@Tommar
Yes, those are 10uf and 1uf. I would increase to 22uf (c26) 2.2uf (c25) (or 47uf and 4.7uf). Use tantalums, better in that position than elco (i use film, rubycon mu 22uf and 2.2uf, best but also most expensive). And make sure cap is as close to the board as possible, the shorter leads the better. Job of that part of the circuit, in short, is to reduce noise.
Yes, it is usually better to have chips from the same batch unless you trim, but not always the case. Yes, replace riv resistors with metal film ones. Since you have oscilloscope, check if it is oscillating just in case.
Trimming would require you to have low level sine undithered input, and measure output on spectrum analyser. You basically adjust to bring down harmonic peaks. I wouldn't bother honestly if you don't have the required gear.
@MrHifiTunes those resistors should be metal film, and quality matters. Their purpose is to prevent parasitic oscillations. 22R is enough for most cases.
As for trimming, read up above. Mainly due to gear required to do the adjustment.
Yes, those are 10uf and 1uf. I would increase to 22uf (c26) 2.2uf (c25) (or 47uf and 4.7uf). Use tantalums, better in that position than elco (i use film, rubycon mu 22uf and 2.2uf, best but also most expensive). And make sure cap is as close to the board as possible, the shorter leads the better. Job of that part of the circuit, in short, is to reduce noise.
Yes, it is usually better to have chips from the same batch unless you trim, but not always the case. Yes, replace riv resistors with metal film ones. Since you have oscilloscope, check if it is oscillating just in case.
Trimming would require you to have low level sine undithered input, and measure output on spectrum analyser. You basically adjust to bring down harmonic peaks. I wouldn't bother honestly if you don't have the required gear.
@MrHifiTunes those resistors should be metal film, and quality matters. Their purpose is to prevent parasitic oscillations. 22R is enough for most cases.
As for trimming, read up above. Mainly due to gear required to do the adjustment.
You don't want absolute low ESR for decoupling the op amps, you have to try, it is often worse.
UKZ are goods but sometimes too much (and indeed low ESR). Anyway you need more caps to experiment, this is the fun of diy and they are not expensive. You can try w/o soldering first. Always switch off the device when you swap caps or op amps.
Definitly buy the UKZ, some FC, UKG before they are gone, some basic Elna. 47 uF is more than okay and 25V perfectly fine.
I will not dismiss the carbon resistor in the I/V, that is true it is more noisy and not wanted theorically in a feedback loop, but you will not really ear it and the tonal can be interresting according on how is sounding the rest. As usual, proof is in the pudding.
👍 Thanks, very helpful.
Is there a description on how and where to measure the parasitic oscillation?
Is there a disadvantage increasing the value?
Same for measuring the trimming.
I have an oscilloscope
Yes, for ease of use, use online tone generator, this one works fine.
https://www.szynalski.com/tone-generator/
With 22R on output, cable influence wouldn't make a difference, but you can check by connecting cable to dac output, on the end of the cable use connector which is loaded by for example 10k resistor (or whatever your input impedance on the following device is, also if you have info of the capacitance you can load it with that value as well). Hook up osciloscope, and play 1k 10k tune, sine and square. I'm going with that you know how to use osciloscope, so no need to detail that step?
22R fits most cases, i went with 47R in one occasion, simply because it was susumu rs, as a much higher quality resistor. No difference soundwise that i noticed.
You need spectrum analyzer. You can't do it on osciloscope.
Dear All, I purchased this module after I saw some reviews on a Russian forum, this module has SAA7220 and an FPGA, it could drive TDA1540 and 1541 selected through a switch 14bit/16bit also could be configured as NOS/OS. From what I understand of SAA7220 being a filter and all that, 7220 was generally used as the MCLK source in CD players. I know the FPGA would be doing the job of logic gates.
So my question is that if I give I2S output from say a Bluetooth receiver like CS8645/8675 or QCC5125 will reduce the jitter inherent to these modules (I have a poor experience with them especially with AD1865/62 and TDA1541 DACs there is too much of jitter)
Second part what if I connect its I2S input with the I2S output of a USB module like Amanero/XMOS etc.
Here is the link to the product and here is it brief, although considering the source I would rather discount the claims of features by atleast 50%.
Link to the module
Some features as claimed by the seller:
"
This version of the kernel circuit comes from the first version of TDA1541 384K digital version, which can drive TDA1541. TDA1540 to play music formats up to 384K/24bit,
Pin compatible, supports 512FS main clock (jumper can be set),
Supports 16bit I2S and can be directly connected to Raspberry Pi,
Supports one click NOS/OS switching,
Supports 16bit/14bit output settings,
Supports output phase conversion,
Supports balanced driving
Description of each interface of the block:
1. The J2 input of the module is an I2S signal, while J3 and J4 output a 14bit/16bit signal in Simultaneous format,
J5 PHASE is the phase setting pin, and a short circuit to GND is the phase reversal
3. J5 NOS is the setting pin for NOS/OS, and a short circuit to GND is considered NOS; Empty as OS
4. S2 sets the switch for NOS/OS, with a short circuit indicating NOS and a disconnection indicating OS; This foot is directly connected to J5 NOS"
So my question is that if I give I2S output from say a Bluetooth receiver like CS8645/8675 or QCC5125 will reduce the jitter inherent to these modules (I have a poor experience with them especially with AD1865/62 and TDA1541 DACs there is too much of jitter)
Second part what if I connect its I2S input with the I2S output of a USB module like Amanero/XMOS etc.
