS phono amp.

Quote:

It's a fairly clever design.

I maybe believe that if someone can tell me how exactly the power supply works and why it should be better despite it's relatively high ripple it gives