DAC output stage

[Freddie]

What do you think about this?
I'm going to use this to drive my Aleph amplifer directly without any preamp. I will probably use two PCM1716 DACs, one in normal and one in inverted mode to have a fully balanced DAC. But perhaps I will use something better like CS43122, I haven't made up my mind yet.

Comments, Suggestions?

 

[tschrama]

The in- and output- buffer will work very well .. but I would choose to discard the cap's... you couls easely make it DC-coupled with vey low offset.... I'm very intersted in the way you will use the volume controll... .. Do you have a remote controll schematic.. ? or any schematic for the controll of the switches.. ? How are you gonna deal with the transient upon switching?

thijs

 

[HarryHaller]

Ah..... the Borbely Jfet buffer. I was getting ready to try that circuit this week with my passive preamp. The circuit has a DC offset since the dynamic current source conected to the negative supply is operating at a diiferent DC Vgs than the follower. You could put the same resistor after the follower and match the FETs for DC coupling.

I don't know about splitting the K389s since the two fets have the same substate. Use two and paralell the two fets in one for the follower in in the other for the current sink. This circuit needs very clean power supplies also. Don't take the caps out, they provide the feed back. I think this would work fine with the DC offset issue fixed.

H.H.

 

[Freddie]

Thanks for the reply.

The relays will be controlled either by a Altera EPM7064 or by a microcontroller (PIC16F84) which counts pulses and senses rotation of an optical encoder. I haven't decided yet. I will buy an Altera EPM7064 next week, If I find it usefull I will use it, otherwise I will go for the PIC16F84. I have never tried the EPM7064 before...

So if I need an attenuation of -48dB only one resistor will be in the signal way. An attenuation of 60dB would mean that two resistors are in the signal way. Max attenuation is 94,5dB (six resistor in the signal way).

As for the transient when switching, I think the relays are fast enough so there won't be any audible click/pop when changing volume setting. I use relays for switching between different sources on my preamp, and there is no audible pop or click.

As for the remote controll, I will use Holtek HT12-D and HT12-E chips to make a simple transmitter/receiver.
More info about the Holtek chip can be found here.
http://www.holtek.com/

Here's a schematic for a receiver/transmitter using the above mentioned ICs.
http://www.rentron.com/project03.htm

I'm a little unsure about the the crossover frequency I should use for filtering. The filter in the schematic is a Bessel 60kHz 12dB/octave. Perhaps I should use a lower crossover frequency?

/Freddie

 

[Freddie]

HarryHaller,

Yes it's the Borbely Jfet buffer

And the volume controlled is actually based on the one LcAudio used in there Sidewinder preamp.

I'm thinking of using a normal Jfet buffer with a constant current source, something like figure 15c in the article written by Borbely as this will have close to DC offset if the Jfet's are matched.
But I haven't made up my mind yet.

/Freddie

 

[HarryHaller]

Borbely FET follower

Use the Borbely circuit. Put the same value resistor after the the follower fet for DC offset The current sink fet should have the same Idss as the folower. Match them. The resistor is not in the Borbely circuit since it is AC coupled and Erno was trying to save money on those expesive resistors IMHO. Put a 1 Meg resistor to ground from the gate of the follower fets. DC coupled circuits really don't like floating inputs. Your woofers will thank me. 1000 ohm resistors as gate stopers will help the sound. Scott Nixon of Anodyne fame taught me how to voice tube circuits by changing the VALUE of the grid stopper resistors in tube circuits. RF stability and bandwidth limiting are very important to get audio circuit to sound right. Look for RC fiters on inputs and zoble networks on outputs to make a come back. They were there on the old stuff and are even a better idea today with all the RF pollution.

H.H.

 

[Freddie]

Borbely follower

HarryHaller,

Does it look OK?
I simulated it in CircuitMaker, and it showed a DC offset of about 1mV.

/Freddie

 

[HarryHaller]

Borbely buffer

Put a 1K in series with the gate on the lower fet also. The actual
offset will depend on how close the Saturated drain current is for the two fets. When you build it or layout the board try to keep the lead from the 1K resistor as short as you can make it.

H.H.

 

[dorkus]

how does it sound

hi freddy,

did you build your circuit? any impressions on sound quality?

cheers,
dorkus

 

[LBHajdu]

reed relays

Going back to the idea of using relays for the volume attenuator, has anybody ever tried using reed relays for this? I would like to try making a preamp with reed relays however I am worried about contact bounsc and mercury reed relays are too pricey for the number of them that I would need.

As for the output section, I am curious to know if a passives preamp could be used with this Dac. Does anybody know what the voltage swing of this circuit is going to be at the output?

Please keep us informed of your progress I would like to know how this one works out.

 

[jan.didden]

Dac output stage

LBHajdu,

I have had good results using Pickering (UK) reed relays in my preamps. Their 100 series has versions that can be driven directly from low power schottky logic at 5 V. The ones I used have Ruthenium sputtered contacts and are specifically developed for automated test equipment, which means for reliable very low level switching. I think I paid about 3 $ US for them in small quantities. They do credit card sales.

Cheers, Jan Didden

 

[promitheus]

I am not so sure if the loa pass filter in the input can also work as an I/V converter for the DAC.

 

[Nicke]

Isn´t the PCM1716 a delta-sigma DAC that does not need a I-V converter ?
In that case it should work.

And Freddie:
Have you built it yet and how does it sound ?

Nicke

 

[promitheus]

I don´t know about that. I was looking for something for my PCM1704 which needs an I/V converter.
Could a buffer be used for this reason ?

 

[Nicke]

Perhaps you could try the output stage from the Aleph DAC D1 ?
Or search the digital forum for more threads.
Here is one from the digital forum:
I-V converter

Nicke

 

[promitheus]

The problem with the Aleph DAC is that it has balanced input and has a volume.
I would use it if it was possible to use a non balanced input with no volume control.
Do you have an idea how this would be applicated ?

 

[mikek]

achtung!

this source-follower configuration is not i fear, due to borbely....it goes back to the tube era, when push-pull action could only be induced in this fashion as there are no complementary tubes....indeed it was used in all-BJT form by douglas self in his 1976 pre-amp. article in electronics world.

 

[jwb]

The Pass Labs DAC has both balanced and single-ended digital inputs. They are switched with relays. Study the repair manual.

Even without that, you can ground the inverted input of the digital input receiver for the same effect.

If you don't like the volume control, insert a fixed resistor at your desired output amplitude.

 

[dorkus]

mikek, read my reply in the JFET buffer thread. borbely does not make any claims to inventing the topology, he gives credit to the developer of the original tube circuit.

 

[promitheus]

What happens in single ended mode ? One DAC is not working at all or is there a coverter to balanced ?
I must check again the schematic.
This resistor I think sets the gain of the circuit. It would be interesting to know at which gain the distortion is at lowest like in the ZEN balanced preamplifer.