1) agreed
2) agreed; it re-introduces ground as a reference, negating part of the advantage of balanced lines.
3) not agreed, see my reply to FdW above.
Jan
Where can we read about this ? Please at least one published
example of an instrumentation amp with different "legs".
1:3 will do fine
Yes, the answer was a bit verbose, I just noted that he should check the input currents at both legs (using the simulator it's very easy).
P.s. The answer was here MPP e.g. #11349
Last edited:
Use your calculator/confusor, I can not help that most are not in balance... go figure...
Last edited:
P.s. Why I did note that 2 different resistor values where needed.
In general I see electronic circuits as current-defined (not voltage) and that changes your viewpoint, when calculating resistor/capacitor values for operational amplifier circuits you will be surprised what you discover if you change your viewpoint.
Unbalanced circuits offer gain to unbalanced signal, e.g. most likely the signal you do not appreciate
Balanced circuits offer gain to balanced signal, e.g. the signal you intended to handle (in the balanced circuit)
Be sure to balance for voltage an current alike (e.g. impedance)
And; Using the multimeter it is easy to measure voltages (in circuit) but it is near to impossible to measure currents (in circuit, while operating and keeping the work-progress going), for this reason we tend to ignore current, maybe the possibility to always measure (most) any thing in simulation makes the simulator the best tool available (e.g. measure your impedances with Spice you can)
In general I see electronic circuits as current-defined (not voltage) and that changes your viewpoint, when calculating resistor/capacitor values for operational amplifier circuits you will be surprised what you discover if you change your viewpoint.
Unbalanced circuits offer gain to unbalanced signal, e.g. most likely the signal you do not appreciate
Balanced circuits offer gain to balanced signal, e.g. the signal you intended to handle (in the balanced circuit)
Be sure to balance for voltage an current alike (e.g. impedance)
And; Using the multimeter it is easy to measure voltages (in circuit) but it is near to impossible to measure currents (in circuit, while operating and keeping the work-progress going), for this reason we tend to ignore current, maybe the possibility to always measure (most) any thing in simulation makes the simulator the best tool available (e.g. measure your impedances with Spice you can)
Last edited:
I agree, it’s always the same.
And the two amps driving this instrumentation amp see different impedances, but with values in the K ohm this has no effect whatsoever on their output voltage and even if it does, it has no effect because the INA subtract both inputs.
The tolerance of the resistors around the INA has a much larger effect.
Hans
Did anyone ever try, or consider to try, LU1014D in a phono input stage. Considering the size of the chip they should also be very low noise.
- 1/f noise corner frequency is well over the audio band, which makes it improper for audio low noise applications. In fact, it will be as noisy as any regular (non low noise) JFET.
- Ciss is over 1nF. That makes it completely improper for MM application, since the cartridge resonance will be in the audio band, in the KHz.
- Gate current is unspecified, practically around 100nA, which is huge comparing to a low noise JFET. Meaning the input current noise will be a serious problem for MM cartridges.
Don't even think about it.
Last edited:
This is not the case.
You are thinking in terms of differential loading not common mode.
With common mode signals both see the same load with equal R values.
In your mind hold the OP at 0V and imagine both + and - signals going up
(+), they see same impedance.
There are other advantages such as proper cancellation of CM distortions.
TCD
Thank you all for the replies. I thought of trying them in an MC stage, off course, not an MM stage.
I bought them primarily to build a firstwatt F3 clone but since they were "cheap" at the time I ordered a bigger number of devices. Unfortunately I have only limited time and other projects come first, so it will be a while before I may try this.
I bought them primarily to build a firstwatt F3 clone but since they were "cheap" at the time I ordered a bigger number of devices. Unfortunately I have only limited time and other projects come first, so it will be a while before I may try this.
Last edited:
1) C15/C16 are DC blocking caps but also part of the riaa compensation.
If removed, the riaa filter must be modified to cope with the extra bass boost.
The circuit works without them as long as the 1st stage components are all extremely well matched.
If there is a small mismatch (for instance R2 = 401 instead of 400) the circuit still works but we have already 6V offset in the output.
If the mismatch is more serious (R2 = 405 instead of 400) the circuit does not amplify and we have 13V offset in the output.
Anyway this was a very good point as I also do not like the big 10u caps but something must be done to avoid collapsing so I added two small 100n caps after the RIAA filter, recalculated the input resistors on the RIAA and the circuit works flawlessly now.
2) You can remove this connection to GND and will not see any difference in the circuit operation. I did follow the instructions found here Still More Phono Preamps
"On the other hand, adding just one more capacitor and referencing the 2122Hz pole to ground will preserve a much more accurate final equalization, in spite of slight imbalances. (This is something will not show in a SPICE simulation, unless you go looking for it, as SPICE parts come perfectly matched.)"
3)I did see better current matching in the output buffer with this resistor combination suggested by Frans......
Below the new