Hi Ivan,
on the older PCB for Ian'n dac (the one I have) there is option only for Resistor on secondary (R1; R2). On later Version(7.1) there are RIV1;RIV2 (for Primary) and R1; R2 or R1a/b;R2a/b (for secondary) load. On the latest pcb for Ian's dac with 60 mm transformers you state using RIV1/RIV2;R1a/b;R2a/b "default" for passive xlr output. On the pcb version for smaller transformer (V 7.1) for Ian's dac with active option you only list using RIV1/RIV2;R1/R2 (avoid R1a/b;R2a/b) for passive output.
Is there a difference in using small or big transformer in this regard? What should be preferred for xlr passive output? And how do the values relate. If I use R1/R2 1000 ohm what value should be then used for R1a/b;R2a/b?
Sorry for so much questions. I have started to try things out. but I do not have good resistors at hand and have to order some (with high prices) and would like to reduce the values I need.
Regards
Branko
on the older PCB for Ian'n dac (the one I have) there is option only for Resistor on secondary (R1; R2). On later Version(7.1) there are RIV1;RIV2 (for Primary) and R1; R2 or R1a/b;R2a/b (for secondary) load. On the latest pcb for Ian's dac with 60 mm transformers you state using RIV1/RIV2;R1a/b;R2a/b "default" for passive xlr output. On the pcb version for smaller transformer (V 7.1) for Ian's dac with active option you only list using RIV1/RIV2;R1/R2 (avoid R1a/b;R2a/b) for passive output.
Is there a difference in using small or big transformer in this regard? What should be preferred for xlr passive output? And how do the values relate. If I use R1/R2 1000 ohm what value should be then used for R1a/b;R2a/b?
Sorry for so much questions. I have started to try things out. but I do not have good resistors at hand and have to order some (with high prices) and would like to reduce the values I need.
Regards
Branko
As you see my thread is more about my transformers. Thus it was logical thing to concentrate my attention on the trafos itself. At the beginning (in the first revisions of my PCBs) I have simply tried to ignore i/v resistors quality. But today I can declare:
BETTER to use LOW VALUE I/V RESISTORS before transformers, but their QUALITY (in terms of sound for sure) will be PREVAILING. Doesn't matter D46 or D60 trafos.
Regarding splittering the resistor across the secondary coil on to two resistors with CT at GND - it is more about semi-simultaneous (no need to reroute jumpers on PCB at least) usage of both RCA and XLR outputs (but SE will be just half of BAL in this case). I do not use balanced connections in the analog lines in my system, but a single resistor across the secondary is better for SE output. In case of a single resistor across the secondary there is a little difficult with XLR usage - you need to pull off negative end of resistor from the ground, i.e. you need to use some switches/jumpers. Is it better or worse for the next input balanced (XLR) stages to use floating gnd (in case of single resistor) versus fixed gnd (at CT between two resistors) - I have no straight answer - I have no statistics.
BETTER to use LOW VALUE I/V RESISTORS before transformers, but their QUALITY (in terms of sound for sure) will be PREVAILING. Doesn't matter D46 or D60 trafos.
Regarding splittering the resistor across the secondary coil on to two resistors with CT at GND - it is more about semi-simultaneous (no need to reroute jumpers on PCB at least) usage of both RCA and XLR outputs (but SE will be just half of BAL in this case). I do not use balanced connections in the analog lines in my system, but a single resistor across the secondary is better for SE output. In case of a single resistor across the secondary there is a little difficult with XLR usage - you need to pull off negative end of resistor from the ground, i.e. you need to use some switches/jumpers. Is it better or worse for the next input balanced (XLR) stages to use floating gnd (in case of single resistor) versus fixed gnd (at CT between two resistors) - I have no straight answer - I have no statistics.
500R+500R=1000R 🤷♀️
Last edited:
Thank you so much for you patience!
