would it not be ideal to make the circuit first. then we know the PSU requirements...What about the circuit shown i post 288 could that be used as a starting point...??
http://www.diyaudio.com/forums/atta...218589d1302834378-parasound-jc3-phono-pre.gif
http://www.diyaudio.com/forums/atta...218589d1302834378-parasound-jc3-phono-pre.gif
Please, everyone, IF you want to build a Vendetta Clone, go for it. If you fail to get excellent results, please don't BLAME ME! I know for a fact that a successful Vendetta input stage is hard to make, properly. Even Thorsten has had trouble understanding it properly or getting one to measure right. The true noise should be 0.4nV/rt Hz from 400 Hz and above, or 85 dB IHF S/N noise ratio. IF you don't get that, you have a problem. This was independently measured by Bascom King in the Dec. 1990 issue of 'Audio' magazine. I can supply the data if necessary.
I think that this is an EXCELLENT START, MiiB. This is a GOOD solid elegant circuit. It is BETTER that the circuits that I used 30+ years ago, the input stage of the JC-80, until the Vendetta was designed and properly constructed. It is buildable, can and should be, class A, can run open loop or with feedback, balanced in or single ended. That's just for starters.
I saw the phono stage that Joarchim linked to..
It has a different approach to the Riaa-correction... through I dont fancy the output...
schematics
It has a different approach to the Riaa-correction... through I dont fancy the output...
schematics
Added output voltage & a few more carts I found inductance data for. 
AUDIO TECHNICA
AT-33EV 0.3mV 10Ω 22µH
Kiseki
Purpleheart Sapphire 0.4mV 42Ω 120µH
Audio Technica
AT-F3 III 0.35mV 12Ω 25µH
Denon
DL304 0.18 mV 40Ω 50µH
Haniwa
HCTR01 0.35mV 0Ω8 1µ3H
Dynavector (last 2 hi output)
100D 0.2mV 30Ω 80µH
17DS 0.2mV 32Ω 90µH
17D 0.2mV 32Ω 90µH
23R 0.2mV 35Ω 100µH
19A 0.25mV 32Ω 90µH
10X3 2.5mV 300Ω 0.7mH
50A 0.2mV 30Ω 80µH
XX-2 0.23mV 6Ω 18µH
DRTXV-1 0.3mV 6Ω 18µH
DRTXV-1S 0.3mV 6Ω 18µH
20A2 3.6mV 510Ω 1mH
20B2 3.6mV 510Ω 1mH
v.d.Hul
Black Beauty SPX 0.98mV 15Ω7 19µ8H
DDT II Special 0.88mV 17Ω5 15µ9H
Goldring
Eroica LX 0.5mV 8Ω 12µH
HIGHPHONIC
MC-R5 0.12mV 40Ω 100µH
MC-D15 0.12mVR 38Ω 100µH
Grado (all Moving Iron)
Statement1, Reference1, Master1, Sonata1 & Platinum1
0.5mV 2Ω 2mH
Is this a typo, could it be 2µH, with only 2Ω or is the inductandce raised that much from winding on a magnetic material?
From the Grado website
"The outstanding quality of the moving coil cartridge is not the moving coil principle but the very low DC resistance and inductance."
