Sorry I didn't follow much the thread.
maybe the issues of opamp have been discussed many here.
But I am on the middle of choosing opamp among of OPA1611, ADA4627 and OPA1656.
I heard OPA1656 is good but it is dual opamp, can someone share the legs arrangement to adapted with MIRO board?
Thanks
maybe the issues of opamp have been discussed many here.
But I am on the middle of choosing opamp among of OPA1611, ADA4627 and OPA1656.
I heard OPA1656 is good but it is dual opamp, can someone share the legs arrangement to adapted with MIRO board?
Thanks
Pins 2,3,4 are the same function in both packages (the single and dual versions).
So the best is to use this side of the dual package.
Pin 1 is usually not used, NC in the applications, so in effect one can solder directly in all pins at one side, 1,2,3,4. Plus Pin 8 is soldered in, too.
So alltogether 5 pin is fix, soldered.
The other side mainly should be isolated from the pads below, with a small patch of tape.
So the chip is lying on the surface, half of it is soldered in place, ~ half is 'floating'. But it is already quite well defined mechanically.
Then in some way (below or above the chip)
Pin 1 - - Pin 6 should be connected,
Pin 8 - - Pin 7 should be connected, and the chip Pin 8 soldered in, as told before.
This is trickier: Pin8 (with chip soldered) should be connected to trace 7 (under the isolation tape).
A thin piece of wire does it.
So again: pins 5; 6; 7 of the OPA1656 must be isolated from the trace below. But (isolated) trace 7 should be connected to (soldered in) trace+pin 8.
Then, above the tape, connect the floating chip pins:
CHIP PIN7 - CHIP PIN 6;
CHIP PIN5 - - GND to somewhere on the board.
That's all.
I like 1656 very much..
Pic from here:https://www.diyaudio.com/forums/dig...1862-tht-i2s-input-nos-2r-47.html#post6278705
Ciao, George
So the best is to use this side of the dual package.
Pin 1 is usually not used, NC in the applications, so in effect one can solder directly in all pins at one side, 1,2,3,4. Plus Pin 8 is soldered in, too.
So alltogether 5 pin is fix, soldered.
The other side mainly should be isolated from the pads below, with a small patch of tape.
So the chip is lying on the surface, half of it is soldered in place, ~ half is 'floating'. But it is already quite well defined mechanically.
Then in some way (below or above the chip)
Pin 1 - - Pin 6 should be connected,
Pin 8 - - Pin 7 should be connected, and the chip Pin 8 soldered in, as told before.
This is trickier: Pin8 (with chip soldered) should be connected to trace 7 (under the isolation tape).
A thin piece of wire does it.
So again: pins 5; 6; 7 of the OPA1656 must be isolated from the trace below. But (isolated) trace 7 should be connected to (soldered in) trace+pin 8.
Then, above the tape, connect the floating chip pins:
CHIP PIN7 - CHIP PIN 6;
CHIP PIN5 - - GND to somewhere on the board.
That's all.
I like 1656 very much..
Pic from here:https://www.diyaudio.com/forums/dig...1862-tht-i2s-input-nos-2r-47.html#post6278705
Ciao, George
Attachments
Last edited:
Don't waste too much time on oaps ! The important thing is the power supply and decoupling : you can correct many thing by working on it and adapt the sound of the dac to the equilibrium of your hifi devices (speakers, amp, room...etc). Sometimes if not often , the difference are more huge than playing with oaps but keeping the same powersupply-reg-caps arrengment !
Try instead to spend monney on a good R-Core traffo both for oap and the dac chips, try different caps : nasonic FC, FR, eventually the Nichicons of the BOM and take care about local decoupling (the MKPs).
On the board you can for instance use the 5 mm spacing for some lytics instead mkps if you don't find it in 2.5 m spacing (whic is best of course : the littlier the gap the best the impedance of the cap) . You don't always need a huge lytics as well : sometimes 0.1 uf to 1 uF can be enough related to you reg toology... etc ! It is a good board and sota dac chip to experiment with that ! Cool project !
The exercice is to find an equilibrium, a trade off you like.
It's a difficult exercice here because the I/V and buffer is made by the same device and it is two different role...
The opa1656 or the opa861 are certainly enough if not the best choices ! YMMV of course as usual
Try instead to spend monney on a good R-Core traffo both for oap and the dac chips, try different caps : nasonic FC, FR, eventually the Nichicons of the BOM and take care about local decoupling (the MKPs).
