Sound quality question

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all,

For over one month, I was listening music from my PC (Creative XFI sound card installed) with Ultrasone DJ1 headphones. I am sure its not an "audiophile" configuration however this is the situation. In this one month I found some reference CDs of hifi manufacturers like McIntosh and Marantz.. Theyre really good recorded CDs and I listened very high quality records from that CDs that I've never heard before.

Anyway,

Last night I found a chance to listen that CDs with my last LM4702+Lateral MOSFET amplifier, Pioneer DV-585A universal player, Wharfedale Diamond 9.6 speakers. However I am dissapointed!

- I coulndt hear some details in music,
- The mid and treble details are weak,
- The sound comes a bit bassy
- Scene affect is not wide as headphones
- After a while, the music becomes boring and makes sleepy...

This points has been determined after just one hour listening experience... I had to change about 30 songs (without finishing) in one hour. But I could listen four full albums while listening with headphones!

What can you say about this situation? Where am I mistaken? Should I change my speakers or amplifier or CDP?
 
We get used to listening to a system, what ever that system consists of. When we change over to something else, we are subconsciously comparing the two for the first day or so.

It's hard for speaker based systems to compare to headphones as you have now found out. So you need to listen for a bit longer to the speakers, then start looking for what is the weak link in that system. It is almost impossible for any of us to look at your list of components and spot what it is that you don't like.

The only real way to find out is to substitute other components until you have the sound that you can live with. I know this isn't always easy to do but there isn't another 'magic' solution! ;)

One thing you can try is to use the PC as a source instead of the DVD player. :att'n:

If you think that the problem may lie with your chip amp, please post full details here so we can comment further.
 
I agree with Nuuk.

And maybe you need to change some component in your amp to achive the listening-experiance you are looking for.

Any schematic? Does it have high-pass/low-pass-filter that maybe cut of to early or to late? It can take some time to configure your amp to a sound-level that matches your needs.
 
In fact; this is not the first time I dissapionted.
Three or four months ago, I was using an LM4780 gainclone (with low quality parts) and happy with them, until;
I made a Class A headphone amp and tested with my pioneer DV585 and the same headphones. I was shocked! This was the first time I've heard some details on music.. And I said "It was not MUSIC what I've listened until now"..
Then I decided to make a better amplifier. And finished this;
An externally hosted image should be here but it was not working when we last tested it.

The first tests were excellent. It has sharp and detailed sound. No distortion or noise were exist. Or I felt this way. I found happiness again. However its case was ugly and huge. So I decided to make him wait until a good enclosure made. In this period (about two months) I was listening music from my XFI sound card and headphones. However at the same time I got some "audiophile" CD s from somewhere which I ve never listened from my speaker system. I loved this CDs too much. But I coudlnt find anytime to listen with my speaker system until last night.
So this is the story again.
By the way, I dont use any preamp or buffer before this amp. Just a 100K pot and this amp... Also there is no input cap..
As you can see, the input impedance is around 6K8 (or less) Do you think this situation "cuts" some details?
 
I have never worked with LM4702 but...

RIN1 and RIN2 looks to low. I would rise it to 1K - 2K.

RS1 and RS2 looks to low to. I would rise it to a value of 82K to achive a 80mS high pass roll-off. You could go a bit higher, maybe 90K.

CS1 and CS2 also looks to low. The low-pass filter should be between 0.5mS to 2mS. They should be 1nF-2nF. I would try with 1nF first, it gives you 1mS.

Those are my recommendations. Correct me anyone if I am wrong (AndrewT :) ).

It could take some testing with different values to find a sound that you like. To me it looks like you right now is cutting off all details in the music with your current configuration.

Edit: RSN1 and RSN2, the zobel network, you use 10R. Standard is 2R7 and that maybe affects the sound also?
 
Moderator
Joined 2002
Paid Member
I would first try and minimise the number of differences in the two chains so you can spot the offending component.