Rgds
James
AUDIO TECHNICA
AT-33EV 0.3mV 10Ω 22µH
Kiseki
Purpleheart Sapphire 0.4mV 42Ω 120µH
Audio Technica
AT-F3 III 0.35mV 12Ω 25µH
Denon
DL304 0.18 mV 40Ω 50µH
Haniwa
HCTR01 0.35mV 0Ω8 1µ3H
Dynavector (last 2 hi output)
100D 0.2mV 30Ω 80µH
17DS 0.2mV 32Ω 90µH
17D 0.2mV 32Ω 90µH
23R 0.2mV 35Ω 100µH
19A 0.25mV 32Ω 90µH
10X3 2.5mV 300Ω 0.7mH
50A 0.2mV 30Ω 80µH
XX-2 0.23mV 6Ω 18µH
DRTXV-1 0.3mV 6Ω 18µH
DRTXV-1S 0.3mV 6Ω 18µH
20A2 3.6mV 510Ω 1mH
20B2 3.6mV 510Ω 1mH
v.d.Hul
Black Beauty SPX 0.98mV 15Ω7 19µ8H
DDT II Special 0.88mV 17Ω5 15µ9H
Goldring
Eroica LX 0.5mV 8Ω 12µH
HIGHPHONIC
MC-R5 0.12mV 40Ω 100µH
MC-D15 0.12mVR 38Ω 100µH
Grado (all Moving Iron)
Statement1, Reference1, Master1, Sonata1 & Platinum1
0.5mV 2Ω 2mH
Is this a typo, could it be 2µH, with only 2Ω or is the inductandce raised that much from winding on a magnetic material?
From the Grado website
"The outstanding quality of the moving coil cartridge is not the moving coil principle but the very low DC resistance and inductance."
Rgds
James
Last edited:
Added output voltage & a few more carts I found inductance data for.
AUDIO TECHNICA
AT-33EV 0.3mV 10Ω 22µH
Kiseki
Purpleheart Sapphire 0.4mV 42Ω 120µH
Audio Technica
AT-F3 III 0.35mV 12Ω 25µH
Denon
DL304 0.18 mV 40Ω 50µH
Haniwa
HCTR01 0.35mV 0Ω8 1µ3H
Dynavector (last 2 hi output)
100D 0.2mV 30Ω 80µH
17DS 0.2mV 32Ω 90µH
17D 0.2mV 32Ω 90µH
23R 0.2mV 35Ω 100µH
19A 0.25mV 32Ω 90µH
10X3 2.5mV 300Ω 0.7mH
50A 0.2mV 30Ω 80µH
XX-2 0.23mV 6Ω 18µH
DRTXV-1 0.3mV 6Ω 18µH
DRTXV-1S 0.3mV 6Ω 18µH
20A2 3.6mV 510Ω 1mH
20B2 3.6mV 510Ω 1mH
v.d.Hul
Black Beauty SPX 0.98mV 15Ω7 19µ8H
DDT II Special 0.88mV 17Ω5 15µ9H
Goldring
Eroica LX 0.5mV 8Ω 12µH
HIGHPHONIC
MC-R5 0.12mV 40Ω 100µH
MC-D15 0.12mVR 38Ω 100µH
Grado (all Moving Iron)
Statement1, Reference1, Master1, Sonata1 & Platinum1
0.5mV 2Ω 2mH
Is this a typo, could it be 2µH, with only 2Ω or is the inductandce raised that much from winding on a magnetic material?
From the Grado website
"The outstanding quality of the moving coil cartridge is not the moving coil principle but the very low DC resistance and inductance."
Rgds
James
Hi James,
Thanks for putting together this excellent compilation. It is striking how many low-output MC cartridges have inductance considerably greater than 10 uH, and several over 50uH. This shows that it is unwise to make broad generalizations.
In fairness to their use in virtual ground MC head amps, many of those with higher inductance have correspondingly higher DCR (as might be expected). This will tend to mitigate HF rolloff due to the higher inductance.
Cheers,
Bob
Ed I'm afraid you lost me on this one, though it scares me that I once got it "right". I just inherited the last two years of AudioXpress so I can catch up.
Last edited:
This alone is not that difficult. Before you start in again the compromises and supply currents involved do not make for a commercial product anywhere but MC preamps.
Last edited:
It has always been a challenge to make extremely noise designs that also measure pretty well, and especially without coupling caps. I could have, in principle, gotten down to 0.4nV/rt Hz or better, with paralleled transistors, fairly easily, as early as 1968, when I found the relatively low Rbb' of the 2N4401, 2N4403, and especially the 2N4405 parts from Motorola, released that year. Today, we can do it with even one part, usually a PNP.
