RIAA circuit help needed

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
www.hifisonix.com
Joined 2003
Paid Member
You just have to trawl through the specs Im afraid. The OPA1641 looks very good and for the money, the NE5534A is hard to beat.

FET input opamps usually have low input noise currents but the input noise voltage is higher as a consequence of the channel resistance - the lower the better. Bipolars have lower noise voltage but the noise current is higher.

Hopefully I’ll have time in the next week or so to finish the spread sheet and then you can just enter the data and make a direct comparison.
 
Last edited:
Vinyl has much more performance limitations than noise performance of opamps. For instance, how many turntable owners have tangential tonearms? One percent? Or less. That means that geometry will present big problem for 99% of owners. Depending on place of cartridge on the record surface distortions will be very high. Even if you have tangential tonearm who can guarantee that cantilever is ideally set in cartridge body or that needle is positioned ideally when glued in the factory? I've seen many times that people buy more expensive cartridge and get worse sound. That's because it is absolutely necessary to have ideally set geometry if you want to get the performance advantages of better polished diamond. How many users are able to set geometry ideally? Probably very few.

The charm of vinyl is not the charm of great specs. If great vinyl sound is the goal I am sure that my PCB with NJM2114 will suffice. For the perfectionist, nothing better than Bonsai's PCB with two NE5534 will be needed.
 
First sentence from John Linsley Hood's "Valve and Transistor Amplifiers":

"An old friend of mine was in the habit of remarking that an engineer is someone who can do for five pounds what any damn fool can do for fifty."

That's what always guided me in the field of Audio. There must be some limit above which audio engineering becomes senseless and expensive mental acrobatics.
 
SOIC8 footprint pads in Eagle are too long for that. It would be necessary to make new footprints with shorter pads and making new footprint is not an easy task for DIYer. Prasi could do it but question remains how difficult it would be to hand solder such footprint with shorter pads.
It is up to you if you want to do this, as it is your design after all, it was only a suggestion.
Did you rotate the SOIC-8, 90 degrees, that way it does not matter too much if the pads are longer.
 
Actually its well worth going through tutorials on Eagle and learning to do footprints and devices, the pay-off is worth it.


I designed some SOIC8 breakouts a while back without the 90 degree turn - in practice you have to bend the legs a little to fit comfortably! Rotating means less than ideal routing of signals, not a big issue for audio stuff. You can place a decoupling cap on the underside of the breakout, note, in fact you can do footprints for both single and dual opamp decoupling.
 

PRR

Member
Joined 2003
Paid Member
Is there any such opamp that has both very low input voltage and current noise? Usually, if one is very low the other is relatively high.

The free variable is device area.

A large area device can be designed for low hiss voltage. If it can also have high current gain, it can have low hiss current.

BJTs can be bought any size you want. Usually current gain is not critical and it is hard to maintain high gain all across a large device, so we have to be picky.

FETs have to be bigger for the same hiss voltage but input current hiss is very low (in our impedance range). There are a few fat FETs which can be great. Sadly they have gone out of style. The whole idea of JFETs is old, meaning the few new ones we can buy are 5-generation tail-end foundries who don't really care.

*Op*Amps* have to sell in the millions. Chip cost is really die area, how many chips you can get from a wafer. So chip (opamp) design uses SMALL devices. Some less-small than others.

The old-old LM394 was great. You paid a lot and got a modestly large area die, subdivided as 100 individual BJTs so each one could have very high current gain, and the sum had low hiss voltage. That's out of production though there is an equivalent now available from East Europe.

My own feeling is that the '5532 (0r 5534, or -A) will get you there without a any muss or fuss. The few-dB difference on paper vanishes when you drop the needle on the best vinyl. Yes, I know I can hear amp-hiss below surface scratch, but if the music is good it does not matter.
 
SOIC8 footprint pads in Eagle are too long for that. It would be necessary to make new footprints with shorter pads and making new footprint is not an easy task for DIYer. Prasi could do it but question remains how difficult it would be to hand solder such footprint with shorter pads.

Sorry Ivan,
just had a look at this post. i dont normally go though diya index.
if you or anyone need anything, I am just a PM away.

i will do it tomorrow and share you the library file. its 12.35AM here, good night..

regards
prasi
 
Hi Prasi,

At that time I did not know that there is a suitable footprint in Eagle libraries. It's in the opamps.lbr and it's DIP8_SOIC8 combined footprint. It's DIP8 on the top layer and it's SOIC8 on the bottom layer. Very convenient. I shall use it from now on as there is much more high quality opamps in SOIC footprint, although I think that high performance opamps are not needed for good sound. Everything one needs for the good sound is TL072 and NE5532.
 
Hello Ivan,
Even I didnt know that.
Here an idea with provision for pin 4- pin 8 bypass cap (1210) on top side and Soic-8 on the bottom.
regards
prasi
 

Attachments

  • lay.png
    lay.png
    8.2 KB · Views: 76
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.