Do you happen to have 3D models of your transformer. I have to build pcb and would like to make components for eagle cad.
thx
Branko
Do you happen to have 3D models of your transformer. I have to build pcb and would like to make components for eagle cad.
thx
Branko
You are welcome! 🙂
I have attached .step 3D model files for both D46 and D60 trafos in the first message
The simplest way to connect my transformers to your circuit without soldering directly to pins (sometime DIYers wants to try different types in quick manner for example) is to use receptacles from Mill-Max:
1-st way is using receptacles with the soldered to them wires. PN is ED90595-ND
2-nd way is using through-hole type of receptacles for PCB (PN is ED10169-ND)
1-st way is using receptacles with the soldered to them wires. PN is ED90595-ND
2-nd way is using through-hole type of receptacles for PCB (PN is ED10169-ND)
Hi bisesik
You posted this information a while back trying to help me with my D3 based DAC project.
I was struggling with what needed to be done to implement your transformers as it is my first DAC project and it had taken up about a year.
I decided to take a break from it and used it with just i/v resistors installed until now.
I would like to return to it and figure out if it is possible to implement your transformers for some additional gain on the output of my DAC.
After going back through previous posts and feedback I was able to understand a little more about my D3 build, but I am also left with some other questions.
If you review your previous feedback that I quoted you’ll see that the transformers I have were originally made for member tubo. They are wound for 0.5 + 0.5:22 as I understand your labeling.
I had mentioned that I was using two D3 boards in a “dual/mono” configuration. This appears to have created some confusion.
You responded with a wiring diagram that assumed what I meant was to try a “differential” arrangement with the two D3 boards.
This is what confused me as I had heard the term “differential” before, but it always seemed to be associated with someone pursuing balanced (XLR) outputs for their project. That was not my goal as I needed RCA outputs.
I’ve come to understand that the “differential” application as related to the D3 is possible with Ryan’s V2 I2S to Simultaneous board (which I have) but requires there to be an inverted signal fed to each D3 board in addition to a normal signal.
I was NOT feeding an inverted signal to the D3 boards. As I understand it, this amounts to what would be considered “parallel” connection.
This brings up a few different questions.
What benefits was I getting from parallel connection of the D3 boards? (More current?)
Does the wiring diagram that you provided for “differential” connection also work with “parallel” connection of the D3 boards to the transformers…or should it be different?
I’ve read the benefits to using the “differential” arrangement as:
- DC Offset is null by the same TDA1541/A chip differential out
- Channel separation is superior
- THD+N and SNR is greatly improved
I have some other questions regarding output impedance and how the DAC will work with my preamps…I’m worried it may be too high and I am not sure how to deal with that.
Ryan had suggested:
…try using an IV resistor on the primaries and leave in place the current injection from the DAC +5V supply.
You will need to remove the resistor on the output of the transformer.
I’m not sure if that relates to the SMD resistors and caps you said should be removed? I’d rather not screw up my boards trying to remove them if that is the case since I don’t have proper equipment for SMD work or replacement components if I need to put them back in place.
Hi chromenuts,
It was a bit scary to answer instantly when I have seen your post at tuesday (too many letters )🙂 Anyway, let me answer today:
2. Matter of taste
3. Matter of implementation/taste
I do not like PP (BAL) outputs at all for example 🤷♀️
Thanks.
Ivan.
It was a bit scary to answer instantly when I have seen your post at tuesday (too many letters )🙂 Anyway, let me answer today:
Same in terms of current (signal) rising, but parallel connection keeps SE output while differential makes it like PP in terms of harmonics.
Different. Parallel connection of D3 boards works similar to single D3 board in terms of the next transformers stage (like here: https://www.diyaudio.com/community/attachments/901396/).
1. Correct
2. Matter of taste
3. Matter of implementation/taste
I do not like PP (BAL) outputs at all for example 🤷♀️
Sure, possible. Transformers' output (secondary coil) can be connected as SE.
Simple story. You should find the needed i/v resistor value across the primary coil which will suits your preamp input capabilities in best manner.
Agree regarding the iv resistor across the primaries, but some amount of resistance (resistor) across the secondary can help for signal to pass through the trafo as well. Matter of your experiments. To leave the DC-cancellation using DC-biasing (using jfet as I remember) - doubtful (primaries DCR of my trafos is really very small for sensible dc-offset impact on it), but again - matter of your experiments, it is up to you.
These caps/resistors relates to the DC-biasing circuit, so you need to remove them if you want to try without "current injection from the DAC +5V supply".
Thanks.
Ivan.
