Dr. ODD said:
Transformers can work quite well in that regard. If you use output transformers though you have to be a bit careful what you drive them from. If the source's output is capacitively coupled, you need to make sure the capacitance is sufficiently high to prevent low frequency resonance peaking due to the relatively low primary inductance of most output transformers. So instead of the more typical 2-10uF coupling cap, you may need to go to 200-300uF.
My preference has been to keep the electronics single-ended and use a good input transformer. Nice thing about such transformers is that they give excellent common-mode rejection even when fed from unbalanced sources. Plus their primary inductance is far higher than typical output transformers so low frequency peaking isn't a problem.
se
Single ended, like BOZOS? The problem with most DIYers, like me here, is that we cannot buy any good transformer (Jensen, Tango, etc). Even if there is any, the price will be very high.
Steve Eddy, Can it still be made by transistors or mosfets but in high quality circuit? What will be the best possible unbalanced to balanced circuit according to your design?
lumanauw said:
You mean BoSoZ? Bride of Son of Zen is balanced. The Bride of Zen is the unbalanced pre.
Yeah. And they're not exactly cheap here in the US either.
Well, you might want to ask yourself do you REALLY want balanced outputs on your sources? The whole Zen aesthetic is about simplicity after all.
An amp like Son of Zen isn't balanced just so it can accept input from a balanced source. It's balanced topology is actually a rather clever way of ultimately making it simpler than the original Zen amp. It has fewer parts and eliminates the capacitive coupling of the original Zen.
To take an existing unbalanced output and make it more complex just so you can have a balanced output seems rather anti-Zen.
If you REALLY want balanced outputs and transformers aren't an option, you can still achieve it without having to resort to active circuitry. You can passively mirror the output impedance of your unbalanced source and achieve a true balanced source.
For example, if your output looks something like this:
An externally hosted image should be here but it was not working when we last tested it.
You can mirror the resistance values and create a balanced output thusly:
An externally hosted image should be here but it was not working when we last tested it.
se
Hello
I also want an unbalanced to balanced but also that I can use it for balanced to balanced.
I have a few problems that maybe can be solved.
I build a couple of proto's with some problems.
If I use the IRF610 and like to use a 10k potentiometer the bandwidth is to small du to the input capacitane.
Second problem are the outputs are not equal in output only if I remove the 150 Ohm resistor but then I have to much gain and on the other proto I only get the output even when I short the - input to ground and the of coarse the gain change.
Rob
I also want an unbalanced to balanced but also that I can use it for balanced to balanced.
I have a few problems that maybe can be solved.
I build a couple of proto's with some problems.
If I use the IRF610 and like to use a 10k potentiometer the bandwidth is to small du to the input capacitane.
Second problem are the outputs are not equal in output only if I remove the 150 Ohm resistor but then I have to much gain and on the other proto I only get the output even when I short the - input to ground and the of coarse the gain change.
Rob
Attachments
Hi
Ok I build it without the current source and use 2x 1K5 parallel and +/- 50V.
Gain is a bit low 3.65 Bandwidht -3dB 235 Khz difference between + and - output is 1.45dB.
Bandwidht is ok difference in output not.
Here the same deal the output is only the same if you short the gate of the - input.
Rob
Ok I build it without the current source and use 2x 1K5 parallel and +/- 50V.
Gain is a bit low 3.65 Bandwidht -3dB 235 Khz difference between + and - output is 1.45dB.
Bandwidht is ok difference in output not.
Here the same deal the output is only the same if you short the gate of the - input.
Rob
Hi Rob
Your upper circiut is just like the BOSOZ without output caps.
If you just short the 150Ohm resistor in your circiut, you simply get too much gain.
The circiut i just posted has some feedback resistors (39K) and some inputresistors (10K) to achieve the X. This arrangement eat som of the gain, therefore there is no resistance between the sources of the diffpair to reach the original gainlevel.
I made the scematic in the SIMetrix simulator, and yes I have simulated this circiut.
SIMetris is a verry easy promgram to use, and has verry few restrictions in the demomodel. You can download a version at: http://www.catena.uk.com/Pages/download.html
If you need some models, let me know, then I can email you some spicemodels of the most common parts used on these sites.
I have attached the sirciut I just posted, it's easy to simulate when you have install the pogram an the spicemodels, just hit the F9 when the scematic is open.
Regards Henrik
Your upper circiut is just like the BOSOZ without output caps.
If you just short the 150Ohm resistor in your circiut, you simply get too much gain.