Here is the link to the product and here is it brief, although considering the source I would rather discount the claims of features by atleast 50%.
Link to the module
Some features as claimed by the seller:
"
This version of the kernel circuit comes from the first version of TDA1541 384K digital version, which can drive TDA1541. TDA1540 to play music formats up to 384K/24bit,
Pin compatible, supports 512FS main clock (jumper can be set),
Supports 16bit I2S and can be directly connected to Raspberry Pi,
Supports one click NOS/OS switching,
Supports 16bit/14bit output settings,
Supports output phase conversion,
Supports balanced driving
Description of each interface of the block:
1. The J2 input of the module is an I2S signal, while J3 and J4 output a 14bit/16bit signal in Simultaneous format,
J5 PHASE is the phase setting pin, and a short circuit to GND is the phase reversal
3. J5 NOS is the setting pin for NOS/OS, and a short circuit to GND is considered NOS; Empty as OS
4. S2 sets the switch for NOS/OS, with a short circuit indicating NOS and a disconnection indicating OS; This foot is directly connected to J5 NOS"
https://www.mouser.com/ProductDetai...Q_ND7iTToLspxElLMKE125Y443UPnynjsCwf4McWDaOw6
The LEDs specified in the BOM of LD1/2/3/4 of PSU2 are 5mm green.
Is the LED being used for any other function than indicating that the power is on?
Actually, I already have hundreds of 3mm blue LEDs, so I wonder if there is any problem using those instead.
Do the resistors need some time, or can you immediately draw conclusions from changing to others?
I changed the capacitors and the positive differences were immediately audible, but after changing Riv from Polish telpod 2.8kohm to Dale rn55 2.15kohm I had the impression that it was worse. After an hour I went back to the old ones.
I use Burson v7 classic. The manufacturer's website recommends a modification consisting in adding 0.1uf to 1uf capacitors between pins 4 and 7 of the operational amplifiers. What do you think about such a modification?
I also added 1000pf 1% wima mkp capacitors between pins 2 and 3, but I'm not sure if I hear an improvement.
I changed the capacitors and the positive differences were immediately audible, but after changing Riv from Polish telpod 2.8kohm to Dale rn55 2.15kohm I had the impression that it was worse. After an hour I went back to the old ones.
I use Burson v7 classic. The manufacturer's website recommends a modification consisting in adding 0.1uf to 1uf capacitors between pins 4 and 7 of the operational amplifiers. What do you think about such a modification?
I also added 1000pf 1% wima mkp capacitors between pins 2 and 3, but I'm not sure if I hear an improvement.
No need for rail to rail (pin 4 to 7) decoupling, it is done good on miro board. And why would you add 1000pf between inputs?
Are you using same volume setting on preamp with dales? Since comparing to those telpod, you are loosing roughly 1/4 output vrms, which can be interpreted by brain as worse sounding, unless you managed to adjust volume to same dB.
Are you using same volume setting on preamp with dales? Since comparing to those telpod, you are loosing roughly 1/4 output vrms, which can be interpreted by brain as worse sounding, unless you managed to adjust volume to same dB.
will start to solder the shift registers.
To make sure. Where is pin1 going?
Text on PCB is in the same direction as on the chip?
To make sure. Where is pin1 going?
Text on PCB is in the same direction as on the chip?
Yep, that is the proper orientation. You can always jusge by the notch or dot on the footprint, and compare to the chip. Where you see notch, thats the orientation where dot or notch on the chip points to.
👍 Thanks . I thought so but it has been to long. Better make sure before starting.
Job done. Getting better at it near the end. First time to use a heatgun.
Job done. Getting better at it near the end. First time to use a heatgun.
xin chào. sau khi đọc tất cả các bài viết ad1862 của miro1360. tôi quyết định xây dựng của riêng mình với cấu hình nos dac đơn sắc kép. chờ nhận được pcb từ nhà máy.
The board is finished.
I didn't had SMD capacitors (i thought I did). I used some regular ones. Made the legs as short as possible. Probably it's an x7r and not a C0G. Don't know if it's important.
Nichicon as mentioned in the BOM on page 1 isn't available anymore.
I bought some Panasonic's ECA's
The 47uF had a wrong footprint and needed to bent the legs.
Again made them as short as possible.
Is there a way I can test the board?
The USB interface is still on the way.
@1543 and @Paddy Garcia thanks for helping with the board and AD1862 chips 👍🙏
I didn't had SMD capacitors (i thought I did). I used some regular ones. Made the legs as short as possible. Probably it's an x7r and not a C0G. Don't know if it's important.
Nichicon as mentioned in the BOM on page 1 isn't available anymore.
I bought some Panasonic's ECA's
The 47uF had a wrong footprint and needed to bent the legs.
Again made them as short as possible.
Is there a way I can test the board?
The USB interface is still on the way.
@1543 and @Paddy Garcia thanks for helping with the board and AD1862 chips 👍🙏