I recomend simulating this circuit. Then you should be able to look at the trade offs at least in the opamp. As you reduce the input impedance you are increasing the amplifer gain and the input noise of the opamp will at some point become the dominant factor rather than johnson thermal from the input resistor. Also as the gain goes up problems with instability may become an issue due to parasitic capacitance. At the limit (i.e if you continued to reduce the impedance) the output would be driving into the virtual earth as a current amp. I am not sure what the effect of this would be, but clearly you will start to approach this as you reduce the impedance.
I am unclear as to the benifit of this balanced input. It provides an asymetrical load to the cartridge as one end is refered to a virtual earth, approx impedance 0 Ohm and the other is refered to ground through the combination of 51R and 6.19K.
I would think a true balanced approach would have a symetrical load L*2 ( whatever your chosen load is) on both sides and use two virtual earths then combine the signals in another stage. However this is just conjecture for your interest, I haven't really thought it through to a conclusion.
Regards,
Andrew
I am unclear as to the benifit of this balanced input. It provides an asymetrical load to the cartridge as one end is refered to a virtual earth, approx impedance 0 Ohm and the other is refered to ground through the combination of 51R and 6.19K.
I would think a true balanced approach would have a symetrical load L*2 ( whatever your chosen load is) on both sides and use two virtual earths then combine the signals in another stage. However this is just conjecture for your interest, I haven't really thought it through to a conclusion.
Regards,
Andrew
That is true, and it is exactly what I am doing, but as the exact data for the cartridge is not know (other than that the supplier notes an internal resistance of 5ohm and a input impediance for the preamp of 100ohms) I have no way of getting more figures on that. That was why I posed the ‘general’ question about the issue.
The configuration, as given, gives a symmetrical (what ever that means for a cartridge with 2 wires per channel) load of 102.2 ohms (2 times 51.1). Look up the wiki reference a few entries back.
Regards,
Frans de Wit.
Last edited:
I agree with FdW and Gerhard regarding load.
Unfortunately the opamp input sees the cartridge signal in series with 102.2 ohms. The signal should be in parallel with 102.2 ohms for lowest input noise. To realize this with DC-coupled balanced input requires more active devices than one op amp, for example:
http://upload.wikimedia.org/wikipedia/commons/e/ed/Op-Amp_Instrumentation_Amplifier.svg
Unfortunately the opamp input sees the cartridge signal in series with 102.2 ohms. The signal should be in parallel with 102.2 ohms for lowest input noise. To realize this with DC-coupled balanced input requires more active devices than one op amp, for example:
http://upload.wikimedia.org/wikipedia/commons/e/ed/Op-Amp_Instrumentation_Amplifier.svg
The wiki gives no formulas for Zin (not for differential and not for common mode). So I looked it up, the Zin you are referring to is called 'Common mode input impedance' and this one is not important for the cartridge (it is for the common mode signal). It is the differential impedance that is important for the cartridge and it is (in the given case) 102.2ohm.
The information and the formula's can be found on:
Differential Amplifier Common Differential Modes
Regards,
Frans.
Attachments
In this case you did build a high Zin differential amp, and then lowered the deferential input impedance of the circuit by adding resistors parallel to the input of the amp. This works fine, but was not selected by me, due to the fact that I like a as low as possible components count (and this configuration adds at the least two opamps, and one in the signal path).
The configuration that I selected has a low differential input resistant and thus does not need the two ‘extra’ opamps. This is not the best solution in regard to common mode rejection, but it is more of a compromise. I used the following reasoning; the ‘standard’ circuit has no common mode rejection at all, the selected one has (using the .1% resistors) around 80db (or more), and that is a good thing, especially as no extra components are needed (compared to the ‘standard’ solution).
Regards,
Frans.
BTW; There is no such thing as a free lunch.
Last edited:
Hi FdW -
In the example I posted, and using the definitions in your link, Z1,Z2 and Z3 are infinite with ideal op amps, in reality they will be limited by the op amp input characteristics but will be much higher than 100 ohms to well above 20kHz.
In the example I posted, the cartridge and its 100 ohm load would be connected between pins V1 and V2, which places the cartridge in parallel with the 100 ohms. Gerhard's design works like this.
Have you consider transformer step-up for MC?
In the example I posted, and using the definitions in your link, Z1,Z2 and Z3 are infinite with ideal op amps, in reality they will be limited by the op amp input characteristics but will be much higher than 100 ohms to well above 20kHz.
In the example I posted, the cartridge and its 100 ohm load would be connected between pins V1 and V2, which places the cartridge in parallel with the 100 ohms. Gerhard's design works like this.
Have you consider transformer step-up for MC?
Simulation results for the LME49710 show, CMR@10Khz 70db and CMR@30Khz 60db(better than). These figures are good enough for me (given the simple circuit).