On the board you can for instance use the 5 mm spacing for some lytics instead mkps if you don't find it in 2.5 m spacing (whic is best of course : the littlier the gap the best the impedance of the cap) . You don't always need a huge lytics as well : sometimes 0.1 uf to 1 uF can be enough related to you reg toology... etc ! It is a good board and sota dac chip to experiment with that ! Cool project !
The exercice is to find an equilibrium, a trade off you like.
It's a difficult exercice here because the I/V and buffer is made by the same device and it is two different role...
The opa1656 or the opa861 are certainly enough if not the best choices ! YMMV of course as usual
Last edited:
Thank youDon't waste too much time on oaps ! The important thing is the power supply and decoupling : you can correct many thing by working on it and adapt the sound of the dac to the equilibrium of your hifi devices (speakers, amp, room...etc). Sometimes if not often , the difference are more huge than playing with oaps but keeping the same powersupply-reg-caps arrengment !
Try instead to spend monney on a good R-Core traffo both for oap and the dac chips, try different caps : nasonic FC, FR, eventually the Nichicons of the BOM and take care about local decoupling (the MKPs).
On the board you can for instance use the 5 mm spacing for some lytics instead mkps if you don't find it in 2.5 m spacing (whic is best of course : the littlier the gap the best the impedance of the cap) . You don't always need a huge lytics as well : sometimes 0.1 uf to 1 uF can be enough related to you reg toology... etc ! It is a good board and sota dac chip to experiment with that ! Cool project !
The exercice is to find an equilibrium, a trade off you like.
It's a difficult exercice here because the I/V and buffer is made by the same device and it is two different role...
The opa1656 or the opa861 are certainly enough if not the best choices ! YMMV of course as usual
You have a very good 2.5 mm pitch MKT in the Wima catalog.
Now, all is about the shortest path o the ground. Take a look on the board, you can play with 3D layout : for instance using a via/hole for the ground - the nearest from the active pin- and solder the other in direct on the leg of the oap or its socket !
But again the typology of the reg/powersupply of the oaps are making a lot ! Good news : one can experiment as the regs are not on the board ! I'm on the side of the linear regs but there are some cool shunt regs, modern chips as the one talked in the thread; discrete shematics a la P. Rodgic that are fast (darlington without feedback is best imho) etc !
Take two boards : one you find good as reference, then take the second to eperiment. Always the sae tracks you know. Measurments if you can but not mandatory for the fun. And always return to the reference to check ! One rule : always change one part only before checking - to know/listen what happen -
There is worse to loose time that such hobby
You can hav efun without having tons of scope as the main was made by Miro1360 materialised with the board
Now, all is about the shortest path o the ground. Take a look on the board, you can play with 3D layout : for instance using a via/hole for the ground - the nearest from the active pin- and solder the other in direct on the leg of the oap or its socket !
But again the typology of the reg/powersupply of the oaps are making a lot ! Good news : one can experiment as the regs are not on the board ! I'm on the side of the linear regs but there are some cool shunt regs, modern chips as the one talked in the thread; discrete shematics a la P. Rodgic that are fast (darlington without feedback is best imho) etc !
Take two boards : one you find good as reference, then take the second to eperiment. Always the sae tracks you know. Measurments if you can but not mandatory for the fun. And always return to the reference to check ! One rule : always change one part only before checking - to know/listen what happen -
There is worse to loose time that such hobby You can hav efun without having tons of scope as the main was made by Miro1360 materialised with the board
And can I use those slightly thinner 0.1uf 63v film caps of WIMA in C34/35?? I think I may borrow a pair from a friend.
Thanks
Yes, ada4627-1 is lither in bass. I hear less bass in drums etc. But i think i dont have the perfect setup of power supply around ada4627-1. I saw his datasheet and they recomend 0.01 at the ps pins to ground less than 3mm. I have ada4627-1 in an adapter and that adapter is in a socket...then the 0.1uf, so a bit far. I am also using salas ulbib3 on +-12v section. Ada4627-1 sound seems faster than normal lol. Some other opamps sound slower....the highs and background info is very clean and no arsh on highs....highs are perfect , in complex pop music with a lot of background info, everything can be eard .. very good.
Whats your opinion on ada4627-1 sound?
LM6171 have stronger and defined bass in my opinion... very good.
Thanks Miro, will use them and test with my Salas Ultra BiB 1.3 psu for +/-12vdc and a LT3045/3049 for +/-5vdc on the digital side. Currently I have NE5534 opamp but would like to try the LM6171 which is easily compatible and per reviews sounds much better. In the meantime I will also test with Burson Audio V6 classic/vivid but fitting them along with the side caps will be tricky. Lets see how it goes. Also by the way is there any initial setup steps like using the trimmer to set the offset etc.,