yes it works as servo.
I have bypassed the current mirror, bring up the voltage and offset is more stable.
The module allows for easy zero out.
So it works just fine.
I am not blown by the offset created by this circuit however sound is very, very good.
Regarding the distortion, I am downloading the file thanks for the link.
What I can say is that using this discrete module to replace other servo I had used for my previous desing really outperformed sound.
It is strange as numbers would say it should behave worse.
However sound is just more natural and transparent.
Why don't you try it on your desing and tell me whar you think?
BTW on the paradise I have the custom Teflon caps.
I am running my simulated values.
I haven't double checked the accuracy with the instruments yet, but accordingly to the sound, it sounds pretty close.
It probably needs a little tweaking.
I have bypassed the current mirror, bring up the voltage and offset is more stable.
The module allows for easy zero out.
So it works just fine.
I am not blown by the offset created by this circuit however sound is very, very good.
Regarding the distortion, I am downloading the file thanks for the link.
What I can say is that using this discrete module to replace other servo I had used for my previous desing really outperformed sound.
It is strange as numbers would say it should behave worse.
However sound is just more natural and transparent.
Why don't you try it on your desing and tell me whar you think?
BTW on the paradise I have the custom Teflon caps.
I am running my simulated values.
I haven't double checked the accuracy with the instruments yet, but accordingly to the sound, it sounds pretty close.
It probably needs a little tweaking.
I mean if expected offset once encased is +/-15mV or so, then I guess it is up and running.
Right now I am just fitting my shunt PSU with Teflon caps so adjusting the compensation cap and strategically placing bypass caps.
Like I said residual noise is doesn't look like flat trace at 2mv/div.
would you kming p[osting a screen shoot of the residual noise?
Right now I am just fitting my shunt PSU with Teflon caps so adjusting the compensation cap and strategically placing bypass caps.
Like I said residual noise is doesn't look like flat trace at 2mv/div.
would you kming p[osting a screen shoot of the residual noise?
The Paradise R3 seems to be very interesting and I'm planning in building one. I found enough information to build my own PCB. In the mean time after reading all these postsI compiled a detailed Parts List for North America using the usual suppliers Mouser, Digikey and Newark. It doesn't include the pre-regulator circuitry thought. I'll make it later on...This list may be useful to others. Feel free to point any error I may had done. I still to go through all the schematics to make sure no part is missing.
Thanks for the effort, was hoping for a Mouser break down. Any pro's able to validate this, would like to get parts underway since shipment nears.
Post from 7432
Last edited:
With the input shorted and the scope set to 2mV/dif you should not see a straight line but a trace that is aprox. 2mV wide and looks like noise. This is partly because the vertical resolution of the scope is limited. Also that trace will be offset by the output DC when the scope is set to DC. When you can not center the trace switch the scope to AC. In Heseners building guide you can read how to minimize the input and output offset. A scope does not work like a millivolt meter. To get the real noise output you need a true RMS wide band millivolt. The subjective noise measurement needs also an A type filter.
I absolutely don't have a broadband noise type of residual noise.
The preamp is very quite when connected to the system and residual noise at the scope with 1MHz BW has an amplitude perhaps of 0.5-0.8mV pk-pk and looks like there is thermal noise a little of 50Hz in it.
Anyhow, when I short the input the output noise becomes really strange increases instead of decreasing.
I will post a picture of the noise with both input opened and shorted.
Again the preamplifier is extremely low noise and sound is really wonderful, so thank you Joachimo for sharing this design with us.
Could you please post a link to Heseners building guide?
You can find the latest schematic in post 13 of the thread " Paradise Builders".
There is also now a trimmer at the input where you can optimize DC offset at input and output. You may not know that Stefanoo.
I can not tell why the Paradise is noisy with the input shorted. Actually i never tried it.
Nevertheless you get a low noise result and very good sound in practice.
That is what counts to me.
There is also now a trimmer at the input where you can optimize DC offset at input and output. You may not know that Stefanoo.
