This is done all the time.
Every mains powered product that is plugged in and switched on that you touch is:
you connected to the Floating and Isolated secondaries.
It's the safety regulations that keep you safe.
When you are building your own mains powered equipment, YOU become the "SAFETY regulator"
YOU MUST ensure that your are isolated from the Mains !!!!!!!!!!!!!!
Hi and sorry very much to come again on this point but i have here a dac with just the transformer external. Any other thing is inside the dac.
I also remember units from Musical Fidelity with just a transformer external.
Is it really so bad to have an AC umbilical ?
I ask this because voltage transfer should be much more efficient in AC form than DC
Am i wrong on this point ?
I remember for instance that with DC bigger cablles sections are mandatory
Thanks a lot again. Regards, gino
Just how many kilowatts are you supplying?
It is considered 'cheap and dirty' to run AC in an umbilical...
Hi and thanks for the reply.
It is a convenient solution. Powerful transformers are heavy and tend to vibrate. The idea to take them out from the amp chassis sounds very reasonable.
At the same time especially with power amps i believe that the closer the reservoir caps to the amp circuits the better.
This is a picture that i like where caps are very close to the output transistors (fast discharge, low impedance rails, etc.), with all the energy stored in the amp chassis.
Speaking of umbilical with just one transformer to connect i do not know if a quad wires cable could be enough ... maybe another wire to connect ground to mains E ?
In this way the actual amp would be lighter and the transformer could be placed in a heavy box with handles and wheels.
Some transformers are even more than 10 kilos.
There is always personal experience at the origin of the question.
I think that a two boxes amp could be much easier to move around.
One very heavy amp is a problem to move. It could weigh also 20 kilos when the chassis is made out of thick steel.
Also vibrations could be kept away from circuits and this i think it is beneficial as well.
Thanks a lot again.
Kind regards, gino
Last edited:
Why?ginetto61 said:
Yes.
You may be thinking of long-distance power transmission. If you have to send power over many 100's of km then AC has an advantage because you can easily step up the voltage. Not relevant to powering audio equipment.
You still seem to be searching for golden bullets. Why not seek understanding instead? Then in a given situation you can decide for yourself whether to use AC or DC, and where to put the caps.
Hi and yes i was thinking to long distance power transmission
but if there is a principle it should be always valid, on principle at least.
I like the idea to have the transformer in a separate box very much.
Because i know that many tranformers buzz, vibrate and i read that sometimes this mechanical vibration is transformed in electrical vibrations that increase overall noise.
Well more than for golden bullets i am searching for good practices.
But sincerely i like the big transformers out.
Nevertheless i understand also that i am quite alone🙂 in this idea. 😱
Always they are placed inside the chassis with the amp circuits.
Thanks again. Kind regards, gino 😀
The principle of low loss long distance AC transmission is still true, but it is irrelevant. In audio power our aim is low noise, not low loss. We get low loss from using short cables, not high voltages. It is not sufficient for a principle to be true; it also has to be relevant and correctly applied.ginetto61 said:
Good practice must be what 99.9% of all amplifiers adopt: the whole PSU inside the amplifier box/chassis.
You mean that taking out the transformer only could increase the overall noise ? 😕
This is exactly the contrary of what i thought.
Of course the idea was to lower the noise not to increase it.
Ok no more discussion. The evidence is in the popularity of the solution.
That is zero so it is not the right thing to do.
Thanks a lot again, gino
Hi and let me elaborate.
Look for instance here on the low right side ...
24VAC/500mA ... bad design ? this is my fundamental question
Of good 24VDC regulated power supply is full the world, and nevertheless they have decided to go with only an external transformer.
Someone advised me that is better to keep the rectifier very close to the transformer.
So is Musical Fidelity wrong ? and it is not alone
I am sure that an external transformer is absolutely better ... i am sure
Like i am sure that especially for circuits operating in class AB where the current draw is rapidly changing like power amps the closer the reservoir caps to the output stages the better
Instead of big caps far away and small caps close i would do the viceversa.
But to be honest in the end i have seen example of both AC and Dc umbilicals
Thanks again, gino
Last edited:
Keep the transformer secondary cables and the bridge rectifier together to ensure a VERY LOW LOOP AREA.
