Don't you love it when things finally work!
I constantly fight with audio settings on my PC setup. I have to be very careful to set the device sample rate and bit depth exactly the same as those used in the measurement software. Otherwise the OS does sample rate conversion and things get messed up. I open the settings on the software and check the configuration multiple times. Some times the settings revert to some default after I have changed them. A reboot or unplugging the device and rebooting is sometimes required as a driver can get hung up on a setting.
I looked up the spec's for that Benz Ruby 2 and what I found said it had a 5 um tip radius. That radius should give you less than 1% tracing error distortion. I don't know how many hours you have on that stylus, but the higher distortion you are seeing may be due to some wear increasing the effective radius closer to 8 um.
I constantly fight with audio settings on my PC setup. I have to be very careful to set the device sample rate and bit depth exactly the same as those used in the measurement software. Otherwise the OS does sample rate conversion and things get messed up. I open the settings on the software and check the configuration multiple times. Some times the settings revert to some default after I have changed them. A reboot or unplugging the device and rebooting is sometimes required as a driver can get hung up on a setting.
I looked up the spec's for that Benz Ruby 2 and what I found said it had a 5 um tip radius. That radius should give you less than 1% tracing error distortion. I don't know how many hours you have on that stylus, but the higher distortion you are seeing may be due to some wear increasing the effective radius closer to 8 um.
Many many hours 🙂
By the way thanks for contribution, I just quickly went through your bigger post but did not reply yet, all free time was preoccupied with this problem. Hope these computers will behave from now on
Cheers,
Drazen
Last edited:
It seems your A/D converter is causing these problems.
There is still a large gap between both LP measurements.
Not so nice to deal with.
Hans
There is still a large gap between both LP measurements.
Not so nice to deal with.
Hans
So far conclusions can be:
1) your SSM preamp seems to work properly
2) The Orofon LP recording revealed that with a 1Khz tone the sine wave is distorted on the upper side, causing the THD to increase.
This can be caused by several reasons:
A) Cart adjustment is not o.k. , most likely the overhang.
B) Cart has issues, maybe a worn needle. Would be interesting to see how the right channel behaves.
So far seemingly only the left channel was used.
C) Ortofon LP has issues, but this seems the least likely cause.
3) The constant velocity does not give the expected 10dB/dec.
Same possible causes as under 2) A to C.
Hans
1) your SSM preamp seems to work properly
2) The Orofon LP recording revealed that with a 1Khz tone the sine wave is distorted on the upper side, causing the THD to increase.
This can be caused by several reasons:
A) Cart adjustment is not o.k. , most likely the overhang.
B) Cart has issues, maybe a worn needle. Would be interesting to see how the right channel behaves.
So far seemingly only the left channel was used.
C) Ortofon LP has issues, but this seems the least likely cause.
3) The constant velocity does not give the expected 10dB/dec.
Same possible causes as under 2) A to C.
Hans
Hi Hans,
I think you are most likely right, might be my cart is worn. I will need to take it to specialist to look at.
I adjusted the azimuth a bit as my eyes allow, cleaned cart and connected measurement gear in stereo.
Tested again, but result is just slightly better, still well over 2% TDH at 5cm/s 1khz.
However channels are behaving equal enough. Here is shot of square wave both channels driven and tracking test 90um peak (track 13) showing distortion at only one side of wave, as Olsond showed in #503. Both signals offset in scope for visibility..
I think you are most likely right, might be my cart is worn. I will need to take it to specialist to look at.
I adjusted the azimuth a bit as my eyes allow, cleaned cart and connected measurement gear in stereo.
Tested again, but result is just slightly better, still well over 2% TDH at 5cm/s 1khz.
However channels are behaving equal enough. Here is shot of square wave both channels driven and tracking test 90um peak (track 13) showing distortion at only one side of wave, as Olsond showed in #503. Both signals offset in scope for visibility..
Ooooo yes 🤣 ! But I'm not concerned at all, neither it is question of life and death, neither i live from it (even that would be nice) . It is hobby and I really enjoy the process.
It is also a bit romantic as it seems that my cartridge was loosing HF together with me, I'm now half deaf in HF so I don't notice my partner being half mute in the same region 🤣🤣🤣
She still plays beautifully to my ears.
I still need to make pro check, but it seems she needs retipping after 20y of service. Coils seem fine as it could not be that L and R are nicely overlapping on scope otherwise.
Can anyone recommend retipp service in EU?