I can not tell why the Paradise is noisy with the input shorted. Actually i never tried it.
Nevertheless you get a low noise result and very good sound in practice.
That is what counts to me.
yep I agree with you. I mean noise is really, really low and soundwise (with modification I have made let's not open this can of warm up) is really STUNNING!!!
I will however post a picture when I get a chance of the residual noise with and without input shorted.
Regarding, the offset, I have installed it without the local feedback.
I measured an offset on my dummy modifieded DL-103R (I use this to test new design and when I am confident I plug the ZYX4D on, which it is going to happen tonight) is about 2mV.
The cartridge I think is 40ohm static impedance, while the ZYX is 4ohm which should then bring the offset down to 200uV, am I right?
Anyhow, I have no problem with that offset, however I will install the local feedback for the more expensive ZYX, connect at the input a resistor with the equivalent impedance of the cartridge, power up regulate the trimmer and then substitute it with two fixed resistors and then put the ZYX back on and power everything up.
I measured an offset on my dummy modifieded DL-103R (I use this to test new design and when I am confident I plug the ZYX4D on, which it is going to happen tonight) is about 2mV.
The cartridge I think is 40ohm static impedance, while the ZYX is 4ohm which should then bring the offset down to 200uV, am I right?
Anyhow, I have no problem with that offset, however I will install the local feedback for the more expensive ZYX, connect at the input a resistor with the equivalent impedance of the cartridge, power up regulate the trimmer and then substitute it with two fixed resistors and then put the ZYX back on and power everything up.
not the moment now to talk about this
Let's just wait a little and we can discuss about this eventually.
I don't think you understand what I was saying or I mis-interpret what you are trying to say.
Anyhow, the goal is to cancel out the idle current from positive and negative side.
This only happens if you have 0V at the input (kirchoff's law docet).
for the nature of the BJTs and the circuit, with input changes on impedance and not perfect matching of input stage, you will get different offset values.
So the idea is to match voltage input based on the impedance of your cartridge since it is impossible to get such a perfect match on input stage to guarantee you an input offset below 50uV or so for any input value.
You could even adjust the offset while your cartridge is connected but you will have the following downsides:
1) playing with the circuit while cartridge is connected can lead you to short the BJTs at the input which could cause them to burn (less risky since shunt is adopted here but still a possibility to consider) and then cause a big DC offset at the input thus burning the coil.
2) you will have to connect your meter at the input and my concern with doing that and having a low impedance cartidge, is that if you don't have a dedicated low voltage meter, usually a regular one puts out 3.3V or more, you could easily damage the cartirdge's coil. I did that with the Denon since the impedance is higher but it wasn't probably a smart move.
So your best bet would be to connect the resistance equaivalent to the static impedance do the adjustments and then re-connect everything back.
Hope this helps explaining also for those who will connect the phono to their precious expensive cartridge!!
Please pay attention as this is not a JFETs input stage and depending on your build it can have significant input offset enough to cause an irrevarsable damage.
We have talked about the DC offset procedure before and yes it is correct to make a dummy resistor with the DC value of the cartridge then adjust the trimmer and then connect the cartridge. This thread is long so some will have missed that discussion. I will make sure that this procedure is described in the building instructions.
Yes, Salas the product of U x I is always the same, U goes down and I goes up. That is the reason i included the offset trimmer. In my Paradise build though the U x I product was very low. One way to get good initial offset cancelation is to select the NPN and PNP side not only for hfe but also for Ube. I may have had luck but my transistors matched very well in Ube too.
Yes, Salas the product of U x I is always the same, U goes down and I goes up. That is the reason i included the offset trimmer. In my Paradise build though the U x I product was very low. One way to get good initial offset cancelation is to select the NPN and PNP side not only for hfe but also for Ube. I may have had luck but my transistors matched very well in Ube too.
That's included on page 6, albeit very short. Joachim you may want to check or suggest a better wording