Hi Guys,
I was wondering if you could share your experience and/or opinions regarding I/V conversion of a current DAC (TDA1541, AD1865 etc) with a transformer like Sowter 1465 or 1495 or similar good quality transformers. Such transformers are quite expensive and I'd like to make an informed decision based on your experience.
My current setup is a TDA1541 DAC in NOS mode with a passive I/V resistor of 50ohm. This feeds the grid of the Aikido preamp with 6N2P as first stage followed by 6N1P as second stage. Quite a bit of gain but I like the 6N2P sound.
Overall I'm very pleased with the sound however I do suffer of FOMO and I'd like to try a transformer coupled with a resistor. This will cost about US $400 for 2 of them which is expensive.
I'm planning to add a secondary TDA1541 in parallel. A single TDA has a 4mA p-p. Twin will be 8mA p-p. Based on sowter calculation, using the 1465 transformer and 8.2kohm on the secondary winding will generate 2V rms. This should be enough to skip the Aikido altogether.
Based on your experience, it it worth the money to go this route? Did it have an improvement on your current out DAC going from passive I/V resistor only to transformer + resistor? I'm not currently entertaining an active I/V conversion. I'd like to keep the signal path as simple as possible.
Thanks
I was wondering if you could share your experience and/or opinions regarding I/V conversion of a current DAC (TDA1541, AD1865 etc) with a transformer like Sowter 1465 or 1495 or similar good quality transformers. Such transformers are quite expensive and I'd like to make an informed decision based on your experience.
My current setup is a TDA1541 DAC in NOS mode with a passive I/V resistor of 50ohm. This feeds the grid of the Aikido preamp with 6N2P as first stage followed by 6N1P as second stage. Quite a bit of gain but I like the 6N2P sound.
Overall I'm very pleased with the sound however I do suffer of FOMO and I'd like to try a transformer coupled with a resistor. This will cost about US $400 for 2 of them which is expensive.
I'm planning to add a secondary TDA1541 in parallel. A single TDA has a 4mA p-p. Twin will be 8mA p-p. Based on sowter calculation, using the 1465 transformer and 8.2kohm on the secondary winding will generate 2V rms. This should be enough to skip the Aikido altogether.
Based on your experience, it it worth the money to go this route? Did it have an improvement on your current out DAC going from passive I/V resistor only to transformer + resistor? I'm not currently entertaining an active I/V conversion. I'd like to keep the signal path as simple as possible.
Thanks
Attachments
I'd have thought mitigate the risk by having a play with some much cheaper step-up trafos first. For example, these - https://www.aliexpress.com/item/1005004949037079.html
The voltage ratio is around 1:9
The voltage ratio is around 1:9
Wise advice Richard. I tried some of those on one of Ian Canada's dacs. My brother still listens to them. Ian ended up doing a PCB for them as he was pretty impressed with them compared to the more expensive ones he initially designed for...can't remember the brand now....edit Lundahl maybe
I just went through this process recently.
I have a DAC built around Ryan’s D3 boards found here:
https://www.diyaudio.com/community/threads/diy-tda1541a-pcb-d3.328060/
I got it up and running a few years ago and left it with only IV resistors on the output.
It worked because I use highly sensitive speakers and my system had just enough gain…but lacked the ability to handle certain types of music with the desired impact.
I found a pair of barely used IV transformers made by bisesik in the swap meet as part of a bundle sale.
They sat around for a while until I recently got frustrated enough to open the DAC back up and looked for the help I needed to install them properly.
I couldn’t have done it without bisesik’s help…he was great.
The end result is phenomenal. I am very happy.
For reference, I am running two of Ryan’s D3 boards with Taiwan made TDA1541A.
They were originally run parallel for the higher current swing into mk132 IV resistors alone.
When I installed the transformers, I changed the IV resistors out to some hand wound Constantan resistors I made.
I also changed the signal delivery over to differential arrangement.
I’m using some of Ian Canada’s original boards for SPDIF input and the Fifo with upgraded crystals for the reclocking.
I use Ryan’s I2S to Simultaneous board to deliver the signals to the D3 boards.