Keep the bridge rectifier and the smoothing capacitors together to ensure a VERY LOW LOOP AREA.
If the transformer is centre tapped, or converted to centre tapped, then keep the centre tap wire with the secondary cables to ensure a VERY LOW LOOP AREA.
And keep the centre tap with the rectifier to smoothing cap cables to ensure a VERY LOW LOOP AREA. The centre tap wire does not connect to the rectifier, but it does pass through the middle of the group of rectifier terminals to maintain that LOW LOOP AREA.
This requires a twisted triplet from transformer to bridge rectifier AND a twisted triplet from bridge rectifier to smoothing caps.
Part of the reason for the twisted triplet is to give low loop area but also to minimise interference due to the VERY HIGH pulsing currents in the charging loop/s
Keep the bridge rectifier and the smoothing capacitors together to ensure a VERY LOW LOOP AREA.
If the transformer is centre tapped, or converted to centre tapped, then keep the centre tap wire with the secondary cables to ensure a VERY LOW LOOP AREA.
And keep the centre tap with the rectifier to smoothing cap cables to ensure a VERY LOW LOOP AREA. The centre tap wire does not connect to the rectifier, but it does pass through the middle of the group of rectifier terminals to maintain that LOW LOOP AREA.
This requires a twisted triplet from transformer to bridge rectifier AND a twisted triplet from bridge rectifier to smoothing caps.
Part of the reason for the twisted triplet is to give low loop area but also to minimise interference due to the VERY HIGH pulsing currents in the charging loop/s
Hi !
yes. I have recurrent obsessions 😱
But I have decided for the DC umbilical connection between the external power supply and the device, a usb soundcard with 12VDC socket.
Mainly because I do not want to mess too much with the internals.
I should place a regulator and I am not that good at installing things in tight spaces.
Moreover 12VDC power supplies are quite available (I could build something in the next future anyway).
I have opened a similar unit and I can see a 2200 uF cap just after the DC socket 🙄
My guess is that it is acting like a local energy reservoir placed closer to the circuit board ?
I have a question about the max size of this cap. I think that the bigger the better.
But if it is too big it could be seen like a short circuit by the power supply ?
Could 2200 uF be an optimum then ?
The DC umbilical will be about 1/1.5 meter long.
Thanks a lot again.
Have a nice day, gino
yes. I have recurrent obsessions 😱
But I have decided for the DC umbilical connection between the external power supply and the device, a usb soundcard with 12VDC socket.
Mainly because I do not want to mess too much with the internals.
I should place a regulator and I am not that good at installing things in tight spaces.
Moreover 12VDC power supplies are quite available (I could build something in the next future anyway).
I have opened a similar unit and I can see a 2200 uF cap just after the DC socket 🙄
My guess is that it is acting like a local energy reservoir placed closer to the circuit board ?
I have a question about the max size of this cap. I think that the bigger the better.
But if it is too big it could be seen like a short circuit by the power supply ?
Could 2200 uF be an optimum then ?
The DC umbilical will be about 1/1.5 meter long.
Thanks a lot again.
Have a nice day, gino
Last edited:
'The bigger the better' is sometimes true in engineering. Often it is not true.
There is not a simple answer to your question. You need to consider things like how the PSU behaves when it sees a (temporary) short, what the resistance and inductance of the cable will be, what your aim is. In other words, you need to design your circuit.
There is not a simple answer to your question. You need to consider things like how the PSU behaves when it sees a (temporary) short, what the resistance and inductance of the cable will be, what your aim is. In other words, you need to design your circuit.
Hi ! thanks a lot for the very valuable advice.
Usually this kind of units are delivered with very cheap smps wall wart adapters.
Speaking of noise I have not seen much of a difference between a cheap one and a decent linear psu.
Anyway i like the idea to have big uF very close to the circuits that use the power.
I guess I should look for "inrush current limiter circuits/devices" maybe ?
