@lemon:
thanks for your contribution. I agree that it is certainly better to correctly match the JFETs and I am working on this together with miniDSP. So I am quite sure that the DACs will soon be available again with well matched FETs.
If you want to match on your own using an Atlas DCA Pro with curve tracing function will surely give very good results.
I matched the FETs with respect to IDSS using this Page and it works perfectly for this circuit. The MMBFJ310 do not neccessarily need matching (DC offset will be only a few mV if they are not matched) but the BF862 should be matched with +/-0.2mA or better.
regards, Daniel
thanks for your contribution. I agree that it is certainly better to correctly match the JFETs and I am working on this together with miniDSP. So I am quite sure that the DACs will soon be available again with well matched FETs.
If you want to match on your own using an Atlas DCA Pro with curve tracing function will surely give very good results.
I matched the FETs with respect to IDSS using this Page and it works perfectly for this circuit. The MMBFJ310 do not neccessarily need matching (DC offset will be only a few mV if they are not matched) but the BF862 should be matched with +/-0.2mA or better.
regards, Daniel
Daniel thanks for the reply.
It is very positive announcement that the Curryman dac is available again with well matching the jfets!
I was to dillema what to do, to buy some jfet and matching them or to put the R2 and pot to the right position of designer (Joachim).
Finally, I decided to order some jfet and matching them.
From what I see, there is a changed at MMBFJ310 with the MMBFJ111 to Joachim Gerhard Filter Buffer thread.
Do you know what is reason for that? Is it some upgrade or for availability reasons?
It is very positive announcement that the Curryman dac is available again with well matching the jfets!
I was to dillema what to do, to buy some jfet and matching them or to put the R2 and pot to the right position of designer (Joachim).
Finally, I decided to order some jfet and matching them.
From what I see, there is a changed at MMBFJ310 with the MMBFJ111 to Joachim Gerhard Filter Buffer thread.
Do you know what is reason for that? Is it some upgrade or for availability reasons?
Last edited:
just a minimal correction to avoid missunderstandings: will soon be available again
The original buffer design from Joachim used MMBFJ310 and for the JG filter buffer GB it was changed to MMBFJJ111. e.g. EUVL can surely better explain the reasons. I discussed the schematic with Joachim during the "development" and desided to use his original design. However both perform very well in this buffer, no need to worry
Well, I think because it takes a lot of time, effort and cost for a small group of individuals to pull off commercial DIY product(s). They're dependent on small assembly shops, etc. Each time the shop or supplier screws up, I suspect that all involved are eating the cost of the whatever needs to be redone.
We all want it right the first time, but honestly - I get why many who are capable of making their designs available, don't. If something goes wrong (your fault or not) buyers can be very unforgiving.
Sorry - not really targeted at you damuffin; just think we all need to get our minds right about how to support small audio efforts. It'll cost us a bit in patience, but for it we get access to some very leading edge kit.
We all want it right the first time, but honestly - I get why many who are capable of making their designs available, don't. If something goes wrong (your fault or not) buyers can be very unforgiving.
Sorry - not really targeted at you damuffin; just think we all need to get our minds right about how to support small audio efforts. It'll cost us a bit in patience, but for it we get access to some very leading edge kit.
Finally, I used the Atlas DCA Pro and I measured 20 of BF862 by Vgs (on) and Id. I matched 4 of them (3 of them was just and the 4th was very close to them).
The old values was -63.13mV and 51.18mV.
Now the values is -0.12mV and -11.2mV and I am OK.
Hey Abel,
with 3 times the current the resistors R1-R4 will drop more voltage and probably need to be adjusted to have enough voltage at the output. I am currently in Seoul and do not have access to the spice sim but when I am back on weekend i'd take a look. Or just apply U=R*I and do some tests ;-)
regards, Daniel
with 3 times the current the resistors R1-R4 will drop more voltage and probably need to be adjusted to have enough voltage at the output. I am currently in Seoul and do not have access to the spice sim but when I am back on weekend i'd take a look. Or just apply U=R*I and do some tests ;-)
regards, Daniel
curryman DAC with Rasberry Pi
Today I tested the DAC with the Rasberry Pi. Works like a charm
I am using RasbyFi with modified Kernel 3.10 for I2S output at P5 Header. Used my own setup however it is quite similar to the HifiBerry implementation. Currently I am using the ES9018 sound module which works good so far. Tested playback up to 96/24 FLAC, 192/24 will be tested next week
@Abel: One Filter for all will be OK. Most of the Filter effect at SMPS switching frequency and above (>100kHz) is coming from the common mode chokes and R's only add some additional filtering at audio frequencies (where PSRR of regs is quite good).
regards, Daniel
Today I tested the DAC with the Rasberry Pi. Works like a charm
I am using RasbyFi with modified Kernel 3.10 for I2S output at P5 Header. Used my own setup however it is quite similar to the HifiBerry implementation. Currently I am using the ES9018 sound module which works good so far. Tested playback up to 96/24 FLAC, 192/24 will be tested next week
@Abel: One Filter for all will be OK. Most of the Filter effect at SMPS switching frequency and above (>100kHz) is coming from the common mode chokes and R's only add some additional filtering at audio frequencies (where PSRR of regs is quite good).
regards, Daniel
Attachments
With all the aging dual-core laptops some of us have laying around, I wonder what the allure of the Raspberry Pi is? Can you network it and have it grab FLAC from a networked NAS? Grab a console on it from a laptop or phone and tell it what you want to hear? IS there a library app that runs on the Pi? So music which is physically poorly laid down on disc, can be indexed, sorted by Genre, artist etc??? Obviously I have not played with the Pi, they are cute but what can they do that a Laptop can't? I2S direct out would count as a huge plus, maybe that's plus #1?
The Pi is smaller and consumes less power. It can also be interfaced easily with geek HW, because it has simple GPIO pins. Try finding that on any laptop made after about 2000 - the industry dropped support for the LPT port and made everything USB.
On a vaguely related note, the XMOS startKit is one of the cleverest little learning tools for experimenting now available. Lots of GPIOs, analog inputs, LEDs, etc. It can also interface with a Rasp Pi.
XMOS startKIT, a 500MIPS multicore development board for $14.99 | XMOS
There are a huge number of community projects that target the XCore architecture (some or most of which can be adapted to run on the StartKit - the toolchain is identical). A fair number of them are related to digital audio:
Projects developed by XMOS community | XCORE
That is nice but what's its practical use when you still need a computer to start playing an album ? If a button panel + display would exist then it would be stand alone device with no fan etc. Connectivity is the buzzword but maybe one tends to forget that many people don't want pc's in the living room.
Jean Paul, are you referring to the XMOS startKit or to the Pi?
For the Pi you won't need a computer. E.g. RaspyFi runs MPD (Music Player Daemon) and many clients exist for iphone/Android and even for WP8. So its really standalone and low power (and sounds quite good).
For the Pi you won't need a computer. E.g. RaspyFi runs MPD (Music Player Daemon) and many clients exist for iphone/Android and even for WP8. So its really standalone and low power (and sounds quite good
).- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.