Hi All,
I have a DAC that based on AD1862 dac chip, currently it uses opamps (1 for each channel) for I\V and LPF in it's output.
I'm looking to buy a kit (assembled or not) to replace the opamps, after a long search in this forum and others I couldn't find this kind of "ready made" kit.
I've seen all the I\V stage schematics in this forum but I have a very limited access to components in my area of living so building one from scratch is almost impossible for me.
I'll appreciate your help finding this kind of a kit to buy online.
Cheers,
Guy
I have a DAC that based on AD1862 dac chip, currently it uses opamps (1 for each channel) for I\V and LPF in it's output.
I'm looking to buy a kit (assembled or not) to replace the opamps, after a long search in this forum and others I couldn't find this kind of "ready made" kit.
I've seen all the I\V stage schematics in this forum but I have a very limited access to components in my area of living so building one from scratch is almost impossible for me.
I'll appreciate your help finding this kind of a kit to buy online.
Cheers,
Guy
...I also would like to understand please the principal of using transformers as an output stage, as far as I know using a resistor for the I\V conversion (passive I\V) is not recommended due to added distortion.
when using a transformer there is a resistor that from my understanding is doing the I\V conversion and the transformer is for voltage "amplification" (correct me if I'm wrong) so what is the idea behind it if a resistor is still doing the I\V conversion?
when using a transformer there is a resistor that from my understanding is doing the I\V conversion and the transformer is for voltage "amplification" (correct me if I'm wrong) so what is the idea behind it if a resistor is still doing the I\V conversion?
I don't think transformers are a great choice as part of an I/V stage though for interfacing to cables they are excellent. A step up transformer makes a good voltage amplifier but needs a low impedance source for driving it, a DAC output with a shunt resistor isn't it really. I suspect that the reason for using it is - being a passive device the transformer isn't upset by high levels of high frequency noise that emanate from DACs.
Good question, I was thinking that as I was typing my earlier reply.
'Low impedance' is of course relative, in this case its the ratio of the driving impedance to the primary shunt inductance which seems to be the relevant ratio. For a step-up transformer the shunt inductance is relatively low as the primary has so few turns.
There's a good paper which covers issues such as this which you can download from Plitron's website. Actually, several very good papers there.
@bguy - you're asking if my discrete I/V can handle 1mA? Yes that's also the output current of the TDA1387 which I designed it for. I've not looked at the AD1862 datasheet in some time, can't recall if its output is bi-directional or not. So the biassing current sources might need changing.
'Low impedance' is of course relative, in this case its the ratio of the driving impedance to the primary shunt inductance which seems to be the relevant ratio. For a step-up transformer the shunt inductance is relatively low as the primary has so few turns.
There's a good paper which covers issues such as this which you can download from Plitron's website. Actually, several very good papers there.
@bguy - you're asking if my discrete I/V can handle 1mA? Yes that's also the output current of the TDA1387 which I designed it for. I've not looked at the AD1862 datasheet in some time, can't recall if its output is bi-directional or not. So the biassing current sources might need changing.
Last edited:
The ESS chips have pretty low output impedance and work very well in voltage mode as it is, so you can kind of do whatever the heck you want and it will work fine.
The AD1862 is one of the harder to deal with properly if I recall correctly. It has a high output impedance current source (more of a true current source than the ESS) and it's a small current as noted above. I am not sure its compliance is in the same league as even something like a PCM1792.
I would use an op-amp for this. I realize that's not the "cool" thing to do. If you absolutely insist on not using an op-amp, then look to common base / common gate circuits like the Jocko I/V or Zen/Sen/whatever the current iteration is.
Hi,
the current outputs of the DAC-Chip are quite low in impedance ... ADI specs a value of 2.1k.
The input impedance of a I/V converter stage should therefore be very low to not form a voltage divider that introduces distortion.
Also some DACs feature output protection diodes which start to conduct at already 200-300mV, also increasing distortion.
Hence a resistor as I/V converter wouldn´t be a first class choice.
Things do change though when You put a ´current Buffer´ between DAC output and resistor.
In its simplest form this could be just a single transistor.
Connected in common base/common source configuration, the DAC current is fed into the Emitter or Source of the transistor, which is a low-ohmic node.
Impedance values may range from a few Ohms down to a few mOhms, depending on the devices, circuit topology and bias settings.
The current fed into this node appears at the Collector (Drain) of the transistor which forms a high-impedance output --> current source output.
Values can range up into the MOhms, hence a far better current source than the DAC itself.
Also the voltage compliance can be much larger than the DAC´s (~+-1V).
The resistor value can be chosen as high as wanted ... even so high that a Power-DAC is possible that feeds the speakers directly.
Put at this position the resistor is a close to perfect I-to-V converter.
The resistor may then be followed by a (voltage-) Buffer stage to drive the cabling and the attached load.
A specially suited converter stage for the AD1862 is the JFET-based CEN that EUVL described in this Link
IIrc there was -or still is- a Kit available .... if not...it´s very simple to build.
A second useable converter topology is the single-ended grounde base or single-ended grounded source circuit.
The former with excellent measurements but heavily energy inefficient, the latter making use of common small signal transistors.
There are many incarnations and variants of all these topologies around and You can choose whichever You like most.
And there´s of course always a helping hand here to tune a chosen circuit to Your demands.
jauu
Calvin
the current outputs of the DAC-Chip are quite low in impedance ... ADI specs a value of 2.1k.
The input impedance of a I/V converter stage should therefore be very low to not form a voltage divider that introduces distortion.
