Tda1545 Tda1543 Tda1541

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Yes and No.

TDA1545 and TDA1543 are pin compatible, but not replacable. You should change/remove the caps and/or resistors on the Vref pin (7). But this would probably be the waste of time.

Generaly speaking, you could swap it also with TDA1541, but that demands more work. You’ll need new pcb and new supply for DAC, and also new analog stage.

Pedja
 

Eldam

Banned
2012-09-13 12:25 am
France
TDA 1545 is an excelent dac chip with a special good sound in highs...

IMHO, you will have better result to try a light tweaking like a good Panasonic os con sepc or Nichicon FP for the 5V and the ref pin (need a very stable feeding). Some thread in DIYAUDIO talk about that.

Swapping with a TDA 1543 or 1541 is a lake of time because you need to change to much things, and of course work on the buffer stage after the TDA1545.

Just my two cents of course (I tweak manyTDA1543 an 45 CD players)
 
Last edited:
In my experience, TDA1545 has been trickier to get right than TDA1543, but well worth the extra effort as it sounds more transparent when the implementation is fully optimized. The main reason TDA1543 tends to be preferred amongst DIYers I think is because its more benign in terms of glitching, being a bipolar technology device. A top-notch TDA1545 output stage can't use a bog-standard opamp for I/V because of their dislike of RF. Passive has worked best for me.

I agree Eldam its worth paying attention to the pin7 drive as noise or ripple on this pin isn't strongly rejected - the basic PSRR of the TDA1545 is rather sucky (30dB from memory).
 
TDA1543 use I2S as input while TDA1545 uses EIAJ.there's TDA1543A which use EIAJ but hard to find.

i used 3*TDA1545 and they sounds pretty good combined w/ 12AX7 output. :)

you might want to tweak the supply such as changing the caps w/ the better one, adds bypass caps, or even change the regulator to something like LT1083 if the original one was 7805. (1083 was variable output.you'd better set the voltage at something like 3V before connecting it to the DAC then adjust to 5V after connected to DAC)
 
As power supplies for TDA1545A, I've tried a few. TL431 shunt, BJT-assisted TL431 (lower Z-out) and TL431 as Vref feeding an EF. I prefer this last one (the shunts improved a lot with lots of electrolytics on them) but I'm not yet done with PSUs.... A normal 3pin reg feeding a single TDA1545 is going to have highish Zout due to the very low current demand (3mA or so).

+1 for the question about the PNP on the 723. This chip is very low noise, lower than 317 as far as I recall.
 
Last edited:
As power supplies for TDA1545A, I've tried a few. TL431 shunt, BJT-assisted TL431 (lower Z-out) and TL431 as Vref feeding an EF. I prefer this last one (the shunts improved a lot with lots of electrolytics on them) but I'm not yet done with PSUs.... A normal 3pin reg feeding a single TDA1545 is going to have highish Zout due to the very low current demand (3mA or so).

so..the vref and Vdd was separated?..it might be a good idea!.Vref might absorb ripples from the main supply and separating the Vref supply could help!
 
I was using about 2,200uF as decoupling for pin7, same resistor values as in the DS. I figure passive filters do the best job of noise reduction.

you splitted the regulator out or ?..i don't really get it but what i get is you split the Vdd and Vref capbank using a CRC network :confused:

this is what in my mind :
regulator - C |-R -C (Vdd)
**********|-R -C (Vref)

am i right?
 
I was using RCR where Rs were about 15-22k (can't recall exactly) and C was 2200uF. VDD to first resistor of RCR chain, pin7 to the second one.

When I had the BJT assisted shunt, I put 20,000uF on its output and this improved the bass rather a lot. Trouble is - 20,000uF is rather bulky!

whoa?..20,000uF?.isn't that overkill?.i used something around 1047uF here and bass is pretty good :confused:
 
More capacitance always sounds better to me, overkill or not. Currently my DAC has close to half a farad of decoupling (on the active stages, about 20,000uF on the DAC chips) and the bass sounds truly awesome :) My amp hasn't caught up yet though, still has only about 50,000uF....:p Long overdue for upgrading. Good job caps are cheap here :D

i have chemi-con 150,000uF 100V here..is it useable? :D :D
i'm afraid it would destroy my trafo and regs :p
 
I parallel up lots of small caps, have tried using a big single one (68,000uF) but didn't sound as good. Here I get 3,300uF/10V and solder them up 40-50 to a board in parallel. So one board is over 0.1F. Takes a while to power up with this much uF but the wait is worth it :D

i'll give a try.more caps in parallel means lower Z which means lower ripple (but too big capacitance might kill your regs :D ) .i have 50pcs of 1000uF panasonic FC here :p

but it takes space much :scratch: .should i place the caps outside?:eek:
 
Yes, this huge quantity of caps won't fit on the boards. For my DAC I've built 4 100,000uF PCBs which I have used as the 'feet' for the analog PCB. They can be mounted underneath by a few cm of thickish wire (say 20AWG). I only use these caps to decouple LF, the HF is handled by a big stack of 1206 22uF ceramics on the chips themselves.

When the caps are on the output of a reg, the ripple current is negligible and low ESR isn't a requirement as there are so many in parallel that the wiring begins to dominate over the internal resistance of the caps. Hence Panasonic FC are definitely overkill in this application - use very cheap 'lytics where the ESR isn't disclosed :)

I shall be interested to learn if you hear a difference - for me it improves the soundstage at LF, so I get more of the original acoustic coming through. Also rhythm is enhanced, more likelihood of foot tapping....
 
Yes, this huge quantity of caps won't fit on the boards. For my DAC I've built 4 100,000uF PCBs which I have used as the 'feet' for the analog PCB. They can be mounted underneath by a few cm of thickish wire (say 20AWG). I only use these caps to decouple LF, the HF is handled by a big stack of 1206 22uF ceramics on the chips themselves.

When the caps are on the output of a reg, the ripple current is negligible and low ESR isn't a requirement as there are so many in parallel that the wiring begins to dominate over the internal resistance of the caps. Hence Panasonic FC are definitely overkill in this application - use very cheap 'lytics where the ESR isn't disclosed :)

I shall be interested to learn if you hear a difference - for me it improves the soundstage at LF, so I get more of the original acoustic coming through. Also rhythm is enhanced, more likelihood of foot tapping....

hmm...i'll search for cheap one..maybe chemi-con GXE is enough?..

anyway, for OP, i would recommend you to improve the PSU first before going more far :) .PSU is the key for good audio
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.