Most of the info of the process is toward the end of the D3 thread and bisesik’s transformer thread
here:
https://www.diyaudio.com/community/threads/output-transformers-for-dacs.327794/page-44
I have a DAC built around Ryan’s D3 boards found here:
https://www.diyaudio.com/community/threads/diy-tda1541a-pcb-d3.328060/
I got it up and running a few years ago and left it with only IV resistors on the output.
It worked because I use highly sensitive speakers and my system had just enough gain…but lacked the ability to handle certain types of music with the desired impact.
I found a pair of barely used IV transformers made by bisesik in the swap meet as part of a bundle sale.
They sat around for a while until I recently got frustrated enough to open the DAC back up and looked for the help I needed to install them properly.
I couldn’t have done it without bisesik’s help…he was great.
The end result is phenomenal. I am very happy.
For reference, I am running two of Ryan’s D3 boards with Taiwan made TDA1541A.
They were originally run parallel for the higher current swing into mk132 IV resistors alone.
When I installed the transformers, I changed the IV resistors out to some hand wound Constantan resistors I made.
I also changed the signal delivery over to differential arrangement.
I’m using some of Ian Canada’s original boards for SPDIF input and the Fifo with upgraded crystals for the reclocking.
I use Ryan’s I2S to Simultaneous board to deliver the signals to the D3 boards.
Most of the info of the process is toward the end of the D3 thread and bisesik’s transformer thread
here:
https://www.diyaudio.com/community/threads/output-transformers-for-dacs.327794/page-44
Attachments
Thanks for the information. It sounds like trafos do improve the overall sounf quality.
Abraxalito, I see manufacturers specifying that there is a difference between current and voltage transformers. Do you know in which category the trafos you specified belong to as the info doesn't specify.
Abraxalito, I see manufacturers specifying that there is a difference between current and voltage transformers. Do you know in which category the trafos you specified belong to as the info doesn't specify.
Good afternoon Dromichet,
EXACTLTY my question a couple of years ago.
I got a very fine functioning (DIY) NOS-TDA1541A DAC with 3 output options:
Please read :
https://www.diyaudio.com/community/threads/dac-with-no-output-stage.213606/ which has a horrible title as it has an output stage....
The 2 biggest BUTs are:
Silvercore and Sowter are serious tranny-suppliers (and of course there will be others but i have no experience with them).
Fantastic DAC. Currently i use a PCM63K which is on par with the TDA1541a. It's the design that will determine the grand outcome; both laddder-DAC's perfectly capable of portraying MUSIC.
If you have the money available please choose the OPT/IT route; not the most easiest but certainly the most musical. And NO long interlinks between DAC and pre-amp; .....that kills everything as you're pushing your DAC-chip to do something it was not designed to do.
Please forgive my awkward english.
Regards,
Reinout
EXACTLTY my question a couple of years ago.
I got a very fine functioning (DIY) NOS-TDA1541A DAC with 3 output options:
- an opamp (AD844.....this serious compomised details....)
- a tube-version (6DJ8 or 6C45; made them both and are ok....but not great)
- transformer with i/v-resistor. A no-brainer after listening. This is the way to go BUT.....
Please read :
https://www.diyaudio.com/community/threads/dac-with-no-output-stage.213606/ which has a horrible title as it has an output stage....
The 2 biggest BUTs are:
- the trannies has to be of serious quality;
- the length from the DAC to the pre-amp must be taken into account (read: short !). So no strange long cables as the DAC-chip itself is the "powering force" which on its own is hilarious low -powered. If that DAC is nearby no problems should arise. Same accounts for strange resistance at the pre-amp; please make that transformer-output-DAC (in essence a DAC-schip with tranny-converter !) to see an easy load.
Silvercore and Sowter are serious tranny-suppliers (and of course there will be others but i have no experience with them).
Fantastic DAC. Currently i use a PCM63K which is on par with the TDA1541a. It's the design that will determine the grand outcome; both laddder-DAC's perfectly capable of portraying MUSIC.
If you have the money available please choose the OPT/IT route; not the most easiest but certainly the most musical. And NO long interlinks between DAC and pre-amp; .....that kills everything as you're pushing your DAC-chip to do something it was not designed to do.
Please forgive my awkward english.
