• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Improving A Vintage Phono Stage For Less Than $2.00!

Leestereo

Leestereo

Super Member
I have a Toshiba/Aurex SY-335 preamp that is up for a restore/upgrade. Its a mid-range unit that uses, like many vintage phono stages from the late 70s and 80s, the 4558 op amp. As is, the phono stage sounds average at best. I wanted to see if I could improve the phono performance by modifying the circuit components. A look at the schematic and a few calculations revealed that the the low end starts rolling off at 34Hz (F3) due to R03 and C59 and the RIAA equalization is also off at the 318µs and 3180µs time constants (R05, C03, C05 and R07, C07) when compared to the theoretical calculations. See http://hifisonix.com/wordpress/wp-content/uploads/2010/10/RIAA-Equalization-Amplifiers-Part-One1.pdf for a brief but good overview of the work of S. Lipshitz, P. Bandaxall and D. Self in this area.

The low end cut-off of the circuit is extended to below 10Hz by changing R03 from 1 kohm to 392 ohms and C59 from 4.7µF to 47µF. The 318us and 3180µs errors were corrected by changing the values of R05 from 560 kohms to 619 kohms and R07 from 47 kohms to 51.1 kohms. Both of the new resistors are standard 1% metal film values and the resultant time constants are now within 1% of the target values. The low pass filter cap, C11, was also removed to improve the high frequency phase response. The stock 4558 op-amp is also limiting phono stage performance due to its slew rate of 1V/µs, lowish Gain Bandwidth Product, and its relatively poor noise specs. The 4558 was replaced with a 4580DD op amp which is much better in all of the aforementioned parameters (5V/µs, 15MHz, 0.8µVrms). Although the 4580 has been surpassed by more recent designs, it is well suited for use in this particular circuit without any need for additional decoupling. So for the cost of 6 resistors, 2 capacitors and 1 op amp (a total cost of about $1.50), you can have a much improved phono stage!
 

Attachments

  • sy330 phono1.jpg
    sy330 phono1.jpg
    67.4 KB · Views: 293
  • sy330 phono1mod.jpg
    sy330 phono1mod.jpg
    64.2 KB · Views: 288
Last edited:
Register to hide this ad
Here is are charts of the "Stock Phono Stage RIAA" and the "Modified Phono Stage RIAA" illustrating the improvement in the accuracy of the RIAA equalized response. The modified circuit has a very smooth response of +/- 0.2 dB. Note that the even the stock circuit response can be described as +/- 0.5dB, 20Hz-20kHz but that hardly describes the response anomalies.
 

Attachments

  • phono stage charts.jpg
    phono stage charts.jpg
    58.4 KB · Views: 226
Last edited:
Can you describe the audible difference between 'before' and 'after'?
Not trolling, just trying to get an idea between the measured and the experienced in this case.
 
mhardy6647 said:
Can you describe the audible difference between 'before' and 'after'?
Not trolling, just trying to get an idea between the measured and the experienced in this case.
Click to expand...

The stock circuit sounded "boomy" to me with a somewhat harsh treble (compared to the phono stage in a NAD 106 preamp or a CJ PV-8 phono). After I made the RIAA comparison chart, it became clear that I was hearing the 160hz "hump" (its only 0.4dB, but it has a very low "Q") and the rising treble. The modified circuit is very nice sounding, "neutral" and "clean" would be apt descriptors; its very close to the NAD now. The change in sound response was subtle, and it took many recordings for me to convince me that I wasn't deluding myself (who knows, I might still be) and it was more easily heard when playing acoustic jazz and classical chamber music recordings. Note that I also replaced the other 5% resistors with 1% metal film resistors, increased the IC supply capacitors to 100µF (low ESR) and replaced the input and output coupling capacitors with low leakage types. The most obvious change in sound occurred after correcting the RIAA and changing the op amp. The other sound changes were more subtle.
 
Last edited:
Very fine job! IMO a decent phono section is critical because the signal quality will never be better than what happens in the first stage. The eq caps should be polystyrene, polypropylene, C0G/NP0 or silver-mica. Most already are but every now and then you find something where they didn't make a good choice.
 
Last edited:
ConradH said:
...IMO a decent phono section is critical because the signal quality will never be better than what happens in the first stage. The eq caps should be polystyrene, polypropylene, C0G/NP0 or silver-mica.
Click to expand...

Completely agree, if you don't get it off the record the first time right, you'll never be able to correct that later. The equalization caps are film: 2 polyester and 1 polystyrene in each channel. Would have preferred metallized polypropylene instead of the polyester (at least they didn't use ceramics); but then again this is a relatively cheap build, not a TOTL unit.

