Headphone stage & preamp output

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Rioc

Post by carlmart »

Ricardo,

You may be right about the 2113, as I would have to isolate the output jack, though you are not about your insistence in using every chance you can in pointing my using 12v single supply as my limitation. It is not.

Please acknowledge that is just your opinion, not a fact. You may name ten reasons why I shouldn't use one, and I may name another ten why I should.

Audio quality won't be affected and I will have plenty of level for the inputs I will be feeding. I may not claim the levels some portable units, using DC-DC converters, get to have, but I don't really need them. My only (minor) boost is for the 48v. A proper DC-DC supply would bring more problems than solutions to my project, the main ones being higher cost (which I don't want), space (which I don't have) and RFI (which I would need to filter, both electronically and (physically).

Nagras, which you probably know about, worked with -10v discrete regulated supply from 18v, and were considered one of the best sounding units ever made. Their universal mic preamps didn't even use transformers, just inductors on their balanced input, which was also discrete. Their 48v supply was discrete, using a long capacitor-diode chain after an AC generator.

In fact I will raise the issue, with my EE partner, on doing one of the prototypes without 12v regulation at all. THAT had told me I didn't need it, as I will always be feeding from NimH or Li-Ion 14.4v batteries. By not using a regulator I may feed with 12v batteries too, same technology and more widely available.

What NWAVguy pages are those? You mean using two chips in parallel? That would be better than the output boost I suggested?

The only option I wonder for the headphone amp, that seems to have enough current, is the LM380/386. But I think an NJM4556 with booster may sound better and have better THD.

Carlos
ricardo
Posts: 262
Joined: Sun Apr 10, 2011 7:24 am

Re: Rioc

Post by ricardo »

carlmart wrote:... your insistence in using every chance you can in pointing my using 12v single supply as my limitation. It is not.

Please acknowledge that is just your opinion, not a fact.
The FACT is that using 12V limits your drive capability into certain headphones. With NJM4556 & 12V, extra boosters will NOT drive headphones better.

Why so touchy about FACTS ? You might also like to STUDY the links & datasheets posted by those trying to help you .. and make an attempt to understand them.

The NWAV site will tell you if a single NJM4556 on 12V will drive the phones you think are important. If so, no problem. Links in my 31may post.

BTW, LM380 is good to generate your Vcc/2
carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Headphone stage & preamp output

Post by carlmart »

I'm already generating my Vcc/2 from the THAT 4315 itself, which has internal resistors you just buffer.

I've asked the question at NWAV's blog, about what will an NJM4556 move using 12v single supply. Do you think two 4556 in parallel would improve things?

Unfortunately there are no curves I could find for the 4556, that might tell what could be reached for each voltage.

Apparently I should be able to get +/-3v RMS, and NWAV mentions that a Beyer DT880-600 needs 43 mW at 600 ohms to hit 110 dB which is 5 V RMS. Isn't that within my capabilities?

Yes, I would like my preamp to move any casket, but I'm not sure it will. Perhaps that is something I will have to accept.
ricardo
Posts: 262
Joined: Sun Apr 10, 2011 7:24 am

Re: Headphone stage & preamp output

Post by ricardo »

carlmart wrote:I've asked the question at NWAV's blog, about what will an NJM4556 move using 12v single supply.
NWAV is AWOL. But he has loadsa good pages on his website which you should read if you are interested in headphone amps.
NWAV mentions that a Beyer DT880-600 needs 43 mW at 600 ohms to hit 110 dB which is 5 V RMS. Isn't that within my capabilities?
That's more than 7Vp & 14Vp-p. So NO. Your OPA won't reach the rails .. even those which claim to do so.
I would like my preamp to move any casket, but I'm not sure it will. Perhaps that is something I will have to accept.
YES.

But its an ENG preamp. You aren't recording Alfred Brendel at the Wigmore Hall.
carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Headphone stage & preamp output

Post by carlmart »

ricardo wrote:YES.

But its an ENG preamp. You aren't recording Alfred Brendel at the Wigmore Hall.
That is a fact I can live with quite well. :)
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Headphone stage & preamp output

Post by mediatechnology »

mediatechnology wrote:One thing this series has going for it that the 5532/34 and LM4562/LME49710/LME49720 don't is that the MC33078 does NOT have differential input clamp diodes.