Regards,
Reinout
What is the best position of the I/V resistor: right at the output of the DAC (at the primary of the transformer) OR at the secondary of the transformer OR distributed at the primary and at the secondary?
It depends on the model it seems. For sowter 1465 and 1495, which are advertised as current trafos, they recommend on the secondary and they also provide calculations for the resistor size. Other models and manufacturers have different recommendations.
I guess, if the resistor is placed in the primary, the conversion happens at that stage with the trafo stepping up the resultant voltage created across the resistor on the primary. If it's placed on the secondary, the conversion happens there. What benefits there are on one over the other I have no idea, maybe it depends on the type of the trafo. Not sure about these statements though, that's what I'm trying to understand.
I guess, if the resistor is placed in the primary, the conversion happens at that stage with the trafo stepping up the resultant voltage created across the resistor on the primary. If it's placed on the secondary, the conversion happens there. What benefits there are on one over the other I have no idea, maybe it depends on the type of the trafo. Not sure about these statements though, that's what I'm trying to understand.
Rainout, thanks for the detailed info. My connection between the DAC and pre-amp is about 2 feet using decent cable. Would that be too long?
I've not met that distinction before - is that something Sowter is saying? I'll go take a look at their blurb and see if I might learn something.
<later> Just read through Sowter's FAQ, couldn't find anything about current vs voltage trafos. Got a link elsewhere?
Last edited:
I used to take the view that its better on the secondary side but these days I'm not so sure either way. The argument for a secondary side R is that the trafo's leakage inductance supplies some useful lowpass filtering, however leakage inductance isn't generally well defined so its tricky to arrange it to hit a specific cut-off frequency. One reason for caution in going for a secondary side R is the compliance at the DAC's output to a large extent depends on the leakage inductance so its tricky to predict whether the output compliance voltage will be exceeded.
By any chance did 'voltage' vs 'current' come from this page : https://www.sowter.co.uk/dacs.php ? If so then its a distinction in the types of DACs that trafos are suited to, not different types of trafos. The trafos suited to voltage output DACs tend to be 1:1 type rather than the step-up ones well suited to current output DACs.
You are right, I assumed they're specific ones for each application based on this line.
"Analog voltage out DAC's cannot be used with 1465 or 1495."
"Analog voltage out DAC's cannot be used with 1465 or 1495."
I've done just that but have a pair of 1541s in balanced configuration using Ian Canada's I2S to PCM interface that produces bit perfect + - R & L data.
I use the 1465 transformers but also, a pair of low noise 100R resistors, and resistors across the secondary to reduce the primary impedance.
I am pleased with the results and the output is high enough to not need any buffer or gain stage. I was surprised just how clean the audio was thanks to the cancellation in the transformers of common mode nasties.
The output offset current of the 1541s can also be ignored because the transformers are bifilar wound so with two dacs, the magnetising effect cancels out.
I would recommend that you first try active I/V conversion with AD811,AD844 or OPA861 or some discrete I/V stage.
You can also do a tube I/V conversion, but with a grounded grid.Tubes for this must have a higher gain and must be able to work with a very small cathode resistor.
A transformer, or in your case a passive I/V conversion with a resistor is not an ideal partner for a current DAC due to the high Zin no matter how much some claim otherwise . The ideal Zin would be 0ohm, which can only be achieved by active conversion. The sound of passive I/V is always a bit withdrawn and even blurred in some parts of the audio spectrum unlike active I/V which is more dynamic, cleaner and more precise. However, someone likes passive I/V, but does not use the current DAC to its full potential.
I thought the DC resistance of the primary coil (6.7 Ohm for Sowter 1465) and the reflected DC resistance of the secondary should be taken into account.
At DC of course the DAC sees a very low resistance.
With AC, even with the reflected secondary load, the IV resistor swamps this.
The main reason for not using an I/V resistor with the 1541 is the 2ma zero signal current offset. This should be nulled if using a single ended dac with any transformers ut not relevant to my balanced design or other bifilar wound transformers. Most 1541 critics concede that the value of the i/v resistor doesn't result in audible distortion provided it is below 100r.
