Suggest you get a 2M Blue stylus for your cartridge. The Red stylus is really marginal, even by entry level standards.
jeff
Also less surface noise with Blue , I will upgrade it with replacing the stylus to Blue 👍
'R11 does the same as R123 in Technics su V6 , isn't it?'
Not sure - I have not played with it enough to make a definitive comment on it. IIRC, Lipshitz commented that it wasn't entirely necessary - the post RC filter (R31 and C3) to pull the top end within +-0.2dB seemed to be more important. With R11 = 120 Ohms, here is the result (the light blue trace is with the 120 Ohms):-
Not sure - I have not played with it enough to make a definitive comment on it. IIRC, Lipshitz commented that it wasn't entirely necessary - the post RC filter (R31 and C3) to pull the top end within +-0.2dB seemed to be more important. With R11 = 120 Ohms, here is the result (the light blue trace is with the 120 Ohms):-
Thanks! In Technics I remember that resistor giving less phaseshift and removing some peaking in the 1...10Mhz region...although a bit distant memory to be very accurate.
Whilst that is true at 1KHz the signal off the disc is 20dB higher at 10KHz due to the RIAA equalisation. With all the gain in the one chip I suspect you are pushing it into distortion at high frequencies. Increase R5 as suggested and add a second stage on the output to make up the gain.
An ortofon red is not a cheapy cartridge. If used at 1.5 g or lower should not damage records unless stylus is broken. My fleet of LP's was not damaged by a Grado FTE MM cart on AR turntable 1970-78 followed by a Shure M97 Era IV cart on a BIC 940 changer til LP fleet stolen 2019. Of course too light cart weight or fuzz on the needle or surface can cause distortion. Wash your LP's, I do.
Circuit below makes a Colombia piano LP Peter Serkin 3 Beethoven Sonatas sound exactly the same as the CD, barring the usual ticks & pops of dirt which my ear tunes out since I grew up with LP's. That's on SP2-XT speakers with Peavey CS800s amp. More extreme High Freq test tracks include Peter Nero Young & Warm & wonderful (solo Steinway grand top octave) and Martin Denny Hawaii (lots of fizzy percussion). Disclaimer, the SP2-XT stop at 17 khz and my tinitus masks any sound from 14 khz up.
You'll note circuit in Herald RA-88a has no input cap, no cart loading cap, 10 uf output cap, and 1k load resistor to ground cap on feedback pin. Much simpler feedback network. The cartridge is loaded enough by a 6' rca cable, about 600 pf. I live in a high RF environment, extensive filtering was installed by me to keep AM & FM radio from coming in on the power supply line or from the output, and a 33 pf ceramic cap was parallel and installed by Herald (OEM) on the RCA input jack. There is a 33 pf ceramic cap from output to feedback op amp pins not shown, which prevents 1 mhz oscillation of the fast 33078. 4558 OEM op amp in the position was hissy and packaging & PS errors by Herald caused a lot of hum, that I stomped out. 33078 has equivalent drive to a 5532 with less PS current required. 33078 also was available when I bought op amps, and 5532 DIP was not.
Last edited:
Was this or a worn-out stylus already excluded as possible cause?
There are a lot of possible refinements of the phono preamplifier, but I don't see anything so wrong with it that it could explain gross distortion.
If you still have a recorder output on your amplifier, if it isn't buffered and if the recorder is connected but switched off, that could also cause distortion - but that's three ifs in a row. I had that issue with a Technics SU-V40 amplifier and a Philips CDR775 recorder.
Last edited:
I have a record with the first album of Tori Amos reissued in 2012 I think...At some moment on a particular track some harshness or distorted sound in the highs is heard.I thought first it was that the vinyl was damaged , but then I bought the CD.The same part was identical on the CD.Her voice wasn't properly compressed for a few seconds here and there on the whole album, but in one place that was very obvious and it was a recording mistake.
This RIAA EQ network is not accurate enough. The above plot is for 20 Hz to 20 kHz. Below plot is wide band out to 20 MHz. The high source impedance on the inverting input also introduced more noise than is necessary - try to keep it below 200 Ohms for better noise performance.
This will give lower noise. You can omit the DC blocking cap if you use a low offset JFET input opamp with Vos <500uV, or you trim it out - you can do this by splitting R8 into 47k in series with 1k and injecting the offset adjust current into the junction of the two resistors from a 470k resistor to a 25 turn 10k trimmer across the rails. Note if you use this appeoach, you must have C7 in place. Decouple the offset trim node with a 10uF X7R cap. You should AC couple the output of U2 to the next stage as shown with C5.
The peak current into the EQ network is 16mA at 1.2V input at 20 kHz - so again, adequate and well within the capability of modern opamps. You can expect an absolute peak output at 20k Hz out of the cartridge to be 46 dB above the 1kHz value (20 dB for the RIAA pre-emphasis + 26 dB headroom) so for a 5mV output cart, this is 1V. Hitting 1V input will be a rare occurrence, but it can happen.
The peak current into the EQ network is 16mA at 1.2V input at 20 kHz - so again, adequate and well within the capability of modern opamps. You can expect an absolute peak output at 20k Hz out of the cartridge to be 46 dB above the 1kHz value (20 dB for the RIAA pre-emphasis + 26 dB headroom) so for a 5mV output cart, this is 1V. Hitting 1V input will be a rare occurrence, but it can happen.
OK, and what kind of dielectric?
Most capacitors range from good to very good, but class 2 ceramic capacitors (X7R, X5R, Y5V, Z5U) are bad to very bad. Then again, their tolerance is usually 10 % or higher - they vary so much with just about everything that there is no point in making them to tighter tolerances.
Most capacitors range from good to very good, but class 2 ceramic capacitors (X7R, X5R, Y5V, Z5U) are bad to very bad. Then again, their tolerance is usually 10 % or higher - they vary so much with just about everything that there is no point in making them to tighter tolerances.
Last edited:
@totally analogue @dreamth
@MarcelvdG
I changed R5 to 220 to decrease the gain to 30dB , it got just a bit better, still distortion at high frequencies are there at some levels. I guess the problem would be the low impedance of RIAA network. I'm going to increase R5 to 1k and re-calculate the EQ network for the sake of higher impedance and see if it works better. I was just wondering why D Self published such design on Elektor in 2012 when in real world it doesn't work properly?
@MarcelvdG
I changed R5 to 220 to decrease the gain to 30dB , it got just a bit better, still distortion at high frequencies are there at some levels. I guess the problem would be the low impedance of RIAA network. I'm going to increase R5 to 1k and re-calculate the EQ network for the sake of higher impedance and see if it works better. I was just wondering why D Self published such design on Elektor in 2012 when in real world it doesn't work properly?