Hi bisesik
Thank you so much for taking the time to answer my questions!
Your response did leave me with a few more questions regarding the parallel connection of my D3 boards and using the transformers in that scenario.
My research and understanding on the parallel scenario was based on some information posted by ecdesgns in the Ultimate NOS TDA1541 thread who stated:
“For parallel mode you can parallel the outputs on each chip and use an I/V resistor with half the usual value and +4mA bias or active I/V converter circuit. Full scale current doubles to -8mA!”
and
“So if we want higher output voltage and a maintain low output impedance we need higher full-scale current (TDA1541A chip outputs in parallel) so we can lower the I/V resistor value.”
I was under the impression that there was a direct correlation between how much current was available at the output of the D3 boards (being doubled in parallel) and the signal voltage level after i/v conversion by the resistor.
It wasn’t clear to me if lowering the i/v resistor value was actually part of the reason why more signal voltage was seen after i/v conversion?
Ecdesigns comments on i/v resistor type and value as well saying:
“a 120R Welwyn W21 wire wound resistor between +240mV and TDA1541A output. With TDA1541A full scale current of 4mA this results in 480mVpp. This is still below the clipping limit of around 600mVpp and distortion stays below the audibility threshold of roughly 1%. It is -very- important to use a low noise wire wound (not bulk metal foil !!) resistor here.”
and
“Low passive I/V resistor values give lowest distortion but poor resolution as most of the LSBs are burried deep below the noise floor.
With 12.5 Ohm passive I/V resistor (50mVpp) LSB is only 0.05 / 65536 = 762 nanovolts. Noise levels of up to 1 millivolts in an I/V stage are quite common. So 6 MSBs will stay above the noise floor while the other 10 hide below it. Suppose one could still hear signals that are below the noise floor one might just be able to squeeze out 8 bits resolution”
So at this point I am wondering if lowering my i/v resistor value (they are currently 110 Ohms) will increase the signal voltage level seen after the i/v resistor conversion?
If so, that would give me a stronger signal and lower output impedance.
I’m assuming I would have to re-adjust the D3 output biasing to null dc offset properly with a new lower i/v resistor value.
The only question that remains is if this still does not give me a satisfactory signal level for my system, and I need to also implement the transformers, do these transformers multiply the output impedance by the same factor when connected in a single ended configuration as they do in the differential configuration(x484)?
Thank you so much for taking the time to answer my questions!
Your response did leave me with a few more questions regarding the parallel connection of my D3 boards and using the transformers in that scenario.
My research and understanding on the parallel scenario was based on some information posted by ecdesgns in the Ultimate NOS TDA1541 thread who stated:
“For parallel mode you can parallel the outputs on each chip and use an I/V resistor with half the usual value and +4mA bias or active I/V converter circuit. Full scale current doubles to -8mA!”
and
“So if we want higher output voltage and a maintain low output impedance we need higher full-scale current (TDA1541A chip outputs in parallel) so we can lower the I/V resistor value.”
I was under the impression that there was a direct correlation between how much current was available at the output of the D3 boards (being doubled in parallel) and the signal voltage level after i/v conversion by the resistor.
It wasn’t clear to me if lowering the i/v resistor value was actually part of the reason why more signal voltage was seen after i/v conversion?
Ecdesigns comments on i/v resistor type and value as well saying:
“a 120R Welwyn W21 wire wound resistor between +240mV and TDA1541A output. With TDA1541A full scale current of 4mA this results in 480mVpp. This is still below the clipping limit of around 600mVpp and distortion stays below the audibility threshold of roughly 1%. It is -very- important to use a low noise wire wound (not bulk metal foil !!) resistor here.”
and
“Low passive I/V resistor values give lowest distortion but poor resolution as most of the LSBs are burried deep below the noise floor.
With 12.5 Ohm passive I/V resistor (50mVpp) LSB is only 0.05 / 65536 = 762 nanovolts. Noise levels of up to 1 millivolts in an I/V stage are quite common. So 6 MSBs will stay above the noise floor while the other 10 hide below it. Suppose one could still hear signals that are below the noise floor one might just be able to squeeze out 8 bits resolution”
So at this point I am wondering if lowering my i/v resistor value (they are currently 110 Ohms) will increase the signal voltage level seen after the i/v resistor conversion?