On the other hand, if it is to trust eBay, people ask much over 1000€ even for well used Ruby 2, mine is on top of that naked and I'm first - only owner. Today I don't have such budget as 20y ago 😢, but even when I pay 30% for ebay, I can still get new Denon DL103R and go for few beers, and all is good again 🙂
This is stereo "square" without offset, L and R rather close together, im not sure if this is close enough, but seems good to me:
Dont know if anyone tried this for fun, good knock (like you would knock the door) on the plinth makes for a good dance with squares, they just miss eyes
PS @Hans, about earlier confusion related to Ortofon square; now I see on LP it says duty cycle 3:7, picture makes sense now.
The specification of 3:7 was confusing. It's a 70% duty cycle waveform, high 70% of the time. Why they didn't just say that I don't know. I had to look at the record with a microscope to figure out what the polarity was. A positive mono bass wave pulse moves the groove towards the outside of the disk.
Figure 1. Outside of record is to the right. Longer segment of the trace jogs to the right.
Figure 1. Outside of record is to the right. Longer segment of the trace jogs to the right.
Now I'm quite sure I can not contribute this discussion by good preamp (main topic of this subject) measurements before I repair my cartridge or get new one.
What I can still do while TT is in one piece is to touch again upon overload margin: as seen by phono preamp:
For this purpose I recorded from "part 2 of Laure Anderson US live album" LP, songs O Superman (as Hierfi likes) and Talkshow ( as I find it very high dynamic in HF region), luckily they are one after another on this LP.
The recording stream is as follows:
1. My tired Ruby cart playing not so perfect record...... It's 0db should be around 0.4mV @ 1khz as we concluded. That should be +20db@ 20khz according to riaa, or 4mV.
2. 2 x SSM2017 at total (linear) gain of 66db. Max 0db @20khz should than be 4mV x 66db (x2000V/V) = 8VRMS@20khz.
3. than it goes to passive riaa equilizer, LF is intact, mids (eg 1khz is reduced by x10 (20db) and HF is reduced by x100 (40db)
4. This is than amplified by second stage (linear) with OP2134 by 20 db (see post #339 for sketch of complete preamp)
4. This final signal is then sent to RME soundcard input . RME input has capacity to receive approx 5 V peak to peak .
5. Card input I set to +4db gain as there was no 0db setting.
6. Digitally it is ADC-ed @ 192/24 and that is recorded by Audacy sw.
7 Finally saved as WAV (Microsoft) 32-bit float PCM in wav. file
My main point is # 2 above; peak HF from LP after 66db amp is about 8V RMS
In fully differential mode amp with +-18 Volts DC can deliver 20+VRMS easily. Since these are short term pulses, RIAA R-C network will not see issues neither.
Above is for HF high end, and I believe nobody records 0db in 20kHz , so no overload margin is needed, even there is some left.
This is the picture of whole recording in Audacy, (left part is O Superman, right is Talkshow)
And this is highest peak I could see in above recording, cant see clipping here....:
The whole record is 1.6 GB in 192/24 resolution, whoever wants it (it is also good to listen) I will send via wetransfer, just PM me with address where to send. Looking for comments.
Cheers,
Drazen
What I can still do while TT is in one piece is to touch again upon overload margin: as seen by phono preamp:
For this purpose I recorded from "part 2 of Laure Anderson US live album" LP, songs O Superman (as Hierfi likes) and Talkshow ( as I find it very high dynamic in HF region), luckily they are one after another on this LP.
The recording stream is as follows:
1. My tired Ruby cart playing not so perfect record...... It's 0db should be around 0.4mV @ 1khz as we concluded. That should be +20db@ 20khz according to riaa, or 4mV.
2. 2 x SSM2017 at total (linear) gain of 66db. Max 0db @20khz should than be 4mV x 66db (x2000V/V) = 8VRMS@20khz.
3. than it goes to passive riaa equilizer, LF is intact, mids (eg 1khz is reduced by x10 (20db) and HF is reduced by x100 (40db)
4. This is than amplified by second stage (linear) with OP2134 by 20 db (see post #339 for sketch of complete preamp)
4. This final signal is then sent to RME soundcard input . RME input has capacity to receive approx 5 V peak to peak .
5. Card input I set to +4db gain as there was no 0db setting.
6. Digitally it is ADC-ed @ 192/24 and that is recorded by Audacy sw.
7 Finally saved as WAV (Microsoft) 32-bit float PCM in wav. file
My main point is # 2 above; peak HF from LP after 66db amp is about 8V RMS
In fully differential mode amp with +-18 Volts DC can deliver 20+VRMS easily. Since these are short term pulses, RIAA R-C network will not see issues neither.
Above is for HF high end, and I believe nobody records 0db in 20kHz , so no overload margin is needed, even there is some left.
This is the picture of whole recording in Audacy, (left part is O Superman, right is Talkshow)
And this is highest peak I could see in above recording, cant see clipping here....:
The whole record is 1.6 GB in 192/24 resolution, whoever wants it (it is also good to listen) I will send via wetransfer, just PM me with address where to send. Looking for comments.
