Have a look at the schematic in post 11309, bases are at 0V,
so emitters must be at -0.6V, collector - and reference voltage -
should be (at least slightly) positive.
Oscillation can probably be prevented by small caps between
ops output and neg input.
so emitters must be at -0.6V, collector - and reference voltage -
should be (at least slightly) positive.
Oscillation can probably be prevented by small caps between
ops output and neg input.
I got rid of the oscilation (near 6MHz) with a zobel in the output of 1st stage.
Also increased IPS current to 2mA
1st stage gain is 54dB
2nd stage gain is 26dB (Not a very good distribution but it works for a final gain of 60dB @ 1khz)
Re = 800ohm might be noisy
Gain in 2nd stage can not go higher or coupling caps C12 C21 might become too large.
Below you can see the antiriaa response with and without the zobel.
If we use a DOPA for U3 we can increase headroom by using higher voltage rails for the 2nd stage 🙂
Also increased IPS current to 2mA
1st stage gain is 54dB
2nd stage gain is 26dB (Not a very good distribution but it works for a final gain of 60dB @ 1khz)
Re = 800ohm might be noisy
Gain in 2nd stage can not go higher or coupling caps C12 C21 might become too large.
Below you can see the antiriaa response with and without the zobel.
If we use a DOPA for U3 we can increase headroom by using higher voltage rails for the 2nd stage 🙂
Attachments
Right now I am building it this way passive, so a double RIAA is not necessary.
I think it is better to use circuit 2. An INA instead of the simple circuit 1.
I would raise the input idle to something like 10 mA for lowest voltage noise. I do not think that Re is playing a big roll in noise.
How high is the value of R load in your simulation ?
I think it is better to use circuit 2. An INA instead of the simple circuit 1.
I would raise the input idle to something like 10 mA for lowest voltage noise. I do not think that Re is playing a big roll in noise.
How high is the value of R load in your simulation ?
Attachments
I think this will only work if you use a series R on the opamp outputs. Otherwise the output is flat, no?
Jan
Rload is 5k so to have 2mA idle.
Replacing Rload by 1k puts idle near 10mA. nothing else changes
Last edited:
Here you have the arrangement with the INA.
With this new IPS current and different gain settings, the zobel seems to not be needed.
Here we have 1st stage gain = 42dB for a total gain @ 1khz of 61dB
PS: to avoid current flowing into the cartridge, the NPN bjt should be a dual monolytic item.... maybe a biss transistor.
With this new IPS current and different gain settings, the zobel seems to not be needed.
Here we have 1st stage gain = 42dB for a total gain @ 1khz of 61dB
PS: to avoid current flowing into the cartridge, the NPN bjt should be a dual monolytic item.... maybe a biss transistor.
Andreas, I think U1 and U2 supply that. The feedback loop keeps the bottom of the emitter resistors at the voltage to support the right Q1, Q2 collector current versus the 5V ref.
But wasn't this supposed to be an (anti) RIAA? As I see it, it has a flat freq response.
Jan
For me it is not completely sure that these two transistors will work as a
differential pair with full CMRR if the emitter courrent source is missing.
differential pair with full CMRR if the emitter courrent source is missing.
Jan, this is not an anti RIAA. It is an MC RIAA. You have quite high resistor values in the passive RIAA. Do they affect the noise ? I also think I like the current source version better.