If so, that would give me a stronger signal and lower output impedance.
I’m assuming I would have to re-adjust the D3 output biasing to null dc offset properly with a new lower i/v resistor value.
The only question that remains is if this still does not give me a satisfactory signal level for my system, and I need to also implement the transformers, do these transformers multiply the output impedance by the same factor when connected in a single ended configuration as they do in the differential configuration(x484)?
Simple story. Lower i/v - lower signal swing.
🙂 What if there is no any noise level in an i/v stage? I/V resistor+Transformer is a passive i/v stage. Sure, there will be PSU's noise later, at preamp/amp block, but it is out of the DAC+i/v stage scope. In fact I do hear the bettering in SQ even between 10R vs 5R in i/v resistor value, so it is up to the particular system/case/listener.
This is absurd. Reducing the current-voltage resistance will lead to a decrease in the output voltage of the signal.
Lowering of i/v resistor will DECREASE the signal voltage level. Yes the signal will be stronger (lower output impedance) but such bettering can remains as just "a virtual bettering". To make it real in most cases means that you should decrease (to lower) the next stage input impedance to allow your signal to be transferred with larger signal current.
First question to be answered. Why the DC-offset should be eliminated? Because this 4/8mADC will be applied to the music (AC) signal output and will interfere the output transistors and its normal working within the AC signal (music) "generating". Second question. How it can be fully eliminated? Simple - you need to short the output to the ground. But where is our AC (music) signal then?.. Shorted to the ground as well. Means - no sense. Third question. How to short the output at DC but to leave some amount of resistance (correct termin is impedance) at AC (music) signal?
And here is the brilliant solution is coming. To use my trafos, which has a very low DC resistance of the primary coils but enough (or not enough - depends from your LF appetit) inductance at LF to allow 1541 to see just a couple of ohms or even part of ohm at DC (DC-offset) and much more impedance at AC (music) signal. So that is why I prefer to avoid active DC-offset cancellation circuit as in case of low-DCR coil, the DC voltage attached/applyed to the DAC output (during current DC-offset convertion on resistance) becomes really low compared to the impedance (at AC) seen by DAC.
yes, the output impedance will be higher after any SUT. Step Up Transformer - means Transformer of impedances which will up-transform your impedance from primary to secondary by factor if the squared meaning of the turns ratio. 110R on a 1:11 trafo will be transformed to 110*11*11=13.3k.
So your output for two D3 boards will be: 0.0028A*110*11=~3.4Vrms with 13.3k of output impedance. So if your next stage is say 50k, then 50k will be reflected to primary as 50000/121=413R.
413R||110R =87R
Then: 0.0028A*87*11= you will get ~2.7Vrms at the input of your amp/preamp.
Thank you again Ivan for taking the time to try and help me.
I guess the real issue here is my lack of understanding with how these 0.5 + 0.5:22 transformers I ended up with were intended to be used.
That, as you say, is dependent on the “system/case/listener”.
Did you give the example of a 1:11 transformer because this is how my transformers would function if they are connected in a single ended manner as you illustrated?
I was using the transformation ratio in your previous statement that they multiplied the impedance by 484 times…but that was the differential connection example.
Also, I thought that running two tda1541 parallel doubled the output current…wouldn’t that mean I had more like .004A coming out of my D3 boards?
I’m trying to follow your math and work backwards from my worst case scenario.
The worst scenario, I think, is my DCB1 based preamp with only 20k of input impedance being presented by its pots.
It seems like the transformers in this situation need to be used with a very low value i/v resistor…but how low can I go and still maintain a sufficient output voltage after the transformers?
I need to understand the transformer ratio and do the math.
The thing I hope I have going for me is the paralleled D3 boards putting out double the current…which allows for me to use a lower i/v resistor value and still maintain a decent voltage level after i/v conversion.
Also, it seems the tda1541 wants to see a low impedance on its output.
My DCB1 preamp was modified with a pair if Jensen jt-123-flpch transformers on its output for an additional gain of about 6dB to help with trying to run my DAC with only a i/v resistor on the output.
I use the DCB1 to drive a variety of my DIY amps…my favorite scenario is using my F2J clones to feed high efficiency fullrange speakers on the top of my open baffle and/or in my back loaded horns.