The circiut i just posted has some feedback resistors (39K) and some inputresistors (10K) to achieve the X. This arrangement eat som of the gain, therefore there is no resistance between the sources of the diffpair to reach the original gainlevel.
I made the scematic in the SIMetrix simulator, and yes I have simulated this circiut.
SIMetris is a verry easy promgram to use, and has verry few restrictions in the demomodel. You can download a version at: http://www.catena.uk.com/Pages/download.html
If you need some models, let me know, then I can email you some spicemodels of the most common parts used on these sites.
I have attached the sirciut I just posted, it's easy to simulate when you have install the pogram an the spicemodels, just hit the F9 when the scematic is open.
Regards Henrik
Hi Rob
I have build a XBSOZ without currentsink (source), and i have done some tests with and without, the difference is verry subtle, and I think i like the resistorsourced the most - but hard to tell really.
When you drive the resistorsourced circiut unbalanced, you wount get quite the same amplitude on both sides, the undriven side will be approx. 5% lower than the driven as far as I remember, this has caused a great deal of concern.
I have build a XBSOZ without currentsink (source), and i have done some tests with and without, the difference is verry subtle, and I think i like the resistorsourced the most - but hard to tell really.
When you drive the resistorsourced circiut unbalanced, you wount get quite the same amplitude on both sides, the undriven side will be approx. 5% lower than the driven as far as I remember, this has caused a great deal of concern.
Hi, Henrik,
Nice design you have. It uses the SOX principle (the feedback resistor)
Want to ask you a few things.
1. What is the current in the CCS (What voltage is LM329)?
2. Mr. Pass has said that to make unbalance to balance converter, CCS is a must, cannot use Resistor load. Is in your simulator, both the + output and -output has the same voltage (compared to ground)? Nothing is bigger or smaller than the other?
3. You uses unbalance PS (+40/-20V). Could you design with balance PS (maybe like +/-30V) and what resistor value should be change? the 750ohm?
If the distortion is not very different I would like to use balance PS. But if the figures are very different, I will stick to your original design.
Nice design you have. It uses the SOX principle (the feedback resistor)
Want to ask you a few things.
1. What is the current in the CCS (What voltage is LM329)?
2. Mr. Pass has said that to make unbalance to balance converter, CCS is a must, cannot use Resistor load. Is in your simulator, both the + output and -output has the same voltage (compared to ground)? Nothing is bigger or smaller than the other?
3. You uses unbalance PS (+40/-20V). Could you design with balance PS (maybe like +/-30V) and what resistor value should be change? the 750ohm?
If the distortion is not very different I would like to use balance PS. But if the figures are very different, I will stick to your original design.
Hi lumanauw
1) 70mA, and LM329 is 6.9V.
2) In the original paper to the balanced Zen Linestage (BOSOZ) Nelson stated, that this circiut could be used also as a unbalanced to balanced converter. But the current sinks makes an almost 100% diffrential match between te outputs when driven unbalanced, so technically speaking it's better than the about 95% match you get with resistors at the sources in an X'ed circiut. I gues that's why Nelson reccomend the use of current sinks instead of resistors.
3) You can use any rail voltage you like, as long as you adjust the emitter resistor in the current sink to draw about 70 mA, and more than 60V rails might be overkill. The currentsink just needs to have about 5V or so across the Collector and the Emitter of the BJT.
Regards
1) 70mA, and LM329 is 6.9V.
2) In the original paper to the balanced Zen Linestage (BOSOZ) Nelson stated, that this circiut could be used also as a unbalanced to balanced converter. But the current sinks makes an almost 100% diffrential match between te outputs when driven unbalanced, so technically speaking it's better than the about 95% match you get with resistors at the sources in an X'ed circiut. I gues that's why Nelson reccomend the use of current sinks instead of resistors.
3) You can use any rail voltage you like, as long as you adjust the emitter resistor in the current sink to draw about 70 mA, and more than 60V rails might be overkill. The currentsink just needs to have about 5V or so across the Collector and the Emitter of the BJT.
Regards
Hello Henrik
I did some simulation and build it.
If you replace the source resistors with a currentsource the positive and negative outputs are the same problem solved.
Thanks for the program link and I am intrested in the models if you can mail me those ill be even more thankfull.
Regards
Rob
I did some simulation and build it.
If you replace the source resistors with a currentsource the positive and negative outputs are the same problem solved.
Thanks for the program link and I am intrested in the models if you can mail me those ill be even more thankfull.
Regards
Rob
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.