Cheers,
Drazen
Last edited:
It’s an interesting theme to know how much overload can happen.
Therefore It would be nice to also have a 0dB@1Khz recording with exactly the same setting.
That will give you the reference in the Audacity plot.
Now your recordings comes to ca. 0.5 or 0.35rms in Audacity, but were lands the 0dB on this same scale from +/-1.
Hans
Hi Hans, good point.
Here is picture in scale of one of highest peaks of music recording and 1khz @5cm/s . I think ratio is as expected,
Highest audio peaks can be seen at 0.35 and peaks of 0dB@1Khz at 0.075.
That is roughly 13 dB above 0dB.
So it’s not that wrong to keep 14dB or a factor 5 as an overload margin, which means in your case 4Volt as maximimum output after Riaa, independant of frequency.
(((0.4mV*66dB)/20dB)*20dB)*5 = 4Vrms
Hans
That is roughly 13 dB above 0dB.
So it’s not that wrong to keep 14dB or a factor 5 as an overload margin, which means in your case 4Volt as maximimum output after Riaa, independant of frequency.
(((0.4mV*66dB)/20dB)*20dB)*5 = 4Vrms
Hans
Dear Hans,
Exactly my point, and since it is fully differential amp, it is actually just 2VRMS per + or - leg.
Earler i repediately measured about 735mV at 1khz@5cm/s amplified by 66db before RIAA , x 5 times =3.67 V, for this purpose that is same as 4 V you calculated.
The picture changes at 20 khz and then is 0.4 mV +20db(RIAA) +66db(amp) = 86db x 0.4 mV =8 Vrms (4V per differential leg) , still manageable easily but overload margin is might be 3 and not 5 with +-18v PS, this is good for me... No high frequency is recorded so high, that would hurt ears 🙄
Exactly my point, and since it is fully differential amp, it is actually just 2VRMS per + or - leg.
Earler i repediately measured about 735mV at 1khz@5cm/s amplified by 66db before RIAA , x 5 times =3.67 V, for this purpose that is same as 4 V you calculated.
The picture changes at 20 khz and then is 0.4 mV +20db(RIAA) +66db(amp) = 86db x 0.4 mV =8 Vrms (4V per differential leg) , still manageable easily but overload margin is might be 3 and not 5 with +-18v PS, this is good for me... No high frequency is recorded so high, that would hurt ears 🙄
What is interesting and not related to subject is that I got mail with reply from Hans this morning at 10:47, but this is not visible here on the thread
🤔
However, back to subject, here is that reply:
"At 20Khz the caculation remains the same as for 1Kz.
0dB at 20Khz would be 20dB higher or 4mV but extra attenuation in the Riaa network of 20dB filly compensates this.
So, the 4vrms overload or 2Vrms per leg after Riaa is frequency independent.
Hans"
My comment:
This is correct Hans, but we talk about different things. I am talking about overload before RIAA as there is where it can happen. Passive RIAA attenuates HF with 40db, no chance for signal after that to go into clipping, but before riaa it is possible.
🤔
However, back to subject, here is that reply:
"At 20Khz the caculation remains the same as for 1Kz.
0dB at 20Khz would be 20dB higher or 4mV but extra attenuation in the Riaa network of 20dB filly compensates this.
So, the 4vrms overload or 2Vrms per leg after Riaa is frequency independent.
Hans"
My comment:
This is correct Hans, but we talk about different things. I am talking about overload before RIAA as there is where it can happen. Passive RIAA attenuates HF with 40db, no chance for signal after that to go into clipping, but before riaa it is possible.
One of the reasons I switched from using the SSM2019 was that it produced a voltage out. Post passive RIAA typically requires significantly more headroom (as evidenced in this thread) because of the boosting of high frequencies. This headroom isn't nearly as problematic when using current mode output driving a passive RIAA as the resistive element is grounded as per below. It should be also noted that passive capacitive termination into ground in both cases can filter out broadband signals to the extent dependant upon the quality of the capacitors used.
I recently completed a fully balanced phono stage using a variant current mode RIAA but haven't done much testing for overload margins yet. I took a day off to attend the Toronto AudioFest on Saturday... leaving much to drool about behind. Several mono-blocks, taking up the size of ottoman's, sounded very good. It also seemed that turntables were more prevalent than last year.
I recently completed a fully balanced phono stage using a variant current mode RIAA but haven't done much testing for overload margins yet. I took a day off to attend the Toronto AudioFest on Saturday... leaving much to drool about behind. Several mono-blocks, taking up the size of ottoman's, sounded very good. It also seemed that turntables were more prevalent than last year.