Injured dac kit

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Last night while wiring in a on/off switch on a diy dac kit (Hong Kong-no name)and almost finished, I managed to pull a solder tab and break the hair thin wire attached to the primary side of the power tranny, aaaaargh! Talk about cheeeeeap.

So where to go for a replacement?

Primary= 0-110v 0- 110v (either wired for 110 or 220) something closer to 120v would be better.

Secondary= looks like 12v tap + a 5v tap. I'm just not completely sure at this point, simple wiring instructions call for DC input of 12v ; 5v. This looks to be a regulated supply.

Here's a link to the same dac kit as mine:

It doesn't seem like there is much I can do with the broken side as I kinda dug into it last night trying to find anything left of that tiny @#%!!$& wire.

Thanks for any help/suggestions.


Why not ask the seller as according to his eBay feedback profile, he is an excellent communicator? Surely, it is not too much to ask of him what the secondary voltages should be when you are already a paying customer?

Seeing the markings on top of this transformer, it is unusual for the primaries and VA rating to be stated clearly, but with no mention of any secondary voltages. This probably means that it is a 'special' made for this particular application (unlike generic 6,9,12, or 15V etc.) and the makers often don't print these 'unusual' voltages when smaller manufacturing 'runs' are involved.

Peter Daniels' similar kits have (IIRC) 8V secondaries, for example, although a 9V would usually be OK for that application.

I have used 10V secondaries in similar circuits, and don't forget that merely 1V AC difference on the secondaries gives around 1.4V DC change when rectified by a full bridge rectifier.

The problem with simply guessing is that if you end up with a too low voltage on the secondaries, there could be insufficient headroom for the regs to regulate properly, but if voltages are too high there is the risk of excessive heat-dissipation in the regs etc., to contend with.

Thanks Bobken,

I'll give that a try, although this isn't the original seller of mine. Perhaps he would still enlighten me. He did send an email as to what he would charge for a replacement (?) xformer $40, I will also ask the original seller if he has a clue.

One other thing, I plan on trying to check right side of primary (other half 0-110) to see if I can get a reading on secondary. Probably using my dim bulb tester and hope for no shorts.

Will let you know how it turns out.

OK, well after digging in deeper here is what I have found:

Xformer secondary is center tapped supplying LM78XXC series regulators, two- LM7812C for analog output, one-LM7805C for digital side.

LM7812= 12V output, requires 14.6V to maintain line regulation, rated @ 1.0A typical load up to 1.5A w/ proper heatsink.

LM7805C= 5V output, requires 7.5V to maintain line regulation, rated @ 1.0A typical up to 1.5A.

Since the 7812's require 14.6V to keep them working, are 1A rated and that the only labeling on the secondary side of the original xformer states 20VA, it would seem conceivable that a suitable replacement would be a 18 or 20VCT @ 1A secondary.

20V x 1A = 20VA.

Oh btw, I couldn't get a reading on the injured xformer, a disaster from my haste, stupidity, and a few glasses of wine.

Any thoughts on my ramblings?


With regret, I am not going to take the responsibility for your outlay on a transformer, here, and I would rather you discovered what the original transformer secondaries are rated at, both in voltage and current output capability. As already mentioned, there are quite a few factors involved, and I have little knowledge of what these are in reality.

Clearly this circuit has more than the TDA chip to satisfy for PS requirements, since as I said, the max. usable supply for this chip is around +9v (and IIRC around 56mA for current), and you say some of the regs give +/-12v output.

As it seems like a centre-tapped arrangement here, and there is a need for +/- 12v like you suggest, and a +5v supply for other parts of the circuit, it is not certain how this is achieved, so some guesswork comes into the matter. If there were more secondary tappings than 3 showing here, I would have guessed that the +/- 12v secondaries would be at a higher VA rating than the single 5v is, as this would probably be more usual here, and there is no need for the 2 secondaries to be the same VA rating as each other. They can be both different voltages and VA ratings.

However, probably (but not certainly) the 2 secondary tappings are used for the separate +/- 12v supplies initially, and one of these (the +12v side) is regulated down to +5v in a second regulation step. I have no idea what the currents being drawn by these various circuit requirements are, but would assume that the 2 x 12v requirements are likely to be similar. If the 5v requirement is not excessive (as doesn't seem likely, but this is only another guess), then it is also likely that the 2 secondaries are similar in VA rating as you suggest. Unfortunately, the max ratings for these regs don't help much, as their combined capability is higher than you will get out of a 20VA transformer like this, and you haven't taken into account in your final calculations that there are 2 secondaries. Normally, the VA rating will be for the entire transformer's output, so with 2 x 20v AC outputs at 1A like you suggest, this would be a 40VA transformer.

To adequately supply the max. apparently 12v supplies, I would have thought that dual secondaries in the region of 12/15 volts will be adeqaute, but as I already said, it is most likely that this is a special voltage here to suit this application, and I have previously pointed out the snags if you get this wrong.

Disregarding your possible local mains voltage difference compared with the primary voltage-rating of the chosen transformer, and that if not fully-loaded there will be a consequential rise in output voltage as well, a nominal 12v secondary output device should be OK.