As far as generalised negative comments about sound quality are concerned, they are not true in my case. I have a R2R discrete logic dac from member Chiurtu and can swap quickly from one to the other. (also have a Dacmagic and Cambridge 651 (dual WM8740 and dual CC dacs and the differences are quite subtle)
The R2R dac (opamp output) has a slightly more sharp edged treble compared to the more restrained 1541 dac but neither suffer from lack of detail nor "blurriness". In the end it's which you prefer - just don't let the naysayers put you off.
With AC, even with the reflected secondary load, the IV resistor swamps this.
The main reason for not using an I/V resistor with the 1541 is the 2ma zero signal current offset. This should be nulled if using a single ended dac with any transformers ut not relevant to my balanced design or other bifilar wound transformers. Most 1541 critics concede that the value of the i/v resistor doesn't result in audible distortion provided it is below 100r.
As far as generalised negative comments about sound quality are concerned, they are not true in my case. I have a R2R discrete logic dac from member Chiurtu and can swap quickly from one to the other. (also have a Dacmagic and Cambridge 651 (dual WM8740 and dual CC dacs and the differences are quite subtle)
The R2R dac (opamp output) has a slightly more sharp edged treble compared to the more restrained 1541 dac but neither suffer from lack of detail nor "blurriness". In the end it's which you prefer - just don't let the naysayers put you off.
I deliberately did not mention ordinary op amps but CFA op amps which give a completely different sound with current DACs.
The difference in the price of two CFA op amps and two quality transformers is large so it is more logical to try the cheaper option first. A cheap option can sometimes surprise you with the result.
I have had the NOS TDA1541 for years and I must admit that the passive I/V with the tubes behind them really sound great. In the meantime, I made a much better I/V conversion with tubes where the signal is fed to the cathode (ground grid), which I tested on the PCM1702.
The sound of such a conversion is again completely different from the classic passive I/V with tubes and is much more dynamic and cleaner, a completely different story. It may be a more expensive option than a transformer and certainly much more complex, but the result justifies the investment.
Today, after more than 15 years of tinkering with various I/V conversions, I returned to the CFA op amp the AD811 which Walt Jung first introduced as an I/V stage back in 1992 precisely for the TDA1541. From its output I directly drive the next CFA amplifier Le Monstre. I can hardly compare that combination of I/V conversion and amplifier to anything I've heard so far.
However, everything is a matter of taste including the choice of I/V level and I only gave my opinion.
For comparison, one image shows the tube I/V stage (ground grid with 6HM5) and the other AD811 I/V in the same DAC.
The difference in the price of two CFA op amps and two quality transformers is large so it is more logical to try the cheaper option first. A cheap option can sometimes surprise you with the result.
I have had the NOS TDA1541 for years and I must admit that the passive I/V with the tubes behind them really sound great. In the meantime, I made a much better I/V conversion with tubes where the signal is fed to the cathode (ground grid), which I tested on the PCM1702.
The sound of such a conversion is again completely different from the classic passive I/V with tubes and is much more dynamic and cleaner, a completely different story. It may be a more expensive option than a transformer and certainly much more complex, but the result justifies the investment.
Today, after more than 15 years of tinkering with various I/V conversions, I returned to the CFA op amp the AD811 which Walt Jung first introduced as an I/V stage back in 1992 precisely for the TDA1541. From its output I directly drive the next CFA amplifier Le Monstre. I can hardly compare that combination of I/V conversion and amplifier to anything I've heard so far.
However, everything is a matter of taste including the choice of I/V level and I only gave my opinion.
For comparison, one image shows the tube I/V stage (ground grid with 6HM5) and the other AD811 I/V in the same DAC.
Attachments
Just an update..... I haven't purchased any transformer yet to try but I did buy a board to add a secondary 1541 in parallel using Nichicon 200nF and 670pF.
A big improvement in sound. It extended the high end and made the sound crispier enlarging the soundstage. Overall I'm very happy with the sound.
I haven't given up on the idea of alternatives to passive resistor I/V yet but will stay like that for now. Eventually I'll get bored of the sound and will want to tinker again.
Thanks again for all your input.
A big improvement in sound. It extended the high end and made the sound crispier enlarging the soundstage. Overall I'm very happy with the sound.
I haven't given up on the idea of alternatives to passive resistor I/V yet but will stay like that for now. Eventually I'll get bored of the sound and will want to tinker again.
Thanks again for all your input.