In those days jfets were not developed enough and were VERY NOISY in comparison. By 1971 or so, however, Crystalonics (sp) released the C413N, or later, the 2N6550 jfet. I have measured less than 1 nV/rt Hz with these devices in the midrange and some were very good down to 10 Hz. This opened up the use of jfets for low noise input stages. Unfortunately, the manufacturer did not GUARANTEE the low noise of the parts, so I have handful still that are completely useless for audio input stages. However, Siliconix started to make high current switches, at 1/5 the price or less, and they measured essentially the same. This was the input device used, the J110, in the early Levinson JC-2 phono stage. Later these devices got noisy, and we were stuck once again. In 1978 or so, Toshiba introduced the 2SK146 dual pair. This was VERY quiet, and a superior replacement for the J110 or 2N6550 in virtually every respect.
Still, there was no complement for any of these devices. About this time, however, Hitachi came up with a complementary pair of low noise jfet devices, that were pretty good, but then Toshiba came up with the 2SJ72, and this fit the bill. Before, we had achieved 0.4V/rt Hz only with bipolars, but now, we could make a direct coupled circuit input and output with complementary jfets, and the Vendetta was born. I have kept the .4nV/rt Hz standard for the last 38 years for a commercial product, but lower is possible, yet it is getting harder to get the jfets that do the job. They just don't want or precisely know how to make them, anymore.
What is ironic is that in 1967, when I told Ortofon that I was going to make a transformerless mc preamp, they showed me the door.
In 1973, when Mark Levinson and I offered to license them my design, they were unimpressed, and in 1978, they claimed to have designed it themselves.
It is the same, historically:
It is impossible.
It is possible, but unimpressive.
We invented it.
And so it goes.
In those days jfets were not developed enough and were VERY NOISY in comparison. By 1971 or so, however, Crystalonics (sp) released the C413N, or later, the 2N6550 jfet. I have measured less than 1 nV/rt Hz with these devices in the midrange and some were very good down to 10 Hz. This opened up the use of jfets for low noise input stages. Unfortunately, the manufacturer did not GUARANTEE the low noise of the parts, so I have handful still that are completely useless for audio input stages. However, Siliconix started to make high current switches, at 1/5 the price or less, and they measured essentially the same. This was the input device used, the J110, in the early Levinson JC-2 phono stage. Later these devices got noisy, and we were stuck once again. In 1978 or so, Toshiba introduced the 2SK146 dual pair. This was VERY quiet, and a superior replacement for the J110 or 2N6550 in virtually every respect.
Still, there was no complement for any of these devices. About this time, however, Hitachi came up with a complementary pair of low noise jfet devices, that were pretty good, but then Toshiba came up with the 2SJ72, and this fit the bill. Before, we had achieved 0.4V/rt Hz only with bipolars, but now, we could make a direct coupled circuit input and output with complementary jfets, and the Vendetta was born. I have kept the .4nV/rt Hz standard for the last 38 years for a commercial product, but lower is possible, yet it is getting harder to get the jfets that do the job. They just don't want or precisely know how to make them, anymore.
What is ironic is that in 1967, when I told Ortofon that I was going to make a transformerless mc preamp, they showed me the door.
In 1973, when Mark Levinson and I offered to license them my design, they were unimpressed, and in 1978, they claimed to have designed it themselves.
It is the same, historically:
It is impossible.
It is possible, but unimpressive.
We invented it.
And so it goes.
Last edited:
Hi,
I don't know. The Vendetta sure is very good, justly legendary and there is a certain elegance that suffuses the design that is appealing.
However, you have to re-tune the circuit for using different J-Fets. No need to use V-Grade, use BL Grade, adjust the circuit values.
Also, use the later version schematic, hint, the inverting input of the output amp is used and the 750 Ohm forms the input resistor of the inverting Amp AND input stage load.
Incidentally, I used this same style once with an INA103 set up as Howland Current Pump on the input and an OPA637 in the second stage, powersupplies inspired by Sutherland/Emitter without regulation, just passive filtering LC and RC, EQ using NOS precision Tinfoil/Polystyrene Cap's.