A sample audio file can be found at:
http://www.diyaudio.com/forums/digital-line-level/35621-diy-adcs-4.html#post2015941
Yes I did consider adding a step-up and rejected it, I know there are many pro's (and even con's
) but these counted the most; transformer CMR (not the best there is [or you need a very, very, very… spectacular expensive transformer]) and transformer noise. Plus it adds an extra component to the equation, and as before, for me, simpler is better!Regards,
Frans de Wit.
Trust me, most manufactures have no idea how to best load their cartridges. I build everything from a virtual 0 Ohm input to 47kOhm loading, balanced and un-balanced and all circuits worked just fine when optimised.. Over the weekend i build a low noise version of your input stage and i will report how it performs against my INA.
I am using the OPA1611 that waited to be put into good use.
I am using the OPA1611 that waited to be put into good use.
Cool, I’m looking forward to that.
The thing I am doing is; design the shunt regulators. And optimize them using the sim.
Also, the cartridge manufacturer gives an internal (generator) resistance of 5ohm, selecting 20x 5ohm as the impedance for the preamp is not that bad. I assume you could go as low as 4x 5ohm and loose only a few micro volts drive (of the 600 uV available).
Regards,
Frans de Wit.
Last edited:
This is a design from a totally different perspective than the one I created. For one, it is way more complicated
It would really be nice if we could someday compare our results, some kind of 'pre-amp-battle' who knows, maybe one day.Having a good look at it, there is one thing I wonder about; the capacitors C2/C3 and C15/C16 around OP3 must be of extreme small tolerances, the normal low tolerance capacitors are 1% and I think that will not do (gives only 40db CMRR). Are you going to select matched values?
Regards,
Frans de Wit.
I'm almost done with the shunt regulator design
P.s. Sorry to react a bit late.
Frans, you may only lose a few microvolts of signal level - but I believe you will losse a HUGE amount of fine detail, resolution and ambience.
I am a strong supporter of very lightly loading allmost all MCcarts (47k) as it always sounds more alive to me. The only cart I have tried that needed a lowish value of loading R is the IKEDA, which needed something like 100R to sound really natural.
But then, my preamps are tube with 250 to 300V of supply voltage, and around 1 MHz of zero NFB bandwidth, so overload from any ultrasonic resonances just do NOT affect them.
Regards, Allen
I build your input circuit over the weekend with different values. I am using 10 Ohm resistors at the input and 200 Ohm for feedback. I must admit it sounds excellent. No big difference to my post 45 design. See more on my MPP tread. I use the OPA1611.
Yes, the caps in my discete INA have to be selected.
Yes, the caps in my discete INA have to be selected.
Allen, come down to earth please. When you visited my we used the input circuit of post 45 and i had loaded down the cart with two 300 ohm resistors to ground. We had tremendoes, insane good sound. You build with tubes and i build with silicon. That can not be compared. I tryed 47kOhm with my circuit and it sounded thin and boneless.
Ikeda
Hi Joachim,
Did you use a IKEDA cart when you listen to Mr. Wright's amp's?
I've listen to the IKEDA but was not my cup of tea! It's a good cart. don't get me wrong.
But I prefer the Shelter 901 and that's the one I'm using at the moment. But I do use the IKEDA arm, it's a straight one not the curved one!
If Mr. Wright want to listen to a MC cart at 47K than we got to respect that!
Again it's a matter of taste, and we cannot argue about that.
Just my 2 cents.
Best regards,
Audiofanatic
Hi Joachim,
Did you use a IKEDA cart when you listen to Mr. Wright's amp's?
I've listen to the IKEDA but was not my cup of tea! It's a good cart. don't get me wrong.
But I prefer the Shelter 901 and that's the one I'm using at the moment. But I do use the IKEDA arm, it's a straight one not the curved one!
If Mr. Wright want to listen to a MC cart at 47K than we got to respect that!
Again it's a matter of taste, and we cannot argue about that.
Just my 2 cents.
Best regards,
Audiofanatic
I agree ! I just find it not good to jump to quick conclusions and then never go back. We used a Lyra Titan i and i tryed all kinds of resistors and loading before Allen arived.
Be not affraid, Allen and i have more in common than you think. We are fighting quite fircely for the same goal: the best posible sound. I have a lot of respect for Allens work and i asume he respects me too. If we can share effords we may have an even better sound in some month. I eagerly await his amp back and i am working on even better phonostages so this is the never ending story.
P.S. : never heard a Shelter so far. What i whould really like to try is the Hyper Eminent though.
Be not affraid, Allen and i have more in common than you think. We are fighting quite fircely for the same goal: the best posible sound. I have a lot of respect for Allens work and i asume he respects me too. If we can share effords we may have an even better sound in some month. I eagerly await his amp back and i am working on even better phonostages so this is the never ending story.
P.S. : never heard a Shelter so far. What i whould really like to try is the Hyper Eminent though.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.