:o Evil! Truly EVIL !!?!
Not sure I'm following you ricardo.

The MC33078-series actually does make a pretty good HS peak detector, clamp and poor-man's comparator.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Headphone stage & preamp output

Post by mediatechnology »

The FACT is that using 12V limits your drive capability into certain headphones.
+1. Let's add physics to the list too.
Low impedance phones with a 12V supply may not present a problem; my Sony MDR-7650 are 63 Ohms with a sensitivity of 106 dB/mW.
I get plenty of drive for myself with only a couple of volts.

In response to the question about which general-purpose op amp would be "better," from a low-voltage standpoint, the LME49710, LME49720 LM4562 run down to 5V; the 5532, for example, is about 10V minimum.
FWIW the 4556 operates to 4V.
For the HPA the 4556 is likely the better choice because it's got the drive current.

In the headphone amp carl you could parallel 4556 sections for more current if you use current-sharing (ballast) resistors.
If it were me, in your application, I'd go that route rather than adding current boost transistors and losing additional voltage swing from the external transistors.
ricardo
Posts: 262
Joined: Sun Apr 10, 2011 7:24 am

Re: Headphone stage & preamp output

Post by ricardo »

mediatechnology wrote:
mediatechnology wrote:One thing this series has going for it that the 5532/34 and LM4562/LME49710/LME49720 don't is that the MC33078 does NOT have differential input clamp diodes.

:o Evil! Truly EVIL !!?!
Not sure I'm following you ricardo.
Insclutable Oliental attempt decadent Western joke :ugeek:

OK, I LIKE OPAs with input clamping diodes. I tend to put them on TL07x cos they don't have them .. especially if they can be overloaded.
carlmart
Posts: 169
Joined: Mon May 19, 2014 6:11 am

Re: Headphone stage & preamp output

Post by carlmart »

mediatechnology wrote:
+1. Let's add physics to the list too.
Low impedance phones with a 12V supply may not present a problem; my Sony MDR-7650 are 63 Ohms with a sensitivity of 106 dB/mW.
I get plenty of drive for myself with only a couple of volts.
Probably a mistake I made myself many times: it's Sony MDR-7506. And I have a few of those. Those are the industry reference nowadays, and I do believe I can move them quite high with 12v.
In response to the question about which general-purpose op amp would be "better," from a low-voltage standpoint, the LME49710, LME49720 LM4562 run down to 5V; the 5532, for example, is about 10V minimum.
FWIW the 4556 operates to 4V.
For the HPA the 4556 is likely the better choice because it's got the drive current.
I got to believe a 4556 will be a better for all my opamps in this project, instead of the MC33178.
In the headphone amp carl you could parallel 4556 sections for more current if you use current-sharing (ballast) resistors.
If it were me, in your application, I'd go that route rather than adding current boost transistors and losing additional voltage swing from the external transistors.
Do you mean something like this for the paralleling?

My question is how do I connect my Vref for the lower opamp.


Carlos (not Carl)
Attachments
parallel_OPA.jpg
parallel_OPA.jpg (70.76 KiB) Viewed 15513 times
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

Re: Headphone stage & preamp output

Post by JR. »

I have used variants on that (paralleling two or more opamps) to get more drive current.

The use of degeneration resistors in series with each output prevents the multiple opamps from fighting each other over mV offset differences.

If you instead connect the feedback resistor Rf to the junction of the two 10 ohm resistors you will actually allow the opamp to control the final output.

I do not know if you need that zobel network.

For the headphone application connecting the feedback to before the 10 ohm resistors, looks like a 5 ohm R in series with the output, probably not a huge problem for 600 ohm cans. For commercial headphone amps I often used significant R to accommodate the wide range of headphone impedances (3.2 ohm - >600 ohm )

JR

PS: the one time I paralleled opamps was for a different application inside the product. For robust headphone amps I prefer discrete transistor buffers. Whenever something I design interfaces with the outside world, I always assume the worst from customers.
Cancel the "cancel culture", do not support mob hatred.
Post Reply