12vAC x 1.4 = 16.8vDC, less say 1.2v for diode losses gives 15.6 DC, *theoretically*, and this should give enough headroom to be suitable for the 12v regs from what you know, without wasting too much in excessive heat dissipation.

This would mean 830mA (0.83A) secondaries if you use a 20VA transformer, and again entirely at a guess, I would think that this will be fine.

I think that you will need to resolve this for yourself now, but I still urge you to find out what the original transformer outputs nominal voltages are, as presumably this is known to work properly.


P.S. I just noticed that both 12v regs are positive from their numbering, but this doesn't make much material difference to what I said.
The decription tell us it uses CS8416 receiver, CS4397 DAC and NE5532 OPAMP. There is no current sucker like the TDA154X series DAC. It specifies maximum current consumption of only 35mA. Anything better should be surplus.

Mostly likely the +12V, -12V are for the OPAMP NE5532, the +5V is for digital. The +5V should be derived off the raw rectified +12V from the transformer. I do note that there seems to be a regulator on the main PCB, which may provide the +5V for the DAC from the regulated +12V.

So you probably just need a 20VA or better transformer with 24V seconday with center tap, or two 12V secondaries.


Having realised after posting that both of your regs are positive polarity (assuming that they are both LM812Cs) another more likely possibility exists here, which I previously discounted when I thought that there was a +/- 12V supply, in addition to a +5V supply.

This highlights the dangers of any guessing (as I have mentioned several times) but this was due to my assumptions that more frequently there would be +/- similar voltage supplies, than 2 separate supplies of exactly the same voltage.

The calculations I suggested are unaltered, as per my P.S., but the assumptons I was forced to make initially are now less likely to be correct here.

It now seems more likely that there are 2 secondaries, one for the dual 12V supplies of the same polarity, and one for the 5V supply. This being the case, your suggestion that both secondaries might be the same VA rating is now not very likely to be so at all, and it is anyone's guess how this 20VA rating is apportioned. One is quite probably in the region of 12VAC, and the other maybe 5 or 6VAC, but one could be say 5VA and the other 15VA, or any other possible combination.

I therefore strongly recommend, once gain, that you find out what the original voltage and current ratings were. If this is quite impossible, the only way then to be confident would be to measure the currents drawn by each secondary (which is what I would do here) but, assuming that they are different as is now most likely, I doubt that an 'off the shelf' transformer will be satisfactory, anyway.


Edit. This was written before seeing the intervening post, so I will read and digest that next to see if it throws any more light on this matter. In the meantime, are both regs as you say? i.e. LM812s, as this same +12V polarity, compared with the dual polarity issue, makes quite a lot of difference here where the likelihood of secondaries is concerned.
Sorry, 7912 is correct. So there is the - / +12v. I will try and attach a pic of underside of PS board with proper markings.

I am quite certain that reg for digital side +5V is derived solely from 12v feed. Actually rather simple layout on the board :

bridge rectifier(full wave)- filter caps-bypass caps-regs-bypass caps-filter caps. The line out from PS board is: +12v/AG(analog ground)/-12v/DG(digital ground)/+5v in that order.

Bob, I did ask the seller that I purchased from as well as seller of the current Ebay listing ie: link in previous post, have not heard back.
However, current seller will sell me same transformer for $40 including airmail. Perhaps I should just take him up on his offer, but where's the fun in that? Seriously though, original xformer has 110v primary, so running 120v+ here is going to up voltage.
10v may or may not make too much differance.

Happy ending-hopefully.

Hi Chris & kf_tam,

Now that the regs issue is quite clear, I revert back to my original suggestion, which agrees with kf_tam's recommendation as well. From different standpoints we have reached precisely the same conclusion.

2 x 12V secondaries (20VA transformer) each providing 0.83A should do the trick here then, as far as can be ascertained, and I hope that it works out for Chris OK.

Guessing is never ideal, especially where someone else's money is concerned, but I don't think that we can do any more to help with this now. Taking the 'safe option' and going for an original spec transformer would give you 120/110 x 12VAC secondaries *theoretically* (assuming that they are nominal 12V secs.), which is just over 13VAC, of course.

Although I did originally look at the seller's spec from the reference in the first post like 'kf' has very sensibly just done, due to age (!) I suppose, I forgot this info was available when I came back to the subsequent replies, so I made it harder for myself, rather stupidly. :xeye:

Sincere Thank You

Bob, thank you for all your in depth help, also kf_tam for helping to shed light on this.

Even though my initial carelessness caused a detour in completing my dac project, it has also been an educational experience with the help of two kind individuals here,

I will let you know how she sounds.

Hi Chris,

From my point of view, you are very welcome, and it is always of interest to know the outcome in instances like this.

Diagnosing faults and/or predicting someone's requirements is rarely an 'exact science' when dealing with matters at arm's length like this, but hopefully we haven't let you down in this case.

Good luck.

Happy Ending

Thanks to all that have responded, I can say that I am once again a happy camper.

Ordered a 24v center tapped 2A transformer from Marvac, not only saved $ but was a drop-in replacement to boot. Measured output voltage from supply board and was spot on at 12V.
I have it all buttoned up and in my living room setup, between a Toshiba SD-3960 and a Musical Fidelity X10v3 tube buffer, I must say it sounds very nice indeed.

Now on to the next project.

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.