I measured the signal levels off some test discs.
My cartridge is Denon DL-103 specified 0.3 mV @1kHz @5cm/s
I use a SUT Lundahl LL1678 with 1:16 turns ratio. The reflected load is 47k / (16x16) = 180 ohms.
My preamp has about 50 dB gain @1kHz (a bit too high).
My first test disc (MHV - Hungarian Disc Co.) has a 80mm/s 1kHz reference track, where I measured 4.20 V output. This scales down to 2.63 V @5cm/s. Divided by the phono gain 315 gives 8.35mV. Divided by the SUT gain 16 gives about 0.5mV. This is much higher than the specified sensitivity of the Denon, which I don't understand yet.
Another disc I tested is the HFN&RR 300Hz "torture" tracks:
+12 dB - 6.19Vrms
+14 dB - 7.62Vrms
+16 dB - 9.40Vrms
+18 dB - 12.40Vrms
A good phono preamplifier must have at least that high overload margin. Multiply the above RMS by 1.4 to get the peak 17.5V, 35V peak-to-peak. With a more common phono gain 40dB this is relaxed by a factor of about 3, that is 12Vpp. Did you expect it?
My cartridge is Denon DL-103 specified 0.3 mV @1kHz @5cm/s
I use a SUT Lundahl LL1678 with 1:16 turns ratio. The reflected load is 47k / (16x16) = 180 ohms.
My preamp has about 50 dB gain @1kHz (a bit too high).
My first test disc (MHV - Hungarian Disc Co.) has a 80mm/s 1kHz reference track, where I measured 4.20 V output. This scales down to 2.63 V @5cm/s. Divided by the phono gain 315 gives 8.35mV. Divided by the SUT gain 16 gives about 0.5mV. This is much higher than the specified sensitivity of the Denon, which I don't understand yet.
Another disc I tested is the HFN&RR 300Hz "torture" tracks:
+12 dB - 6.19Vrms
+14 dB - 7.62Vrms
+16 dB - 9.40Vrms
+18 dB - 12.40Vrms
A good phono preamplifier must have at least that high overload margin. Multiply the above RMS by 1.4 to get the peak 17.5V, 35V peak-to-peak. With a more common phono gain 40dB this is relaxed by a factor of about 3, that is 12Vpp. Did you expect it?
The DL-103 output is rated at 40Ω and the output voltage is specified into a 100Ω load. You have loaded the cartridge with 180Ω, which will add ~1.2dB to the output of the cartridge compared to 100Ω load. 80mm/s is +4.1dB above 50mm/s. Under your test conditions, you would expect to see 0.3mV + 5.3dB = 0.55mV ± the manufacturing tolerance of the DL-103, and the actual gain of your preamp.
There was a lot of good stuff written about overload margins in RIAA preamps in Wireless World in the 1970s, some of it by Douglass Self IIRC. He's a member here and has published many books on hifi circuit design.
Electronics for Vinyl
There was a lot of good stuff written about overload margins in RIAA preamps in Wireless World in the 1970s, some of it by Douglass Self IIRC. He's a member here and has published many books on hifi circuit design.
Electronics for Vinyl
Last edited:
John, your calculation looks correct, but something is still not clear.
I measured the cartridge output directly, using a Tektronix AA501 Distortion Analyzer in Level setting, with 400Hz filter.
First measurement was without load, the output was 0.95 mV @1kHz @8cm/s.
Second measurement was with 100R load, the output was 0.91 mV (L ch). From these values I calculated the internal resistance, which is 4.4R in contrast to the specified 40R. How is it possible? I double checked the signal levels, load resistance, everything is correct and repeatable. R ch is similar.
The nominal output with 100R load is 0.91mV x (5cm/s / 8cm/s) = 0.57mV, which is within tolerance.
Also calculated the transformation ratio of the Lundahl with 47k load on the secondary. It gave about 1:12.5 rather than 1:16.
Important is that it sounds good
I measured the cartridge output directly, using a Tektronix AA501 Distortion Analyzer in Level setting, with 400Hz filter.
First measurement was without load, the output was 0.95 mV @1kHz @8cm/s.
Second measurement was with 100R load, the output was 0.91 mV (L ch). From these values I calculated the internal resistance, which is 4.4R in contrast to the specified 40R. How is it possible? I double checked the signal levels, load resistance, everything is correct and repeatable. R ch is similar.
The nominal output with 100R load is 0.91mV x (5cm/s / 8cm/s) = 0.57mV, which is within tolerance.
Also calculated the transformation ratio of the Lundahl with 47k load on the secondary. It gave about 1:12.5 rather than 1:16.
Important is that it sounds good
A moving coil cartridge will have an output impedance that varies significantly with frequency. Consequently the loading used for a cartridge affects frequency and phase response, and causes the ideal load to be a hot topic on consumer audio forums. Of course the 'ideal load' is system dependent (i.e. turntable, tonearm, cartridge alignment, cable, preamplifier, etc), so one person's nirvana is quite likely someone else's tragedy. No matter, we still like to tell everyone what they need to do! But I digress...
There have been many different incarnations of the DL-103, some with different output levels to the common 0.3mV. If you are trying to replicate Denon's specification be sure you are using the correct spec for your DUT and are measuring it under the same conditions Denon used, preferably with the same source (test record) and test setup.
I don't pretend to know any more than you and my post was to show how assumptions, that actually might just be approximations, and small variances soon add up. The analogue world is fraught...
Although the above is the equivalent circuit for MM, MC will be very similar, just different values due to difference balance of masses. Ze at the top is the output impedance. Just to make it more complicated, some of the elements will be frequency and time dependent too!
