How about the Pete Millet LR Phono? I've built one but am yet to test it. It uses the AD797 that many in this thread have been talking about but the signal path is free of capacitors...inductors and resistors only, including in the passive RIAA stage.
Here is a link.
Here is a link.
Hi,
I wonder about the choice of the AD797 in that P.Millet phono.
It the AD797 were used for a low-MC or high-MC input stage ok, but for a phono MM input stage it doesn´t even better the old NE5534.
The same holds true for the second stage, where the input impedance the AD797 sees is considerable higher than its optimal noise impedance.
The ADs also run unlimited in bandwidth ... inviting oscillations ... a small cap across the feedback resistor may be a desirable safety measure -with suboptimal PCB layouts it might even be a necessity.
In comparison inductors are alot worse than caps ... less linear, larger tolerances, more parasitics and prone to any magnetic fields.
P.Millet noted in his LRPhono-pdf : "I looked at active LR EQ (in the feedback loop of an opamp), but soon discovered that inductor imperfections made it very difficult to create a stable design ... Has anybody succeeded in building an active LR EQ?"
Says alot, doesn´t it?
So be aware that this circuit may have issues ... asking for tweaking and reworking ... I´d regard it rather for the more experienced builder who can track down and solve problems .... who owns and uses a proper set of measurement devices.
Here´s a amplitude plot of a PlatINA-style phono with RC-RIAA (Neumann, subtract +2dB@20kHz for pure RIAA).
Linearity and channel match is clearly better.
jauu
Calvin
I wonder about the choice of the AD797 in that P.Millet phono.
It the AD797 were used for a low-MC or high-MC input stage ok, but for a phono MM input stage it doesn´t even better the old NE5534.
The same holds true for the second stage, where the input impedance the AD797 sees is considerable higher than its optimal noise impedance.
The ADs also run unlimited in bandwidth ... inviting oscillations ... a small cap across the feedback resistor may be a desirable safety measure -with suboptimal PCB layouts it might even be a necessity.
In comparison inductors are alot worse than caps ... less linear, larger tolerances, more parasitics and prone to any magnetic fields.
P.Millet noted in his LRPhono-pdf : "I looked at active LR EQ (in the feedback loop of an opamp), but soon discovered that inductor imperfections made it very difficult to create a stable design ... Has anybody succeeded in building an active LR EQ?"
Says alot, doesn´t it?
So be aware that this circuit may have issues ... asking for tweaking and reworking ... I´d regard it rather for the more experienced builder who can track down and solve problems .... who owns and uses a proper set of measurement devices.
Here´s a amplitude plot of a PlatINA-style phono with RC-RIAA (Neumann, subtract +2dB@20kHz for pure RIAA).
Linearity and channel match is clearly better.
jauu
Calvin
Attachments
Last edited:
Yeah, you may well be right, but Pete who is not prone to embellishment does go on to say that it sounds better than his other (RC) phonostages.
My personal experience with choosing inductors or caps in the signal path is that I have always preferred to get rid of the cap. This is without exception.
Last edited:
Seems the OP has Left the building.🙄
While various factions argue the Number of Angels on the head of a Pin.
A simple stupid Phono like a Le Pacific Will please.
Simplest/fewest parts has Serious merit
While various factions argue the Number of Angels on the head of a Pin.
A simple stupid Phono like a Le Pacific Will please.
Simplest/fewest parts has Serious merit
Currently playing on a Boozehound i tried and bought. Plays extremely well with my system.
As Jason is on a long holiday I had to find an already finished one.
As Jason is on a long holiday I had to find an already finished one.
I use Erno Borbely slightly modified EB-804/419 (added two more JFETS pairs in parallel at the input stage, decreasing R7/R8 to 348 Ohm - so use 4 JFETs in parallel instead of two to decrease noise) for a long time - it is quite expensive to build, but, on my opinion, it worth it.
Boozhound if you purchase just the pcb its a quick, easy and inexpensive build. Its a nice phono and good value vs performance.
Pearl is more expensive but equally from what I have read significantly better performance. I have built the BH in the past and considering the Pearl at the moment.
Pearl is more expensive but equally from what I have read significantly better performance. I have built the BH in the past and considering the Pearl at the moment.
Yes Boozhound is sure VFM. I have listened to both of them. AD797 phonostage betters both of them for LOMC use.
Regards
Sachin
Hi,
not specifically a lownoise design.
Have You thought about the AD745/743 instead of the OPA604 as MM input?
And paralleling the JFETs or using others for MC?
jauu
Calvin
not specifically a lownoise design.
Have You thought about the AD745/743 instead of the OPA604 as MM input?
And paralleling the JFETs or using others for MC?
jauu
Calvin
It is a passive riaa eq so the s/n can't be very high.
The s/n is close to 80 dB so it is enough, measured.
This ratio still the same with MC stage running.
Paralleling the fet was not so important for me; it has a good overload of 6mV rms in input.
The MM has a max overload of 155 mV at 1 khz
The main goal is to reach the best riaa response with a good overload.
The Opa2604 has a max Vdc at +/- 24 volt and for this reason I use it.
Walter
The s/n is close to 80 dB so it is enough, measured.
This ratio still the same with MC stage running.
Paralleling the fet was not so important for me; it has a good overload of 6mV rms in input.
The MM has a max overload of 155 mV at 1 khz
The main goal is to reach the best riaa response with a good overload.
The Opa2604 has a max Vdc at +/- 24 volt and for this reason I use it.
Walter
With the Pearl 2, purchased with JFET's you know they are real.
I'm building an EAR834, but banging my head up against a PSU that won't hold a steady voltage.