This is going to be very complicated. 🙁
Thanks a lot again, gino
You do not need inrush current limiter circuits/devices, as such are build into the regulators you may use... If you may use Schottky as rectifiers, which are specified for enough high continuous current, then these devices it may tolerate hundreds amps as pick currents... At least the best and safe current limiter (for inrush picks) is the transformer itself. Your power system it can never take a larger current (both picks or continuous) than the transformer it can deliver (without burning...). Some experimenting and done!
Last edited:
Hi ! thank you very much indeed for your valuable advice.
From my last post there has been an evolution.
I will keep the mains transformer out of the box and put the voltage regulator board in the case as close as possible to the dac.
I have a feeling that an AC umbilical is a better solution than a DC one.
So the situation has changed.
The transformer is likely to be a EI type, or maybe a R-core.
But i am attracted by EI types after seeing them used commonly in control units and lab power supplies.
There must be a reason why i do not see toroidals.
So the real issue now, still open, is the regulator solution.
Thanks a lot for the useful directions.
My goal now is to select a kit, assembled or not, and maybe replace some parts with higher quality ones, like the Schottky diodes you mention or better caps, in order to enhance its performance as much as possible.
But I really would like to start from a kit or at least a ready pcb.
I need 12VDC/1A at the output of the regulator (actually the dac's current draw is about 0.55 A at 12VDC).
As i said in the end i prefer the AC umbilical solution.
I like to keep the transformer out like i see in some preamps.
If you have any suggestion for a very good regulator kit i am all ears.
It is my main and only interest now.
Unfortunately this dac is quite sensitive to power supply quality.
But with the right one can be quite decent.
I only look for the flattest and lowest noise floor possible without having to use batteries (too messy).
Thanks a lot again, gino
Last edited:
Hi Gino
Indeed, an AC ombilical is a very right approach in your case, and is at least better to have the transformer out of the DAC box. More confortable/convenient is to have it all in the same box. However, if I may chose myself, I will chose to place the transformer as long as possible far from the rest of the DAC circuits/PCB, but not necessary out of the same box. If outside the DAC box, then is no problem to have it just beside. Few teens cm longer AC connection, it mean really nothing. However, I will prefer an no center tap approach... As your load is a so low current, and stable, as a preamp or DAC or whatever in this range it can be, then is no whatsoever problem with a longer AC connection. And yes, there is preferable a AC power connection, than a DC one.
A quite different situation (AC ombilical) it may be when about a (high) power amp, or even a power supply for a welding device...
You may know that the main reason it is used EI transformers in lab power supplies is the lower costs of a such transformer, but not about their special qualities. The EI transformer is worst as efficiency, weight, magnetic field emissions , and noise isolation from AC main. A toroid transformer is much more efficient, very low magnetic field around it, lower weight, smaller dimensions, but still be worst when about noise isolation from AC main. Both these transformer types may need AC filtering before it, and the EI one a good magnetic shielding over it.
The best it come out R core. This one is both most efficient of all, even smaller for the same power, very low magnetic field emission, and provide best AC noise isolation (lowest capacity between its windings).
If you will use a EI type, then there is a must to place it outside the DAC box. If you may use a toroid or even better a R core, then there is no any problem to have it inside the DAC box. Actually I see a such solution as most reasonable for your approach.
More, an R core transformer is quite easy to be adjusted for best dissipation on the regulator device, as it is easy demountable. However, in your case, for that low power/current you need, it may not be necessary to adjust a such R core...
To buy a regulator kit it may be a good solution for you. Then change and improve it as you can. Good (for simple circuits) regulators are LT10xx series, or MIC29xxx. You may want to use the adjustable version of a such, which provide a better control or improvement capabilities for even better noise performances. I will not want use 78xx series...
So, you may find some kits based on the above suggested regulators. A 3A one it may fit quite well for your needs...
Use the largest filtering capacity after regulator, and a moderate high capacity after rectifier bridge (to not burn the diodes at power on...).
Doing so, you may get a output ripple lower than few µV, and nothing of HF noises. However, the worst (noisy) linear PSU is much, much better than the best SMPS you may find out there... SMPS should never be used to power audio/video devices. Unfortunately there is just the opposite on the today marked...