Also some DACs feature output protection diodes which start to conduct at already 200-300mV, also increasing distortion.
Hence a resistor as I/V converter wouldn´t be a first class choice.
Things do change though when You put a ´current Buffer´ between DAC output and resistor.
In its simplest form this could be just a single transistor.
Connected in common base/common source configuration, the DAC current is fed into the Emitter or Source of the transistor, which is a low-ohmic node.
Impedance values may range from a few Ohms down to a few mOhms, depending on the devices, circuit topology and bias settings.
The current fed into this node appears at the Collector (Drain) of the transistor which forms a high-impedance output --> current source output.
Values can range up into the MOhms, hence a far better current source than the DAC itself.
Also the voltage compliance can be much larger than the DAC´s (~+-1V).
The resistor value can be chosen as high as wanted ... even so high that a Power-DAC is possible that feeds the speakers directly.
Put at this position the resistor is a close to perfect I-to-V converter.
The resistor may then be followed by a (voltage-) Buffer stage to drive the cabling and the attached load.
A specially suited converter stage for the AD1862 is the JFET-based CEN that EUVL described in this Link
IIrc there was -or still is- a Kit available .... if not...it´s very simple to build.
A second useable converter topology is the single-ended grounde base or single-ended grounded source circuit.
The former with excellent measurements but heavily energy inefficient, the latter making use of common small signal transistors.
There are many incarnations and variants of all these topologies around and You can choose whichever You like most.
And there´s of course always a helping hand here to tune a chosen circuit to Your demands.
jauu
Calvin
Hi,
the CEN (complementary) uses complementary JFETs iIrc.
That way the signal current splits into two halves travelling along the upper and the lower part of the circuit, combining again in the I/V resistor.
The SEN (single-ended) uses JFETs of same polarity (N-channel).
Here the lower part JFETs function as a constant current source and the upper part JFETs need to take over the full signal current.
As the complementary JFETs only need to handle half of the signal current the THD can be even lower.
Also the input impedance should be lower as the complementars are connected in parallel signal-wise.
Unfortunately are complementary JFETs seemngly becoming unobtanium, but the circuit can easily be translated into the bipolar transistor world.
See post #43 and post #62of the Sen-CEN-ZEN thread (#62 w. LTSpice-file)
Then just add a Buffer stage and You´re done.
For the DAC1862 it doesn´t matter which circuit it sees, as long as the I/V converters bias current is larger than the DAC´s peak output current, hence when its >1mA.
With common small signal transistors You´d be running the stage probabely on 4-10mA bias ... hence nothing to worry about.
jauu
Calvin
the CEN (complementary) uses complementary JFETs iIrc.
That way the signal current splits into two halves travelling along the upper and the lower part of the circuit, combining again in the I/V resistor.
The SEN (single-ended) uses JFETs of same polarity (N-channel).
Here the lower part JFETs function as a constant current source and the upper part JFETs need to take over the full signal current.
As the complementary JFETs only need to handle half of the signal current the THD can be even lower.
Also the input impedance should be lower as the complementars are connected in parallel signal-wise.
Unfortunately are complementary JFETs seemngly becoming unobtanium, but the circuit can easily be translated into the bipolar transistor world.
See post #43 and post #62of the Sen-CEN-ZEN thread (#62 w. LTSpice-file)
Then just add a Buffer stage and You´re done.
For the DAC1862 it doesn´t matter which circuit it sees, as long as the I/V converters bias current is larger than the DAC´s peak output current, hence when its >1mA.
With common small signal transistors You´d be running the stage probabely on 4-10mA bias ... hence nothing to worry about.
jauu
Calvin
Last edited:
AD844
If you gan get access to some AD844΄s, i highly recommend them for I/V duties. I used triple stacked Ad844΄s per channel in my AD1865 dac and it sounds really great. Havent used it in AD1862 but im planning to do so. Take a look at this thread
http://www.diyaudio.com/forums/digital-source/227677-using-ad844-i-v.html
If you gan get access to some AD844΄s, i highly recommend them for I/V duties. I used triple stacked Ad844΄s per channel in my AD1865 dac and it sounds really great. Havent used it in AD1862 but im planning to do so. Take a look at this thread
http://www.diyaudio.com/forums/digital-source/227677-using-ad844-i-v.html
Hi,
instead of a stacked pair/triple of AD844s (a pair seems the minimum to cope with the +-1mA of DAC current) a single OPA860 or OPA861, which use the same topology as the AD844, may be used.
The OPA861 featuring a OTA only (Operational Transconductance Amplifier), the OPA860 also featuring a integrated Buffer, like the AD844.
If You want to design Your own Buffer something like a diamond made from bipolars is what is typically used in the integrated circuits.
If You want to use the bipolar I-V linked to in my last post You may even use the same bipolars in a diamond buffer.
jauu
Calvin
Here´s a fully dimensioned sim.
instead of a stacked pair/triple of AD844s (a pair seems the minimum to cope with the +-1mA of DAC current) a single OPA860 or OPA861, which use the same topology as the AD844, may be used.
The OPA861 featuring a OTA only (Operational Transconductance Amplifier), the OPA860 also featuring a integrated Buffer, like the AD844.
If You want to design Your own Buffer something like a diamond made from bipolars is what is typically used in the integrated circuits.
If You want to use the bipolar I-V linked to in my last post You may even use the same bipolars in a diamond buffer.
jauu
Calvin
Here´s a fully dimensioned sim.