Try and connect the amplifier to the X-fi and report back on those impressions. I don't have much respect for standalone universal players, the X-fi may yet be a better source than one of those.

I know that my E-mu1212m is a much better source than my heavily modded NAD c521, itself a entry-level player but better than a DVD player.
 
Nordic
Your symptoms sounds like the typical GC without a buffer/preamp.

I was thinking on this way also. Do you think that is a result of low input impedance?

CJ900RR
I have never worked with LM4702 but...

RIN1 and RIN2 looks to low. I would rise it to 1K - 2K.

RS1 and RS2 looks to low to. I would rise it to a value of 82K to achive a 80mS high pass roll-off. You could go a bit higher, maybe 90K.

CS1 and CS2 also looks to low. The low-pass filter should be between 0.5mS to 2mS. They should be 1nF-2nF. I would try with 1nF first, it gives you 1mS.

Those are my recommendations. Correct me anyone if I am wrong (AndrewT ).

It could take some testing with different values to find a sound that you like. To me it looks like you right now is cutting off all details in the music with your current configuration.

Edit: RSN1 and RSN2, the zobel network, you use 10R. Standard is 2R7 and that maybe affects the sound also?

Do R1 and R2 used for input current limiting? If so, this circuit can work without them.

RS1 and RS2 is that value because of to make balance with NFB line. And as my experiences, low value NFB resistors reduce the output noise. So if I increase the RS values then negative and positive inputs become unbalanced. But I dont know the result of unbalanced inputz.

CS1 and CS2 is there to cut HF receiving. I've tested with much higher values like 470pF and nothing changed. However it looks me some weird; if I have problems with high frequencies (details on trebles) so why should I increase this caps value?

For the zobel network; youre right I must change the resistor with 2R7 or 4R7..

Thx.

Sangram
I would first try and minimise the number of differences in the two chains so you can spot the offending component.

Try and connect the amplifier to the X-fi and report back on those impressions. I don't have much respect for standalone universal players, the X-fi may yet be a better source than one of those.

I know that my E-mu1212m is a much better source than my heavily modded NAD c521, itself a entry-level player but better than a DVD player.

I will test this amp with my PC. However its output impedance is much lower than my DV585 (I guess it has 600R and XFI has lower than 10R) So this comparison will not be FAIR. And to move my PC to the living room will be another difficulty.. But I will do that.
 
Hi,
the volume pot output impedance varies from zero to 25K.
This becomes Rs for the power amplifier.
The amp should have an input impedance (Zin) at least 5times larger and maybe benefits from being as much as twenty times larger.
You should have Zin>=125k and upto 500k to match the volume pot you have chosen, NOT 6k8.

Change the 6k8 to 47k and throw away the 100k pot. Fit a 10k pot and try again.
If this improves things then the next stage is a proper buffered volume pot to bring Rs down to about 50r. Fit in a decent cap as a DC blocker.
If it does not improve, then borrow a decent discrete amplifier to determine if the problem is in the amp or the speakers or the room interaction with the speakers.
 
Dear Andrew,
In fact, I plan to make a low output impedance preamplifier stage for that amplifier. And 100K pot (the highest quality pot I have) will be putted before this stage. And a 6u8 or 10u DC blocking capacitor will be added between preamp and amp stages. Anyother components will not be changed.
Do you think this will solve my problem?
 
The overriding theme here is that you like headphones more than speakers. No question you'll hear more detail with headphones, for several reasons. Isolation is better. Separation is better. The amp may perform better into phones. Cabinet resonances aren't a problem. There's no nasty crossover to worry about. No clipping or power problems. Most importantly, and often overlooked, you're almost certainly listening louder. It's easy to listen too loud with headphones, to the detriment of your hearing, but if you look at the Fletcher-Munson curves, you'll see that your hearing of highs and lows is much better at high levels. At least for a while. When I copy LPs to CDs, I do all my quality evaluation and click/pop removal, using headphones. It's just easier to hear everything. OTOH, speakers are a different experience. You get the visceral feel of bass. You get a more realistic sound stage. You can listen with friends. Speakers can be great, but IMO it's tough to get the same definition.
 
Ex-Moderator
Joined 2005
Headphones really are the best for hearing all the musical information in recordings, but like Conrad said they do not offer the same experience as listening to speakers. You don't feel the music in the same way and there is no sound stage unless you have a good imagination and really get into the music. To me, the sound stage is a very key part of the illusion of real reproduction.
 
Wait...
I am confused... Do you say; listening with speakers is more realistic?
If so, why I feel better with headphones?
Yes, some songs sounds equal both on headphones and speakers, but especially with "McIntosh Audiophile Reference CD" and "Marantz 50th Annual CD" there are huge differences between speakers and headphones. And as mentioned headphone is far better than my speakers...
For a comparison; my headphones are not very very good ones. Just 180€, German made mid market things.. And my speakers are not reference series. Theyre about 700-750€ also mid market goods. So we can put them to the same quality range together. So I expect to have at least very close sound quality both of them. And my headphones done this task excellently, but the speakers dont!
Why?
 
I am confused... Do you say; listening with speakers is more realistic?

Ozgur, take one step at a time. Remember your system through loudspeakers needs to be 'put right' first before you draw any conclusions.

Nobody was saying listening through speakers is more realistic than headphones. Each have has its own qualities. ;)

If you are building a buffer for the GC, take it from somebody who has built and listened to quite a few, the discrete Jfet design of Pedja Rogic is clearly the best! :att'n:
 
Dear Nuuk,

The speaker positioning matter is one of my biggest problems. I am married :mad:, and WAF prevents me on many actions. So I afraid that cannot move the speakers. May be I (as myself) can move everywhere in the living room. But believe me, I've tested every "strategic" positions in the room already.
Concerning the buffer issue; I know Pedja Rogics JFEt buffer. I discussed that 5 months ago in this forum. So after that I made a SE headphone amp (as mentioned) and after a while I saw that I can use that headamp as a preamp also! Because its a Class A amp, has very very low noise, great swing capability, very low outpu impedance etc... So I plan to use my headphone amp as a preamp befor my LM4702...
What do you say? Pedja's buffer is made of a JFET and a current source and my buffer is made of a MOSFET and also a current source... The result will be identical or very close I guess..
 
Hi Ozgur,

I'm sorry, you misunderstood me when I talked about getting your system 'right' with the loudspeakers. I meant you should wait to get the whole system working well before making comparisons with the sound when using your headphones. In other words, sorting out your amplifier rather than moving your speakers around.

And yes, your headphone amp should make a good buffer. ;)
 
Nordic said:
Lol well Pedja's buffer includes jfets and a regular IRF mosfet.

http://www.diyaudio.com/forums/showthread.php?s=&threadid=95007&highlight=

I'm chargeing my camera's batteries to show you my semi-smd version later... Next version will use SMD regulators and jfets as well...

As you remember, we talked alot about Pedja's buffer in the past. However, I was planning that buffer with a LM38XX gain clone. But now its subject to use with a LM4702 amp. Anyway, this is not a problem in fact.
I examined the Pedja's circuit after my last posting again. JFET works as amplifier and the others for constant current source as I understand. So my circuit uses an OPA134 as buffer and VAS, IRF510 as output buffer and LM317 as constatnt current source. I think it will work good. I will publish my circuit on next posting.

Nuuk said:
Hi Ozgur,

I'm sorry, you misunderstood me when I talked about getting your system 'right' with the loudspeakers. I meant you should wait to get the whole system working well before making comparisons with the sound when using your headphones. In other words, sorting out your amplifier rather than moving your speakers around.

And yes, your headphone amp should make a good buffer. ;)

Ok this is because of my poor English.

:dead:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.