diyAudio (
-   Digital Line Level (
-   -   NOS DAC with tube buffer output (

billbo 24th May 2013 03:02 PM

NOS DAC with tube buffer output
1 Attachment(s)
Just though I would share this circuit which I used in my TDA1543 nos dac. I removed the op amps and coupled the DAC into this buffer which uses a 6418 tube driving into a source follower to the output. The circuit uses a little feedback for greater stability and can drive very low impeadances down to 1K ohm without any increase in distortion. Although the response is flat up to 35Khz it seems to sound not as crisp and harsh on the highs and after listening for a while it is almost impossible to revert back to the op amp setup. The mids are highlighted and the highs are well presented but not over exagerated. I have only setup the gain at 2 but it has an open loop gain of only about 6.
It is a very affordable solution if you want to try a tube output on your DAC.
I invite anyone out there for your opinions about this. The DAC is a kit I got a while ago from which runs 8 parralled TDA1543's.
Sorry for the poor schematic quality. I just quickly hand drew it.

Billy D...

PreSapian 24th May 2013 08:14 PM

thanks billy, this might be just what i was looking for ;)

was this inspired by the oatley line stage by any chance?

billbo 25th May 2013 02:56 AM

NOS dac tube buffer output
Oatley is just up the road from me (very convenient to get the tubes) and yes I used the k295 PCB and modified it.
The PCB is of very good quality and it withstood all the soldering and desoldering as I was testing different combinations. The other kits he has need more tweaking as there are certain design flaws which I have already spoke to them about. This is by far much better design with a higher supply voltage used and FET source follower output. Feed back put the icing on the cake and improved it even better. If you need open loop gain the feedback network can be omitted. It worked very well without it aswell.

Billy D...

PreSapian 25th May 2013 05:04 PM

and suppose i wanted to do without the trimpot on the input, what gain am i looking at? (5.5 i suppose? [i'm a total noob at this stuff which's why low voltage supply is such an welcoming thing])

AudioLapDance 25th May 2013 05:15 PM

2 Attachment(s)
Since it's NOS, are you planning to add the sin c correction? (~ + 4db @20k)

Parallel RLC in series with the IV resistor (47 ohms) creates a peak @ 20k

Attachment 350703

Sin C Envelope Correction:

Attachment 350706

FAQ: DDS Tutorial - SINC Envelope Correction | EngineerZone


billbo 26th May 2013 09:24 AM

dac tube buffer
The cct is as simple as it gets. Low voltage and small in size comparison. A cheap way to experiment "tube sound". You can omit the trimpot and also the 100k ohm resistor if you want the preamp/buffer to to have the full gain. There is no problem with that. I included it in so I can reduce the output because there was too much drive. No I am not planning to put any post filtering at all. To raise the highs by 4 dB? All the ccts I have seen on the web have not implemented such a thing for the TDA1543 or the TDA1545. The highs don't seem to be lost but that "CD sound" with "rich" highs has been tamed down which has made them sound more clear and more natural and to me that is more pleasant to the ears.

Billy D...

PreSapian 26th May 2013 05:56 PM

i also particularly don't need the roll-off compensation cuz my pi4 speakers start rolling off at 14k, but the circuit is thorsten's design so it has it's merits.

Torchwood421 26th May 2013 07:31 PM

Hi billbo, Thanks for posting this. I happen to have a few of those tiny tubes in my parts bin. :) Will have to try that out. I'm slow at times getting it done. :snail:

billbo 26th May 2013 10:33 PM

dac buffer
If you get around to make up this cct please send some feedback on this thread . I will be curious to read your opinions.

Billy D...

billbo 29th May 2013 12:41 PM

Circuit revised
1 Attachment(s)
I was asked the question in regards to the biasing of the tube in the thread which made me revisit the design of the cct.. In essence the bias takes the average voltage that is across the filament (approx 0.6 volts). In real tests it showed to be about 0.1 to 0.3 of a volt. I considered this is too low and would pose a problem driving the tube positive. This is not a real problem if you can drive high current from the source but it is not recommended for audio. I did more research and redesigned the cct. to have more grid bias. (about 1 volt). The revised cct. has a 100 Ohm resistor which will give 1 volt bias and raised the supply voltage to 30 volts. This has greatly reduced distortion and has made the cct. more stable.
Once again sorry for the hand draw schematic.

Billy D...

All times are GMT. The time now is 12:31 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio