Thanks.
That's what I understood, so one flip flop has Vcc1, the other Vcc2. Digikey showed the 74f74SC (for soic) obsolete, but perhaps I am wrong. Advantage of the soic is it is easy for instance to putthem with all the digital signal on the bottom pcb, the uper layer being DGND, the upper DGND is AGND so the top layer is easy to layout output signal, DEM decoupling and a grill between TDA1541A pins in order each current find its "bus" easily up to pin 15 (aka new-AGND). Ending with only an affordable 4 layers for the masses. It is easy to make the 3 voltages PS and current loop shorts and input PS with the new voltage pins arrengement. Of course, no prob with more layers as even simplier. But I like to think current as a 3D phenomena and i do not thoink in layer but also in stack direction. The current is as far I understood a free back packer turrist that want to travel where he wants according the speed/frequency bus he chooses (and when I talk about bus, it is also about elevator I am talking about as it is 3D.
Well nothing fancy I wanted to do on the hybrid mu : a cascoded CCS for the B+, a HV Hexfet depletation, and a low fet or other mosfet for the low voltage follower part ! simple 0.1 uF cap; something between V7 and V8 of Moglia's website. I thougth about the J113 as a the low fet as it is near than the ECC88 mA (15 to 20 mA in the real world for most of the J113) ! Of course a J310 with 50 mA could be a better line driver. I prefer to92 because it is easier for servicing than smd.
But I let for others the tube pcb section (although 3D cabling is more elegant but not for all), I most prefer the Bjt because my wallet for the I/V, and it seems it is the most prefered while you know about the tubes (but good ones became rare, chocks cost an arm today. And aso Zoran thought of an hybrid (while the 211 seems too much not quieter but certainly more than good enough for a driver/amp/buffer for the few having high efficienty loudspeakers!
Well, my last experience with pcb workgroup and pcb ending in blahxxxx-ction pcb is thanks, but no thanks (not again) ! I see people here using bad TDA1541A pcbs since 2014, lol ! Not my delirium ! This is what we asked Pedja Rogiuc back then to relaunch the AYA diy pcb 🙂 ! Project manager is not for everyone, whatever it is diy !
1205 casing feasible, yup, just more tricky for the 14 DEM and bigger PCB (width : each side of the TDA1541A).
Last edited:
This is something I would like to sort out, please ?
I can make a tht pcb, but it will be easier to make things with soic on the digital front end imho and whatever soic or tht, with ref that are sourcable at Mouser/Digikey/Farnel to be sure of the quality, or at least easily sourcable, imho ! (No problem if you want. I make a pcb THT for your needs because you have parts etc, : pcbs, it is the only thing I can make to contribute to make life easier or firering faster for people that "think" shematics ! 🙂
You think not the same when layouting soic or THT, so we must start from a BOM or with refs having same casing and pins orientation (BJT/FETs/MosFET).
Of course everyone can draw his own pcb, but this is not what I meant there.
I optimized more the shunt circuit. This one is most versatile. Decreased the value od C i RC nets around the 431. Without loosing main chrs. Even lowered down Zout.
.
3rd RC net can be ommiited. This one squared in the sch. The rsults are simulated without this RC net.
.
.
3rd RC net can be ommiited. This one squared in the sch. The rsults are simulated without this RC net.
.
This is my "Memory" IV. NPN and PNP versions. The issues can be very limited nuber of types for P-JFETs... Versions for TDA1541A and TDA1540 (with 2mA Io)
.
Sins are with only 100R Riv. Could be higher of course. These are very simple solutions. I try to combine this Memory concept with other.
.
.
.
Sins are with only 100R Riv. Could be higher of course. These are very simple solutions. I try to combine this Memory concept with other.
.
.
Don't know why the links on the pages do not work but here are direct links:
https://www.audio-perfection.com/vo...formance-simulation-of-tl431-shunt-regulator/
https://www.audio-perfection.com/vo...31-model-development-analysis-and-evaluation/
https://www.audio-perfection.com/vo...r-regulators/testing-tl431-model-performance/
I wrote this now several times.
I will build something on Veroboard, using mostly BIY (ready made) modules and everything through hole, with a few exceptions. Multiple layers will use kapton tape and copper tape (adhesive).
Parts used will be "what I can easily source out here in the boondocks".
I show it, because it illustrates some ideas and may help others to learn. It is not meant as "a project you should copy". Making it worth showing and explaining also helps me get things clear in my own head.
My hope is that peeps ask "why is he doing it that way" and think for themselves.
I'm not guru seeking followers, nor am I teacher looking for students. I'm just throwing breadcrumbs out.
If we want to make a design for machine assembly by JLCPCB etc. that people can order mostly assembled by themselves from there (no, I don't want to sell PCB or stuff like this to people), I would do it very different in a number of areas.
This would include putting IIS2SIM on board with discretes (well IC's but not FPGA/CPLD), using ECL for re-clocking and driving the TDA1541 inputs, probably a footprint for a Chinese Amanero compatible USB module, perhaps a second one to fit another input (BT, SPDIF etc) as china now makes them on Amanero footprint etc. plus perhaps a secondary jitter killer (SI5713?).
That would allow 0603 SMD and hyper modern design approaches and logic choices in TSSOP or TVSSOP as basic defaults without any real penalties, handing over to SOP or SOIC ECL to all the TH stuff at the TDA1541.
All the discrete transistor circuits can use modern SMD Duals that get pre-mounted and have excellent thermal coupling.
And yes, I would order one of these myself.
It's just way OTT and more than I can manage for a one off build.
Thor
I surely purchase a roll of kapton and copper tapes then... I perfecccctly can live with that & a verroboard if everyone can makes his own pcb and helps himself and the time I spent for myself is is not usefull for others, whom have not that time.🙁
It maid sense to me as all I have seen on this bar for tda1541A pcbs is well... not good, at least everyone his own expectations and I surmise as far people are happy, it is okay !
I will try to understand more and translate your BOM with what I have avialable here in Europe... You explained a lot, 😊
I fear there is no more nigth markets here where I live you can eat both bacon and rice and source at the next corner at the same times bags of electronics component ! Mouser/Digikey/Farnel became the almost only sources of this old part of the world called Europe (here France, olympiggs area for few days left) ! Brick & Mortar shops vapored, while clicks & Potars' blasted !😒
Humm I 'd like better finally to seat street and eat a nasi goreng ! Certainly to old to have a guru, not yet for a beer, cheers !
It maid sense to me as all I have seen on this bar for tda1541A pcbs is well... not good, at least everyone his own expectations and I surmise as far people are happy, it is okay !
I will try to understand more and translate your BOM with what I have avialable here in Europe... You explained a lot, 😊
I fear there is no more nigth markets here where I live you can eat both bacon and rice and source at the next corner at the same times bags of electronics component ! Mouser/Digikey/Farnel became the almost only sources of this old part of the world called Europe (here France, olympiggs area for few days left) ! Brick & Mortar shops vapored, while clicks & Potars' blasted !😒
Humm I 'd like better finally to seat street and eat a nasi goreng ! Certainly to old to have a guru, not yet for a beer, cheers !
Last edited:
Somehow I cant reach yhis one from inside the site? Maybe because my brovser is older one Safari osx...
Thanks 🙂
it is more like from Bootcamp to OSX. Back and forth. To use windows spice soft and then back to post the pictures at the forum 🙁
I didnt install parallels... 😕
I didnt install parallels... 😕
There is a feature in almost all forums that truncates web addresses for layout convenience. Perhaps your browser is not able to direct you correctly from those.
Try this (or copy & paste)
Alas, it seems group conversations do not work here at diya. I don't really want to start a separate public thread. Any ideas?
Thor
I don't really want to start a separate public thread , why ??
this thread is mostly made of and around John ideas and some people still want to go and publish that way here on this very thread , it will be easier to open a new thread made of your ideas
.
this thread is mostly made of and around John ideas and some people still want to go and publish that way here on this very thread , it will be easier to open a new thread made of your ideas
.
FWIW following this discussion with interest and learning a lot, unfortunately not able to contribute though as this is way over my head...
I would reckon the discussion is still very much on-topic as "building the ultimate tda1541 dac" is still what this thread aspires to do.
We are talking about coordinating multiple jobs and people and discussing the real minutiae and detail of implementation.
It's a lot of back and forth with little to contribute that is of interest to anyone not involved.
I prefer to keep threads at high SNR.
Thor
I started on TDA15xx based dacs back in early 90's. I even bought+built Thorstens Adagio DAC.
I have built+listened to many many DAC chips (same peripherals), but there is something special about the TDA15xx.
Done right, it is extremely smooth, effortless and "analog" sounding.
It is the only Dinosaur I know built on diff ECL + emitter scaling.
@Thorsten:
Is that the reason why it sounds different than newer CMOS, high sample rate DACs?
Thanks
I have built+listened to many many DAC chips (same peripherals), but there is something special about the TDA15xx.
Done right, it is extremely smooth, effortless and "analog" sounding.
It is the only Dinosaur I know built on diff ECL + emitter scaling.
@Thorsten:
Is that the reason why it sounds different than newer CMOS, high sample rate DACs?
Thanks
H2 ?
But its sound IME has a lot to see with how you decouple the power supplies pins... But there is no doubt the latest production (Asia) had a cleaner, less smooth but still analog (imho the latests are more "fidel", while the european made are more forgiving on bads recordings and less well executed cd players/DAC))
my 2 cents
But its sound IME has a lot to see with how you decouple the power supplies pins... But there is no doubt the latest production (Asia) had a cleaner, less smooth but still analog (imho the latests are more "fidel", while the european made are more forgiving on bads recordings and less well executed cd players/DAC))
my 2 cents
Here is an interesting article on sychronous reclocking:
https://hifiduino.wordpress.com/2012/11/01/synchronous-reclocking/
Also attached the data sheet of Potato Semiconductor's 74G74A (clock frequency 600 MHz). But I think it is not avalable anywhere.
TI's 74F74 is 100 MHz
74S74 is 110 MHz
74AC/ACT74 is 125 MHz
74AUP1G74 is 90 MHz (3.3 V logic!)
The rest (7474, 74HC74, 74HCT74, 74LS74, CD74HC74, etc. all are around 25 to 30 MHz max. clock freq., not interesting for us.
Last edited: