dac I/V convertion with very low distortion

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
So let's say we have minimal DC offset. It may be a good idea to put some filtering after this and then a buffer. I'm thinking the DCB1.

A buffer could be useful with a DAC which outputs only around 1mA peak, such as the PCM1704. For DACs which output significant current, such as does the PCM1794A which outputs around 4mA peak, the resulting output impedance should likely be sufficiently low for almost any line level application without a buffer stage.
 
Thanks very much for the sims Avro. I was thinking in stabilize in temperature the first Ccs that you post with a transistor connected as a diode in series with the led. The tl431 sims nice but I am afraid that it will be the noisier. And difficult to do a complement.
 
Last edited:
You may be right..., however my LT simulator says different.... I wonder if the models aren't sufficient even to calculate CCS impedance.... hmmm So much for for believing that..

In my experience the ltspice do very good simulations with bipolar transistors, very close to reality. But is not very good with jfet, it gives to much gate current leak in every models that I have use, but I don't have investigate the problem further.
 
George, whether a 2.35k output impedance is too high or not will mostly depend on the total amount of shunt capacitance presented by the interconnect cable (typically, about 50pF per foot) plus that of the input of the next gain stage. The output impedance forms a low-pass filter in conjunction with that shunt capacitance.

Ken, I think George are thinking that the load for the dac will be 2.35k.

George, the dac will see a 3.3 ohms impedance. that is the iv converter input impedance ( in the version post here) . the 2.35k is the output impedance of iv converter .
 
Ken, I think George are thinking that the load for the dac will be 2.35k.

George, the dac will see a 3.3 ohms impedance. that is the iv converter input impedance ( in the version post here) . the 2.35k is the output impedance of iv converter .

Yes, thanks for clearing that up, the way Ken worded it it (post 517) which I misconscrewed sounded like the input impedance of your I/V was whatever the output shunt resistor was, which left me scratching my head.
I am watching and your progress with this, hoping it will end up in a kit form or ready made to insert into cdp's and dacs.

Cheers George
 
CCS and impedance... or inverse of Impedance..:)
 

Attachments

  • CCS-imp.JPG
    CCS-imp.JPG
    117 KB · Views: 422
  • CCS's.JPG
    CCS's.JPG
    68.5 KB · Views: 416
I do not get too much gate currents in MicroCap. These two models for ex, give me pA of gate current at Vds = 20VDC:

.MODEL 2SK170V_10M NJF (AF=499.843M BETA=29.5953M CGD=19.8997P CGS=24P IS=10F
+ KF=.001681153F LAMBDA=1M RS=7.19294 VTO=-0.65)

.MODEL 2SJ74V_16M3 PJF (AF=500.034M BETA=35.5713M CGD=106.132P CGS=73P IS=10F
+ KF=.00163621F LAMBDA=1M RS=4.75307 VTO=-0.75)


LTspice cannot be doing something wrong with JFETs?



/S

In my experience the ltspice do very good simulations with bipolar transistors, very close to reality. But is not very good with jfet, it gives to much gate current leak in every models that I have use, but I don't have investigate the problem further.
 
As the pcm1794 has a high output impedance i thought about using the 2sk170/2sj74 pair in the input. I have bought 4 matched pairs from Spencer Cheung, they are not cheap, it cost the equivalent of 1 Kilo gram of BC327 :D . but i am very curious to try them, and this project seems to be perfect for them. So there it is, the version that I going to build.
I have found an elegant soluction to connect the dc servo, instead of connecting the servo at the input as usual, the servo will be connected to the input of current sources, any noise that the servo generate will be filtered by C2 and C3 (I have already test this, and work very well). So this I/V converter will not have output capacitor.
In the schema is absent the ccs of 6,2ma for balance the pcm1794 dc_offset.

I will do also a version with BC327/337 transistors in the input. And i am working in a I/V for the ESS9018 .
 

Attachments

  • tagus fet.png
    tagus fet.png
    37.1 KB · Views: 544
Just for curiosity, the fft of the output with 2Vrms out , 7,8ma p-p.

The impedance of the source is 100k with lower impedance the distortion will be worse. But pcm1794 has a output impedance higher than that, so no problem :cool: .

in the schematic , Q10, Q12, Q13, Q3 will be the BC550c for npn and BC560c for pnp. not bc857c bc847c as this are smd.

edit: This is a ltspice simulation,i will do real measurements when i receive the jfet.
 

Attachments

  • distortion.png
    distortion.png
    63 KB · Views: 491
Last edited:
I have more faith in the BJT version.. I have made phono-stages with Jfets, they always sound nice, but also wailed and cloudy compared to the Bjt's

Like your Servo- very much, the same way I designed one in for the paradise phonostage--there we use the OPA134.
I would make a jfet current source instead of the two resistors setting the current. and then insert the servo with two 100 kohm resistors on each side of the CCS.
I would then use the 6.2mA countercurrent to trim the offset to 0 before engaging the servo.
Then you know the servo is just dealing with thermal drift.
 
Last edited:
MiiB I will do also a bipolar version, that i will post tomorrow, just need to do some final tests tonight. I want to try something with this jfet. And i realy think the jfet version will be very good.

The current source use in this schema is very good and generate very low noise, no need for improvements.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.