One thing I forgot to mention about the sound, I think that the stereo imaging is better now, probably due to better matching of the RIAA equalization between the left and right channels due to the use of 1% resistors instead of 5% ones.
 

Attachments

  • sy-335a.jpg
    sy-335a.jpg
    52 KB · Views: 130
perryinva said:
Interesting. I normally replace 4558 with 5332 or OPA2132.
Click to expand...


I'm guessing you meant 5532 rather than 5332. I actually considered using the 5532, but from my research, the 5532 would be noisier than the 4580. Not that familiar with the OPA2132, but its much higher slew rate suggests that I might have issues with oscillations if I just used it as a drop-in replacement. Also interestingly, the GWP of the both these ICs is poorer than that of the 4580 (15MHz) and from my research, this may be a more important parameter for an RIAA stage than absolute slew rate as long as the latter is at least 3V/us.
 
FWIW, I'm wary of polyester (Mylar). It has a distortion component that's much larger than other films and my guess is that's why audio people tend not to like them. Even if it's not a super high end build it would be interesting to change out those polyesters and report back. I'm in the process of doing the exact same thing with the coupling caps in an ARC preamp.
 
yes 5532, I fat fingered. I'm not an opamp expert by any stretch. What is GWP? I was not familiar with the 4580, but it looks interesting, based on the spec sheet on Mouser, and is audio specific. Where did you find it in pin format? Input noise for 5532 is ultra low at .005uV vs .8uV on 4580. Are you referencing a different noise? I like the very low THD of the 4580, tho. THD is not listed for 5532, but it is a common cheap upgrade.
 
The 4580 is available from Mouser: http://www.mouser.com/ProductDetail/NJR/NJM4580DD/?qs=sGAEpiMZZMtCHixnSjNA6PRGqGBqB8eMLE5yAFNC/Lc=

I'm not op-amp expert either, but I do my best to leverage my formal scientific training when I'm looking at data. The GWP is the Gain Bandwidth Product, sometimes also abbreviated "GB". I have found op amp spec sheets to be difficult to compare since there isn't really any standard format; the data sheets only seem to report parameters which make the device look good!

Regarding the noise specs, the input noise for the 5532 is 0.005µV/√Hz which is not directly comparable to the 4580 0.8µVrms figure. The 5532 noise figure can be converted to a more useful format by specifying the bandwidth and the gain (see http://www.designnews.com/document.asp?doc_id=229503&dfpPParams=ind_182,aid_229503&dfpLayout=article ).
So for a bandwidth of 20Hz to 20,000Hz, the 0.005µV/√Hz number gets multiplied by square root of the bandwidth (i.e., 141.4) and then multiplied by a reasonable (for RIAA?) gain figure of 10 to give 7.07µV (0.005 x 141.4 x 10). This number can now be used to comparable to the 0.8µV number of the 4580, but only in the sense of order of magnitude, not absolutes since the measuring conditions are not likely to be the same between the 2 op-amps.
 
ConradH said:
FWIW, I'm wary of polyester (Mylar). It has a distortion component that's much larger than other films and my guess is that's why audio people tend not to like them. Even if it's not a super high end build it would be interesting to change out those polyesters and report back. I'm in the process of doing the exact same thing with the coupling caps in an ARC preamp.
Click to expand...

I've some polypropylene and polystyrene caps coming in my next Mouser order. How is the ARC preamp rehab going?
 
I just finished the "head amp" MC cartridge section. It had six electrolytics with RTV end seals. I'm not sure if they were made that way or if ARC added the stuff later. I could see under the end that one cap had started to leak. Electrically they were nearing end of life based on rising DF and a wider spread of value than they should have. Heck, they're 37 years old so I'm not complaining. I neatly wired in FC radials because I didn't find an axial I liked. Getting the head amp in and out is sort of a PITA mechanically.

Next I'll freshen up the power supply. Before I start changing the 5 uF Mylar couplers in the output section I want to do some more listening and make some measurements across them. I'm not sure what to install there, as good films will be quite large and expensive.

The SP-4 is sort of a sonic dog t**d, so I'll only sink money in it if an improvement is possible- the potted modules limit what one can do. My simple DIY RIAA using LME opamps and polypropylene caps puts it to shame at the moment.
 
Can't go wrong with Panasonic metallized polyprop caps, inexpensive and not "boutique" but decent.
BTW, IIRC the SP-4 has waaaay too much crap in the line stage sonic path as did all contemporary preamps due to customer expectation of features like tone & balance controls and interstage function switching. All that is required is one simple triode buffer stage with relatively mild gain, say 10-12dB or so. C-J does this with their pre's and are famously transparent tube pres, just keep in mind that if it's done with a single triode plate follower or dual triode stage with cath follower line driver out it will invert polarity of everything passing thru it and will require swapping polarity of speaker leads to correct.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom