I build the Balanced Zen Line Stage ( https://www.passdiy.com/gallery/preamplifiers/balanced-zen-line-stage ) . The objective is to use the stage as a balanced to single ended IV output stage for the es9018k2m DAC:
Setup:
MiniDSP miniSHARC Kit – >4x es9018k2m modules -> Balanced Zen Line Stages -> PowerAmps.
As the setup shows there will be no attenuation between the output stage and the PowerAmps. Therefor I need the output stage to as quit as possible.
The Problem:
I hear a great deal of hiss or sough from the speakers. I had initially set the gain to 20 dB as the circuit suggests but it yields way too much hiss in my opinion. I want the 20 dB gain (30 dB would be even better
) because I want to drive the DAC in current mode.
Can anybody tell me if this is normal behavior? I was under the impression that the Common Mode Rejection Ratio of the circuit should eliminate hum and hiss. There is no hum btw.
I have tried to connect the balanced outputs of the stage to balanced inputs of a powerAmp, but same result. Too much hiss. The only thing that works, is to lower the gain. If I lower it to 5-6 dB I get satisfactory results. But now the gain is too low, and I will have to operate the DAC in voltage mode to get the output level I want.
PS. The Output stage of the es9018k2m module uses an opamp as differential amplifier and it drives the DAC in voltage mode. It is quiet as can be. I thought I could build the same using discrete components.
Setup:
MiniDSP miniSHARC Kit – >4x es9018k2m modules -> Balanced Zen Line Stages -> PowerAmps.
As the setup shows there will be no attenuation between the output stage and the PowerAmps. Therefor I need the output stage to as quit as possible.
The Problem:
I hear a great deal of hiss or sough from the speakers. I had initially set the gain to 20 dB as the circuit suggests but it yields way too much hiss in my opinion. I want the 20 dB gain (30 dB would be even better
) because I want to drive the DAC in current mode.Can anybody tell me if this is normal behavior? I was under the impression that the Common Mode Rejection Ratio of the circuit should eliminate hum and hiss. There is no hum btw.
I have tried to connect the balanced outputs of the stage to balanced inputs of a powerAmp, but same result. Too much hiss. The only thing that works, is to lower the gain. If I lower it to 5-6 dB I get satisfactory results. But now the gain is too low, and I will have to operate the DAC in voltage mode to get the output level I want.
PS. The Output stage of the es9018k2m module uses an opamp as differential amplifier and it drives the DAC in voltage mode. It is quiet as can be. I thought I could build the same using discrete components.
If I understood you correctly you want to take current output from the es9018 to the input of BALZEN which has input impedance of 100k. That won't work - even 1mA from DAC current output will tend to develop 100V on the preamp's 100k input resistance. That's absurd and I suppose you want something else so please clarify.
For example, try to feed the BALZEN's input from DAC's voltage ouput.
Thank you for you replies and sorry for the lack of info. Here are some pictures and a diagram.
@ Zen Mod
PSU is like the original only I use a 2x55V trafo
Common Ground on all channels. (I came across EasyEDA online circuit design, and I made some PCBs )
Hiss is the same on all 4 single ended outputs (and 2 balanced). No difference.
@ Juma
You are right about input impedance. I use 100 ohms at the inputs.
I can make it work I voltage mode, but I wanted to try current mode and I hoped this circuit could provide the gain I needed and be silent at the same time.
Some extra info.
The 1st time I tested the setup, I accidentally mixed up the positive and negative supply voltages. This of course resulted in some popping caps and blown zener diodes. I have replaced all blown caps, and all 8 transistors, plus the broken zener diodes.
@ Zen Mod
PSU is like the original only I use a 2x55V trafo
Common Ground on all channels. (I came across EasyEDA online circuit design, and I made some PCBs
)Hiss is the same on all 4 single ended outputs (and 2 balanced). No difference.
@ Juma
You are right about input impedance. I use 100 ohms at the inputs.
I can make it work I voltage mode, but I wanted to try current mode and I hoped this circuit could provide the gain I needed and be silent at the same time.
Some extra info.
The 1st time I tested the setup, I accidentally mixed up the positive and negative supply voltages. This of course resulted in some popping caps and blown zener diodes. I have replaced all blown caps, and all 8 transistors, plus the broken zener diodes.
Attachments
Last edited:
@6c4c
I'm sorry, but what is PE? Is it positive and negative supply voltage?
I will try to short the inputs, but I don´t expect this to help. I tried disconnecting the Dac, but left the inputs open. This did not change the hiss.
I have a scope at work, I can borrow for the weekend.
@Zen Mod
1st: The missing C5,C6, C17 and C18. They were input caps for blocking DC. I realized I didn't need them (correct me if I'm wrong) and thus shorted them. That is why I left them out on the diagram.
2nd: The zeners. This might be the culprit. I initially had calculated with a current through the zeners of 3-4 mA. But for some reason the 9.1V zeners gave a voltage of over 10V yielding a total of 71V in stead of 63,7 V. Increasing the current seemed to fix this the problem, hence the 20mA.
Yes the zeners look messy. There are 2 reasons for this. The 1W zeners have fairly thick legs. I did not take this into account when designing the pcb, therefore the holes for the zeners are only just large enough for the thick legs of the zeners. This made the assembly a little difficult and resulted in messy looking zeners. On top of this I had to remove some of them after my initial accident, making it even worse . I think I will order some new ½W zeners.
@radule
R16 and 17 are 221 ohm and you are right that is an error. VERY well spotted. I will fix this.
I did not replace all caps. Only the ones that popped. I will fix this also.
I tried replacing the potmeter with fixed resistors. It did held a little, but nothing major.
Regarding matching. I was under the impression that it wasn't that important. Read from Pass Labs: "It is not essential to match the MOSFETs used in this project, but it doesn't hurt either. I have tested the circuit with matched and random parts with insignificant performance differences."
But it properly wont hurt trying
once again I appreciate everyone's feedback.
I have plenty PCBs and time this weekend. I think I will try a new build and try not to blow it all up this time.
I'm sorry, but what is PE? Is it positive and negative supply voltage?
I will try to short the inputs, but I don´t expect this to help. I tried disconnecting the Dac, but left the inputs open. This did not change the hiss.
I have a scope at work, I can borrow for the weekend.
@Zen Mod
1st: The missing C5,C6, C17 and C18. They were input caps for blocking DC. I realized I didn't need them (correct me if I'm wrong) and thus shorted them. That is why I left them out on the diagram.
2nd: The zeners. This might be the culprit. I initially had calculated with a current through the zeners of 3-4 mA. But for some reason the 9.1V zeners gave a voltage of over 10V yielding a total of 71V in stead of 63,7 V. Increasing the current seemed to fix this the problem, hence the 20mA.
Yes the zeners look messy. There are 2 reasons for this. The 1W zeners have fairly thick legs. I did not take this into account when designing the pcb, therefore the holes for the zeners are only just large enough for the thick legs of the zeners. This made the assembly a little difficult and resulted in messy looking zeners
. On top of this I had to remove some of them after my initial accident, making it even worse . I think I will order some new ½W zeners. @radule
R16 and 17 are 221 ohm and you are right that is an error. VERY well spotted. I will fix this.
I did not replace all caps. Only the ones that popped. I will fix this also.
I tried replacing the potmeter with fixed resistors. It did held a little, but nothing major.
Regarding matching. I was under the impression that it wasn't that important. Read from Pass Labs: "It is not essential to match the MOSFETs used in this project, but it doesn't hurt either. I have tested the circuit with matched and random parts with insignificant performance differences."
But it properly wont hurt trying
once again I appreciate everyone's feedback.
I have plenty PCBs and time this weekend. I think I will try a new build and try not to blow it all up this time.
OK time for an update.
I made a setup of 4x 30V PSU’s to get +-60V. This way I could bypass the power section to eliminate any possible errors there (see pics). Result: No difference. Still an audible hiss.
I then decided to build a new stage. This time I didn’t built the power section to keep it simple. In this build I used new components, so everything should be working correctly (see pics). Result: No difference. Still an audible hiss that increases with the gain.