I'm building an EAR834, but banging my head up against a PSU that won't hold a steady voltage.
Hi,
@Waltube
no critique just hints where unused potential might be unlashed 😉
With passive RIAA only the input OPAmp may become the object of overload.
But typical +-15V will suffice comfortably if the first stage´s gain is not too high ... gains that the OPA604 isn´t good for anyway.
24V to 15V means a difference of only 4dB.
I´d rather have an eye on the noise performance where the OPA604 is critical also.
It´s high voltage noise will lead to raised LF-noise (up to ~1-2kHz).
Among the FET-input OPAs the afore mentioned AD745/743 would result in lower noise.
The OPA604 is perfectly ok for the second gain stage.
How did You derive the noise figures?
As You don´t state any voltage or reference levels the figure in itself is of no use.
How many mV in? How many mV out?
Which source impedance value?
Linear weighting or A-weighting?
jauu
Calvin
@Waltube
no critique just hints where unused potential might be unlashed 😉
With passive RIAA only the input OPAmp may become the object of overload.
But typical +-15V will suffice comfortably if the first stage´s gain is not too high ... gains that the OPA604 isn´t good for anyway.
24V to 15V means a difference of only 4dB.
I´d rather have an eye on the noise performance where the OPA604 is critical also.
It´s high voltage noise will lead to raised LF-noise (up to ~1-2kHz).
Among the FET-input OPAs the afore mentioned AD745/743 would result in lower noise.
The OPA604 is perfectly ok for the second gain stage.
How did You derive the noise figures?
As You don´t state any voltage or reference levels the figure in itself is of no use.
How many mV in? How many mV out?
Which source impedance value?
Linear weighting or A-weighting?
jauu
Calvin
Hi
the 78 dB reached in MM with 5 mV in and same in MC with 0,5 mV in as the intl specs; the MM has a gain of 40dB; with MC it is around 68 dB.
The source is 25 ohm in both case, it is AP1 and 2
The s/n is weighted A as intl specs.
It is a good results for the topology used.
Regarding thr overload this was I my idea to get the max possible and you can reach this easily ( less or more) with two cheap batteries of 12v in series.
In the my article in Audioreview I have mentioned others opamp can be used (with some mod on layout) and the LT1115 is the best of noise, 0,9nV but it is single and the max dc voltage is 18 volt that is also fine, it is possible to use 3+3 x 6 volt 1A batteries in series.
Walter
the 78 dB reached in MM with 5 mV in and same in MC with 0,5 mV in as the intl specs; the MM has a gain of 40dB; with MC it is around 68 dB.
The source is 25 ohm in both case, it is AP1 and 2
The s/n is weighted A as intl specs.
It is a good results for the topology used.
Regarding thr overload this was I my idea to get the max possible and you can reach this easily ( less or more) with two cheap batteries of 12v in series.
In the my article in Audioreview I have mentioned others opamp can be used (with some mod on layout) and the LT1115 is the best of noise, 0,9nV but it is single and the max dc voltage is 18 volt that is also fine, it is possible to use 3+3 x 6 volt 1A batteries in series.
Walter
Hi,
"LT1115 is the best of noise ..."
It certainly is a fine device for MCs as its noise minimum falls into the range of ~50Ohm to ~1kOhm.
MMs copper resistance already ranges from ~400Ohm (AT) to 1.6kOhm (Shure) and rises quickly dur to their high coil inductances.
In that range the current noise figure of the OPA dominates.
For a bipolar input OPA the LT1115 performs quite well in this regard, but I doubt it´d be the best choice for MMs noisewise.
One point that makes FET-Input OPAs so attractive for MMs is their by decades lower noise current value, so that they are virtually noise-transparent at least from a couple of hundred Hz on, while bipolar OPAs are getting noisy with rising frequency.
I´m sure that the AD745/743 will give lower noise figures than the LT1115.
I´m not convinced sonically by that device either ... but that is a different field 😉
jauu
Calvin
ps. could You provide me with a copy of Your article?
"LT1115 is the best of noise ..."
It certainly is a fine device for MCs as its noise minimum falls into the range of ~50Ohm to ~1kOhm.
MMs copper resistance already ranges from ~400Ohm (AT) to 1.6kOhm (Shure) and rises quickly dur to their high coil inductances.
In that range the current noise figure of the OPA dominates.
For a bipolar input OPA the LT1115 performs quite well in this regard, but I doubt it´d be the best choice for MMs noisewise.
One point that makes FET-Input OPAs so attractive for MMs is their by decades lower noise current value, so that they are virtually noise-transparent at least from a couple of hundred Hz on, while bipolar OPAs are getting noisy with rising frequency.
I´m sure that the AD745/743 will give lower noise figures than the LT1115.
I´m not convinced sonically by that device either ... but that is a different field 😉
jauu
Calvin
ps. could You provide me with a copy of Your article?
Hi
in the table attached there is a list of the performace on noise.
As you see the LT1115 is one of the best also for different range od source resistor.
In every case, after some test I decided to use the 2604 because the value of 78 dB for the entire circut is reasonable mainly for the topology used and it is good for MC.
In addition the riaa curve is very accurate with a little selection of caps; the resistors are Holco 0,5%-1% H8
About the article I try to send you
Walter
in the table attached there is a list of the performace on noise.
As you see the LT1115 is one of the best also for different range od source resistor.
In every case, after some test I decided to use the 2604 because the value of 78 dB for the entire circut is reasonable mainly for the topology used and it is good for MC.
In addition the riaa curve is very accurate with a little selection of caps; the resistors are Holco 0,5%-1% H8
About the article I try to send you
Walter
- Status
- Not open for further replies.