diyAudio

diyAudio (https://www.diyaudio.com/forums/index.php)
-   Vendor's Bazaar (https://www.diyaudio.com/forums/vendor-s-bazaar/)
-   -   Output transformers for DACs (https://www.diyaudio.com/forums/vendor-s-bazaar/327794-output-transformers-dacs.html)

bisesik 16th July 2019 06:57 AM

Seems you showed the PSU-part of schematics instead of needed...

lasercut 19th July 2019 08:32 PM

I was spending some using the AK4493 voltage out DAC, I was not a fan of this DAC from DIYinHK when I first listened compared to 9038 and didnt use much since getting it. The chip is simple to implement, so fed with good power I think the you can blame the chip as being mediocre (but still very good for the price).

I did notice this chip can drive headphones directly on its output somewhat well , better than a lot of op amps, and to very high levels. I found I got most enjoyment using it this way compared to any buffer/filters but the novelty wears off fast, its a transparent but weak sound.

But then a thought, use your current transformer backwards!
The DAC gets presented a similar load as before (240ohm?), the headphones get 4 times the current, sacrificing those unnecessarily high voltage levels.
The result? Some of the best sound Ive ever heard!
Why? I really have no idea,
This DAC was giving such mediocre performance in normal circumstances, all the transformer provides is more current to the HPs, why doesnt buffering the output achieve something similar? again I have no idea.

This has little reason to sound so good, but how about applying it elsewhere to create far more ideal conditions.

Like with a high current output DAC, ak4499
A transformer built with the absolute minimum possible resistance for the DAC, and low output impedance for the HPs.
I think this would sound incredible, but of course it only intended to work well with high-z HPs.
Though some of the best and affordable HPs you can buy are high-Z (HD650 and HD800)

bisesik 20th July 2019 04:47 AM

Thank you for your experience Lorcan!
Sure you can use the current transformer this way if the output sound volume is enough in such approach for you.
If to simplify then it works better than Op Amps because of its transparancy\fastness and simplicity of the load seen at the DACs output. Moreover it should works better compared to other types of transformers where the possibility of the core (material) possibility are limited by inner properties, including histeresys problems (relation with speed\transparancy). Try to compare in mind the TENS of ACTIVE components in Op Amp and its PSU with only ONE PASSIVE transfomer :)
Regarding load impedance. There is not 240 Ohm (It is the DC resistance of the copper in secondary coil in this case) presented at the DAC output in real, but from 38kOhm@20Hz to 38MOhm@20kHz because it acts as inductance at AC. Let's take the smallest value of 37kOhm for reactive resistance and 300 Ohm of the loaded impedance for your headphones. The DAC in this case will see (including the reflected secondary load):
37kOhm of the primary in parallel with 300 Ohms multiplied by squared meaning of transformer turns ratio => (37000*300*16)/37000+(300*16)=~4.2kOhm.
The load seen by your DAC in this case is 4.2kOhm. I think AK4493 can provide the full-scale signal at such a load without any problem. Let it be 2Vrms (I don't know exactly just for reference), then there should be something about 0.5Vrms at the secondary if I didn't miss something else...

bisesik 20th July 2019 08:54 AM

Another one thing. According to AK4493' datasheet there should bethe load at least of 450 Ohm resistance between output and GND for DC. Seems it is OK if to use 250 (450/2=225) Ohm between outputs (instead of out to GND) according recommendation's satisfaction.
In this terms the new PELSHO version have even more DCR, and should be better in SQ at least within this parameter itself for use with voltage-source-type DACs (like AK4x9x).

lasercut 20th July 2019 09:49 AM

I think this is too oversimplfied in my mind when applied that to a current DAC., not obeying physics
We want a current DAC to see the lowest impedance possible, while passing all current to the HPs or even stepping down with something like 9038Pro/ak4499 in mono mode (''1:0.5'' version current transformer), its impossible to have both?

bisesik 20th July 2019 10:25 AM

First of DACs are "voltage-type" like AK (except 4499) - they works better to the high impedanced load.
Second of them are "current-type" like ES90x8 and they works better to the low impedance load. As you understand, we can not affect to this situation.
Why there are two different types of the DACs on market? I don't know, it is not my question :) My task is to give solution to both of DAC types: current transformers and voltage transformers

androa76 26th July 2019 07:13 AM

1 Attachment(s)
One example of Bisesik transformers use... with totaly Accu suplied ak4495 dac.

lasercut 26th July 2019 12:46 PM

with typical voltage output transformer (1:1 high impedance) what function does it provide?

bisesik 26th July 2019 04:03 PM

1 Attachment(s)
It is very interesting question. To avoid the usual answer that it acts like decoupling+LPF, I asked for TioFrancotirador to share his experience of using Teramoto transformers (same type of core material) with voltage type DAC. Hope it will be helpful.
One addition. All the next transformers will be made in PCB-style.
Attachment 770742

lasercut 27th July 2019 02:56 AM

There is also the differential to single ended conversion, probably the big one for most but I run balanced so overlooked it.


All times are GMT. The time now is 04:34 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 17.65%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Copyright ©1999-2021 diyAudio

Wiki