This whole thing worked very well. The "Current Pump + Inverting Amp" active/passive EQ sounded certainly better than a normal passive EQ for HF and active for the rest. Despite being full of Op-Amp's and stuff that rarely gets respect the actual Phono sounded extremely good, leaving some very expensive commercial units in the dust.
Don't ask for a schematic, I don't have any drawn out, it was a one-off build on a lark. Anyway, it is easy to work out, just stick an OPA637 into the feedback loop of the 2nd stage of the Vendetta (C) and set up an INA103 as Howland Current Pump with around 100mA/V (I used external resistors for the frontend) and add a decently clean +/-16V supply.
Well, you could use the Vendetta frontend with 4 n-channel transistors and only the N-Channel halve.
Well, colour me bad, I have this tendency to consider all active supplies to be noisy, so I tend to use RC Filter chains after fairly pedestrian regulators that help to get the DC stable...
For example 33R/2200uF form a 2Hz lowpass, which will attenuate noise at 20Hz by around 20dB and more higher up. Daisychain a few and noise becomes limited by the passive components. As long as the circuits supplied are strictly class A I find this to work very well...
Ciao T
I don't know. The Vendetta sure is very good, justly legendary and there is a certain elegance that suffuses the design that is appealing.
However, you have to re-tune the circuit for using different J-Fets. No need to use V-Grade, use BL Grade, adjust the circuit values.
Also, use the later version schematic, hint, the inverting input of the output amp is used and the 750 Ohm forms the input resistor of the inverting Amp AND input stage load.
Incidentally, I used this same style once with an INA103 set up as Howland Current Pump on the input and an OPA637 in the second stage, powersupplies inspired by Sutherland/Emitter without regulation, just passive filtering LC and RC, EQ using NOS precision Tinfoil/Polystyrene Cap's.
This whole thing worked very well. The "Current Pump + Inverting Amp" active/passive EQ sounded certainly better than a normal passive EQ for HF and active for the rest. Despite being full of Op-Amp's and stuff that rarely gets respect the actual Phono sounded extremely good, leaving some very expensive commercial units in the dust.
Don't ask for a schematic, I don't have any drawn out, it was a one-off build on a lark. Anyway, it is easy to work out, just stick an OPA637 into the feedback loop of the 2nd stage of the Vendetta (C) and set up an INA103 as Howland Current Pump with around 100mA/V (I used external resistors for the frontend) and add a decently clean +/-16V supply.
Well, you could use the Vendetta frontend with 4 n-channel transistors and only the N-Channel halve.
Well, colour me bad, I have this tendency to consider all active supplies to be noisy, so I tend to use RC Filter chains after fairly pedestrian regulators that help to get the DC stable...
For example 33R/2200uF form a 2Hz lowpass, which will attenuate noise at 20Hz by around 20dB and more higher up. Daisychain a few and noise becomes limited by the passive components. As long as the circuits supplied are strictly class A I find this to work very well...
Ciao T
I have a couple of those, 741 circuit with inputs running at 25mA a side all integrated on one chip. I think they measured .3nV or so. Never went commercial. Knowledge he role of rbb' in noise long predated 1968. It is impossble simply does not apply or ever did.
Last edited:
John,
Which part? Doing the Circuit SE and only N-Channel Fet's?
Sure, it is a compromise. But it works for "Papa Pass" non too shabbily and if we fold the BJT cascode instead of doing it the Pass way we are close to what I am suggesting.
Is it a compromise compared to the original vendetta, sure, but tghen again, everything is a compromise, one must choose which compromises are preferred.
I am not sure if it is "bad" or "good" advise, it was meant to advise how to get reasonable resulting using N-Channel low noise Fet's only.
Ciao T
Which part? Doing the Circuit SE and only N-Channel Fet's?
Sure, it is a compromise. But it works for "Papa Pass" non too shabbily and if we fold the BJT cascode instead of doing it the Pass way we are close to what I am suggesting.
Is it a compromise compared to the original vendetta, sure, but tghen again, everything is a compromise, one must choose which compromises are preferred.
I am not sure if it is "bad" or "good" advise, it was meant to advise how to get reasonable resulting using N-Channel low noise Fet's only.
Ciao T
No, Scott, Bybee is impossible. Low Rbb' was IMPROBABLE. It is difficult to make a generalization fit each and every situation perfectly.
The REASON that I found these devices was because I LOOKED for them. I knew from the IEE 'Electronics Letters' paper in 1966, that low Rbb' was the key. When I went to Ortofon in 1967, THEY told me that their best engineers had studied the problem and found it 'impossible' then they showed me the door.
Only when I started working at Ampex did I have access to a QuanTech Noise analyzer. I was working in in the Instrumentation Division when I was asked to evaluate the En of of number of devices that we used at Ampex, by my boss. He went on to be a professor of Engineering, so he know a lot more about it than me. In any case, the QuanTech was located in the Audio Division, so I got permission to go down the hall and measure the devices we were interested in. Then I thought, why not look at ALL the devices that we use and stock? I found the 2N2905, not too bad. Then, within the year, Motorola came out with a whole bunch of interesting devices like the 2N4126, the 2N4403, and finally, the 2N4405 (the overall low Rbb' champ at the time). I found these devices, BECAUSE I was interested in making MC preamps sometime in future, and I expended the extra effort to find them. Anyone could have done it, IF they had the QuanTech and the desire, but most people don't have this drive, as you well know.
It is amazing how long it took for others to find what I found. Many years!
Hindsight is easy, clearing the path takes effort.
The REASON that I found these devices was because I LOOKED for them. I knew from the IEE 'Electronics Letters' paper in 1966, that low Rbb' was the key. When I went to Ortofon in 1967, THEY told me that their best engineers had studied the problem and found it 'impossible' then they showed me the door.
Only when I started working at Ampex did I have access to a QuanTech Noise analyzer. I was working in in the Instrumentation Division when I was asked to evaluate the En of of number of devices that we used at Ampex, by my boss. He went on to be a professor of Engineering, so he know a lot more about it than me. In any case, the QuanTech was located in the Audio Division, so I got permission to go down the hall and measure the devices we were interested in. Then I thought, why not look at ALL the devices that we use and stock? I found the 2N2905, not too bad. Then, within the year, Motorola came out with a whole bunch of interesting devices like the 2N4126, the 2N4403, and finally, the 2N4405 (the overall low Rbb' champ at the time). I found these devices, BECAUSE I was interested in making MC preamps sometime in future, and I expended the extra effort to find them. Anyone could have done it, IF they had the QuanTech and the desire, but most people don't have this drive, as you well know.
It is amazing how long it took for others to find what I found. Many years!
Hindsight is easy, clearing the path takes effort.
Continuing with this topic, I would like to mention that in 1978, when I first found that Ortofon introduced a pre-preamp with almost exactly the characteristics of the design that they had rejected from Mark Levinson and me, years before, I wrote to a former colleague, Harbo Anderson a letter stating that Ortofon was using my patented design, and I had been taken advantage of, by him, since I had used HIS chalkboard to show why my PATENTED design was really better than typical designs, in his office at Ortofon in Denmark, because I knew that the patent had been applied for, and therefore should be safe from theft.
He replied in about the same way that Scott does when discussing these circuits, that the patented circuit was OBVIOUS, the devices were popular, and besides he had made some minor changes to the design. Sound familiar, anyone?
Of course, ONCE you see and get to know the circuit, it is OBVIOUS! Many of my designs were NOT obvious when they were first introduced, but they sure got popular afterward. And so it goes, and WHY I tend not to publish my internal schematics, here.
He replied in about the same way that Scott does when discussing these circuits, that the patented circuit was OBVIOUS, the devices were popular, and besides he had made some minor changes to the design. Sound familiar, anyone?
Of course, ONCE you see and get to know the circuit, it is OBVIOUS! Many of my designs were NOT obvious when they were first introduced, but they sure got popular afterward. And so it goes, and WHY I tend not to publish my internal schematics, here.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.