My main system has an Audible Illusions 2D which drives a pair of Quicksilver Horn Monoblocks feeding my Tannoy System 15 DMTii. There is a 47k input impedance on the preamp and plenty of gain (30dB).
I guess the real issue here is my lack of understanding with how these 0.5 + 0.5:22 transformers I ended up with were intended to be used.
That, as you say, is dependent on the “system/case/listener”.
Did you give the example of a 1:11 transformer because this is how my transformers would function if they are connected in a single ended manner as you illustrated?
I was using the transformation ratio in your previous statement that they multiplied the impedance by 484 times…but that was the differential connection example.
Also, I thought that running two tda1541 parallel doubled the output current…wouldn’t that mean I had more like .004A coming out of my D3 boards?
I’m trying to follow your math and work backwards from my worst case scenario.
The worst scenario, I think, is my DCB1 based preamp with only 20k of input impedance being presented by its pots.
It seems like the transformers in this situation need to be used with a very low value i/v resistor…but how low can I go and still maintain a sufficient output voltage after the transformers?
I need to understand the transformer ratio and do the math.
The thing I hope I have going for me is the paralleled D3 boards putting out double the current…which allows for me to use a lower i/v resistor value and still maintain a decent voltage level after i/v conversion.
Also, it seems the tda1541 wants to see a low impedance on its output.
My DCB1 preamp was modified with a pair if Jensen jt-123-flpch transformers on its output for an additional gain of about 6dB to help with trying to run my DAC with only a i/v resistor on the output.
I use the DCB1 to drive a variety of my DIY amps…my favorite scenario is using my F2J clones to feed high efficiency fullrange speakers on the top of my open baffle and/or in my back loaded horns.
My main system has an Audible Illusions 2D which drives a pair of Quicksilver Horn Monoblocks feeding my Tannoy System 15 DMTii. There is a 47k input impedance on the preamp and plenty of gain (30dB).
I was wrong, of course. Not 1:11 but 1:22 (0.5+0.5=1). 20k of input impedance, ok. But how much signal voltage do you need? What is the sensitivity of your DCB1 in other words?
That is a very good question!
The DCB1 is a DC coupled version of Nelson Pass’ B1 buffer…it has no gain normally.
I added the Jensen transformers to its output in order to add 6dB of gain.
I’m not exactly sure how I can quantify it’s sensitivity.
The DCB1 is currently playing into my F2J monoblock clones.
The First Watt website states that the F2J has 14 dB of gain and produces 5W into an 8 Ohm load.
The speakers I am using are back loaded horns built around Fostex fullrange drivers that produce an SPL of 96dB from 1 Watt.
When I use a digital source directly (such as my MacBook or iPad Pro) the system will play at very satisfying levels and beyond what I need to listen enthusiastically.
My understanding is the techical information regarding the output voltage on my iPad Pro is:
200kΩ: 1.0355 V RMS @ 0.002% THD.
600Ω: 1.0317V RMS @ 0.002% THD.
300Ω: 1.0275 V RMS @ 0.002% THD.
37.5Ω: 0.968 V RMS @ 0.005% THD.
My MacBook appears similar.
The DCB1 is a DC coupled version of Nelson Pass’ B1 buffer…it has no gain normally.
I added the Jensen transformers to its output in order to add 6dB of gain.
I’m not exactly sure how I can quantify it’s sensitivity.
The DCB1 is currently playing into my F2J monoblock clones.
The First Watt website states that the F2J has 14 dB of gain and produces 5W into an 8 Ohm load.
The speakers I am using are back loaded horns built around Fostex fullrange drivers that produce an SPL of 96dB from 1 Watt.
When I use a digital source directly (such as my MacBook or iPad Pro) the system will play at very satisfying levels and beyond what I need to listen enthusiastically.
My understanding is the techical information regarding the output voltage on my iPad Pro is:
Maximum Output Levels
At 0 dBFS at maximum volume setting with various resistive loads:200kΩ: 1.0355 V RMS @ 0.002% THD.
600Ω: 1.0317V RMS @ 0.002% THD.
300Ω: 1.0275 V RMS @ 0.002% THD.