Footnote: If you have a Dayton DATs system, I think it would be reasonable safe to measure the Zo of a moving coil with it. Might be worth trying it on a dud cart first tho...
There have been many different incarnations of the DL-103, some with different output levels to the common 0.3mV. If you are trying to replicate Denon's specification be sure you are using the correct spec for your DUT and are measuring it under the same conditions Denon used, preferably with the same source (test record) and test setup.
I don't pretend to know any more than you and my post was to show how assumptions, that actually might just be approximations, and small variances soon add up. The analogue world is fraught...
Although the above is the equivalent circuit for MM, MC will be very similar, just different values due to difference balance of masses. Ze at the top is the output impedance. Just to make it more complicated, some of the elements will be frequency and time dependent too!
Footnote: If you have a Dayton DATs system, I think it would be reasonable safe to measure the Zo of a moving coil with it. Might be worth trying it on a dud cart first tho...
Last edited:
Today I re-checked my measuring setup and found that the cartridge load was incorrect. Here are the correct values with real 100R load (1kHz, 8cm/s):
L ch output: 0.650mV
R ch output: 0.660mV
Which makes about 0.4mV @5cm/s.
(Infinite load is 0.950mV and 0.960mV)
The calculated internal resistances are about 46R and 45R.
Denon specified the recommended load >100R, so I believe 180R should be OK.
Last edited:
I repeated the measurements with a signal generator source. I used a "torture" input level of 5mV at the input of the MC step-up transformer, and I considered it +20dB. The output at the SUT was around 72mV (T = 1:14.4), and the phono output was around 28V (Au = 52dB) with 0.77% harmonic distortion. With 0dB input (0.5mV) I measured 2.68V output @ 0.1% THD. This output level is very close to what I measured off-record with a test signal.
I wonder what are the highest practical velocities of vinyls, e.g. the cannon shots at Chaikovsky's 1812 Overture, Beethoven's 5th, ...
I once owned that infamous Telarc ½ speed mastered 1812 Overture recording through which almost no cartridge could track the cannon blasts properly. Whilst it was an ego trip to own a turntable that could play the recording without crapping itself, it wasn't a particularly satisfying performance to listen to anyway. These days I use my turntable for music, not acoustic gymnastics.
IMO, that is just plain POOR mastering engineering. Who wants a record that very few cartridges can track? NOBODY, unless, as you say, someone who is more interested in specs than music.
Thanks for the information. Then it seems the highest modulation velocity is around 30...35 cm/s @1kHz, which is 17dB above 5 cm/s. Then +20 dB is a safe margin. In my case it means 5 mV might come from the cartridge at extreme conditions, and 28Vrms might appear at the phono output. That is 80V peak-to-peak signal level!
(As I mentioned my phono gain is on the high side with 50dB; if we calculate with a more common 40dB, the output swing would be 26Vpp. A phono preamplifier made of opamps with +/-15V PSU can just handle it)
(As I mentioned my phono gain is on the high side with 50dB; if we calculate with a more common 40dB, the output swing would be 26Vpp. A phono preamplifier made of opamps with +/-15V PSU can just handle it)
You know it’s strange I have yet to overload my phono preamp no matter what ratio used on my SUT’s. I have even experimented with EMT XSD-15
(1.1mV) into a 1:30 SUT and have yet to get distortion that could be heard.
I do not have a distortion analyzer, sounded decent to me....
(1.1mV) into a 1:30 SUT and have yet to get distortion that could be heard.
I do not have a distortion analyzer, sounded decent to me....
I have never measured it but, in this review it states it’s output as 1.05mV. Could be wrong though.
Listening #101 Page 3 | Stereophile.com
Listening #101 Page 3 | Stereophile.com
I had a phono preamp design published in audioXpress earlier this year. If you look at fig. 3 in the link below to part 1 of the article, there is a plot of peak output velocities measured by Shure in the 1970's - one of them is at 70cm/sec.
https://audioxpress.com/files/attachment/2735
While this is unusual, it probably makes sense to consider why overload was a big issue, and why it still might be the case nowadays. Across a range of vinyls, you can expect a maximum recorded level to be up to 6 dB at 1 kHz (some references say up to 9 dB). Add to this cartridge outputs that range from 2.5 mV at 1kHz to 5 mV - with some at 6mV so you add another 6 dB and now have a 12 dB input level range to deal with. In most phono amps there is no gain control other than the volume control which comes after the phono stage and therefore allows the listener to adjust for recording level differences between vinyls. Given the dynamic range of recordings of up to 14 dB, you can quickly see how 20 dB overload margin is not a bad design target to cater for most eventualities.
(you can download both parts of the article here You Can DIY! X-Altra MC-MM RIAA EQ Preamp - Part 1 | audioXpress)
https://audioxpress.com/files/attachment/2735
While this is unusual, it probably makes sense to consider why overload was a big issue, and why it still might be the case nowadays. Across a range of vinyls, you can expect a maximum recorded level to be up to 6 dB at 1 kHz (some references say up to 9 dB). Add to this cartridge outputs that range from 2.5 mV at 1kHz to 5 mV - with some at 6mV so you add another 6 dB and now have a 12 dB input level range to deal with. In most phono amps there is no gain control other than the volume control which comes after the phono stage and therefore allows the listener to adjust for recording level differences between vinyls. Given the dynamic range of recordings of up to 14 dB, you can quickly see how 20 dB overload margin is not a bad design target to cater for most eventualities.
(you can download both parts of the article here You Can DIY! X-Altra MC-MM RIAA EQ Preamp - Part 1 | audioXpress)