Indeed, an AC ombilical is a very right approach in your case, and is at least better to have the transformer out of the DAC box. More confortable/convenient is to have it all in the same box. However, if I may chose myself, I will chose to place the transformer as long as possible far from the rest of the DAC circuits/PCB, but not necessary out of the same box. If outside the DAC box, then is no problem to have it just beside. Few teens cm longer AC connection, it mean really nothing. However, I will prefer an no center tap approach... As your load is a so low current, and stable, as a preamp or DAC or whatever in this range it can be, then is no whatsoever problem with a longer AC connection. And yes, there is preferable a AC power connection, than a DC one.
A quite different situation (AC ombilical) it may be when about a (high) power amp, or even a power supply for a welding device...
You may know that the main reason it is used EI transformers in lab power supplies is the lower costs of a such transformer, but not about their special qualities. The EI transformer is worst as efficiency, weight, magnetic field emissions , and noise isolation from AC main. A toroid transformer is much more efficient, very low magnetic field around it, lower weight, smaller dimensions, but still be worst when about noise isolation from AC main. Both these transformer types may need AC filtering before it, and the EI one a good magnetic shielding over it.
The best it come out R core. This one is both most efficient of all, even smaller for the same power, very low magnetic field emission, and provide best AC noise isolation (lowest capacity between its windings).
If you will use a EI type, then there is a must to place it outside the DAC box. If you may use a toroid or even better a R core, then there is no any problem to have it inside the DAC box. Actually I see a such solution as most reasonable for your approach.
More, an R core transformer is quite easy to be adjusted for best dissipation on the regulator device, as it is easy demountable. However, in your case, for that low power/current you need, it may not be necessary to adjust a such R core...
To buy a regulator kit it may be a good solution for you. Then change and improve it as you can. Good (for simple circuits) regulators are LT10xx series, or MIC29xxx. You may want to use the adjustable version of a such, which provide a better control or improvement capabilities for even better noise performances. I will not want use 78xx series...
So, you may find some kits based on the above suggested regulators. A 3A one it may fit quite well for your needs...
Use the largest filtering capacity after regulator, and a moderate high capacity after rectifier bridge (to not burn the diodes at power on...).
Doing so, you may get a output ripple lower than few µV, and nothing of HF noises. However, the worst (noisy) linear PSU is much, much better than the best SMPS you may find out there... SMPS should never be used to power audio/video devices. Unfortunately there is just the opposite on the today marked...
Last edited:
Hi ! thanks a lot again for the very valuable advice.
Free space inside the box is extremely limited. It is enough just for a regulator board. The original smps is little and noisy.
So I searched in the web and found a guy who has replaced it with a linear one getting rid of this noise.
I will post some pics of the kit. I am just waiting for the mains transformer to arrive.
I have a personal experience here. A friend was testing a line preamp prototype using different transformers in the power supply.
Switching on the lights in the room was sending a big bang in the speakers with the toroidal. Much much less with the EI. Just a little click.
I understand that also toroidal can work as isolation transformer, still I am more confident on EI.
For such low VA efficiency is not an issue.
I will follow your advice then. I am just starting now. It will be a long term project. It will take me some months. I am still curious to see what cheap and common units can give me.
I can assure you that the space inside is minimal and the risk to damage something is very high. Instead a regulator board can be fitted with not big effort. If the board is not too big of course.
This is a fundamental point. How large the cap following the regulator can be ?
I have at hand some 16V/2200uF Nichicon Gold. Will 2200uF be too much ? does this depend on the actual regulator used ?
I like this idea a lot indeed. In this way there will be a lot of power reserve just close to the dac.
Very nice indeed ! 🙂
Exactly my thinking ! when i look at the ripple figures i see about 1% of Vout for the good smps and from 20 to 50 times less for good linear psu.
Even high quality digital units like master clock generators use smps.
Actually the dac i am using has dc generation on board.
But again i would like very much that the power supply is not in any way a limiting factor. As i am going to stick with this dac for long i will take some time about this issue.
Thanks a lot again for directing me towards R-core transformers.
I will check immediately. I guess a 12VAC unit should be fine for a 12VDC power supply.
Kind regards, gino
Last edited:
- Status
- Not open for further replies.