37.5Ω: 0.968 V RMS @ 0.005% THD.
My MacBook appears similar.
I’m going to try and follow your math from the previous post.
I’ll set a goal of 1VRMS output (I may be able to live with less).
I hope I’m assuming correctly that my output current from my paralleled TDA1541 D3 boards doubles to approximately .004A.
With my 0.5 + 0.5:22 transformers the 20K pots of my DCB1 reflect back to the primary and result in 20000/484=41.32R.
Assuming my DAC current output is .004A…then to solve for the i/v system impedance value I need to produce 1VRMS:
.004A*?*22=1VRMS
.088*?=1VRMS
?=1VRMS/.088
?=11.36R
However the 11.36R must be the system impedance seen as a result of the the reflected impedance of my 20k pots off of the primary in parallel with the i/v resistor value I choose?
This means to achieve 1VRMS I need to calculate what value i/v resistor in parallel with 41.33R (pot reflected impedance) will produce 11.36R.
Rounding things off to simpler whole numbers:
42R||?R=12R
I used an online calculator which told me the closest I can get is a 42R||17R=12.10R.
So then with a 17R i/v resistor I should be able to achieve:
.004A*12.1*22=1.06VRMS
The impedance at the secondary of the transformers would be:
17*22*22=8.2k
I’m not sure how that would work well into the DCB1 20k pots.
In the past I’ve tried to maintain a minimum 1:6 ratio regarding impedance output to input between stages to avoid problems.
So I’m then left with the dilemma of whether I can live with a lower voltage output signal in order to lower the impedance seen at the transformer secondary?
I’ll set a goal of 1VRMS output (I may be able to live with less).
I hope I’m assuming correctly that my output current from my paralleled TDA1541 D3 boards doubles to approximately .004A.
With my 0.5 + 0.5:22 transformers the 20K pots of my DCB1 reflect back to the primary and result in 20000/484=41.32R.
Assuming my DAC current output is .004A…then to solve for the i/v system impedance value I need to produce 1VRMS:
.004A*?*22=1VRMS
.088*?=1VRMS
?=1VRMS/.088
?=11.36R
However the 11.36R must be the system impedance seen as a result of the the reflected impedance of my 20k pots off of the primary in parallel with the i/v resistor value I choose?
This means to achieve 1VRMS I need to calculate what value i/v resistor in parallel with 41.33R (pot reflected impedance) will produce 11.36R.
Rounding things off to simpler whole numbers:
42R||?R=12R
I used an online calculator which told me the closest I can get is a 42R||17R=12.10R.
So then with a 17R i/v resistor I should be able to achieve:
.004A*12.1*22=1.06VRMS
The impedance at the secondary of the transformers would be:
17*22*22=8.2k
I’m not sure how that would work well into the DCB1 20k pots.
In the past I’ve tried to maintain a minimum 1:6 ratio regarding impedance output to input between stages to avoid problems.
So I’m then left with the dilemma of whether I can live with a lower voltage output signal in order to lower the impedance seen at the transformer secondary?
Your calculations are correct except of the output current fron 1541. 4mA, yes, but 4mA is a peak value not RMS. So you need to use 4/1.41=2.8mArms.
There will be some losses inside the trafo, so the goal is 1.1Vrms.
DAC should see: 1.1Vrms/22turnsratio/0.0028mArms = 16.2 Ohm
There is already the reflected from the secondary coil 42R yes, so we need:
(42*16.2)/(42-16.2)=680/16.2= ~26 Ohm resistor across the primary coil.
No need to worry about impedances ratio between the stages as transformer is transparent in this terms as in fact you matching 16R with 20k, impedances ratio is more than 1000 times.
Something like that.
🙂
There will be some losses inside the trafo, so the goal is 1.1Vrms.
DAC should see: 1.1Vrms/22turnsratio/0.0028mArms = 16.2 Ohm
There is already the reflected from the secondary coil 42R yes, so we need:
(42*16.2)/(42-16.2)=680/16.2= ~26 Ohm resistor across the primary coil.
No need to worry about impedances ratio between the stages as transformer is transparent in this terms as in fact you matching 16R with 20k, impedances ratio is more than 1000 times.
Something like that.
🙂