Working as an acoustical test engineer I know that communicating noise levels can be very difficult. But at 30 dB gain the hiss exceeds the noise from my OLD PC. I tried to measuring the noise levels using the very nice AudioToools App on my IPad (wish they would make it for android, but that’s another discussion). The first picture shows the levels when I run the SE output of the DAC board which has NO hiss. The second picture shows the levels of the Balanced Zen with gain set at 10 dB. Notice the difference from about 500 Hz to 2kHz.
Measuring setup: Icepower 1000asp amp (gain around 27 dB I think?). Beyma 10AG/N speaker standing on the floor. (sensitivity 97 dB, I know this is sensitive and I only used it to prove my point for the measurements. I was using an old book shelve speaker to do the initial testing). IPad at about 20 cm from speaker cone.
Hmm.
I'm having problems uploading my pics. I'll try again later.
I made a setup of 4x 30V PSU’s to get +-60V. This way I could bypass the power section to eliminate any possible errors there (see pics). Result: No difference. Still an audible hiss.
I then decided to build a new stage. This time I didn’t built the power section to keep it simple. In this build I used new components, so everything should be working correctly (see pics). Result: No difference. Still an audible hiss that increases with the gain.
Working as an acoustical test engineer I know that communicating noise levels can be very difficult. But at 30 dB gain the hiss exceeds the noise from my OLD PC. I tried to measuring the noise levels using the very nice AudioToools App on my IPad (wish they would make it for android, but that’s another discussion). The first picture shows the levels when I run the SE output of the DAC board which has NO hiss. The second picture shows the levels of the Balanced Zen with gain set at 10 dB. Notice the difference from about 500 Hz to 2kHz.
Measuring setup: Icepower 1000asp amp (gain around 27 dB I think?). Beyma 10AG/N speaker standing on the floor. (sensitivity 97 dB, I know this is sensitive and I only used it to prove my point for the measurements. I was using an old book shelve speaker to do the initial testing). IPad at about 20 cm from speaker cone.
Hmm.
I'm having problems uploading my pics. I'll try again later.
Last edited:
Well 20 dB was my initial goal, with the dac in current mode (Zin 100 Ohm).
I'm not gonna have a preamp, but I want an output stage for the DAC.
Like I wrote earlier, if I set gain to 5-6 dB the hiss is acceptable, and use the DAC in current mode (Zin 100k ohm) then I get and output level which equals that of the SE output of the DAC board. Only the SE output of the 10USD DAC board is DEAD quit!
I'm not gonna have a preamp, but I want an output stage for the DAC.
Like I wrote earlier, if I set gain to 5-6 dB the hiss is acceptable, and use the DAC in current mode (Zin 100k ohm) then I get and output level which equals that of the SE output of the DAC board. Only the SE output of the 10USD DAC board is DEAD quit!
Last edited:
@ Zen Mod. Just had a glance at the gain article. Looks interesting.
Here is a link for others to use:
http://www.diyaudio.com/forums/diyaudio-com-articles/186018-what-gain-structure.html
Here is a link for others to use:
http://www.diyaudio.com/forums/diyaudio-com-articles/186018-what-gain-structure.html
I build the Balanced Zen Line Stage ( https://www.passdiy.com/gallery/preamplifiers/balanced-zen-line-stage ) . The objective is to use the stage as a balanced to single ended IV output stage for the es9018k2m DAC:
Setup:
MiniDSP miniSHARC Kit – >4x es9018k2m modules -> Balanced Zen Line Stages -> PowerAmps.
As the setup shows there will be no attenuation between the output stage and the PowerAmps. Therefor I need the output stage to as quit as possible.
The Problem:
I hear a great deal of hiss or sough from the speakers. I had initially set the gain to 20 dB as the circuit suggests but it yields way too much hiss in my opinion. I want the 20 dB gain (30 dB would be even better
I have been trying to use a Balanced Zen line stage as a driver for an MOSFET circlotron. The hiss from the line stage is quite pronounced. I'm investigating alternatives at the moment. (I need to convert unbalanced input to balanced output for driving the circlotron. The Balanced Zen in any case needs some modifications to do this well (CCS in MOSFET sources).) But anyway, the hiss is really too loud to be usable.
Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.