are these DC blocker any good?

[3roomlab]

http://www.diyaudio.com/forums/power-supplies/273392-low-noise-toroidal-transformer-vs-normal-toroidal-2.html

there are also many diagrams on internet.

http://bilder.hifi-forum.de/medium/181565/dc-blocker_378260.jpg

has anyone used 1 on your power lines to solve DC problems?

even double trap on both L/N lines
http://www.haoleb.com/pics/dctrap.jpg

(the reason for this question is, on a freak occasion i had my hand on the chasis of the bench DMM, i could feel some AC vibration. i could not draw a solution, so here i am with what i saw above)

[madires]

 To block DC you would need a series capacitor.

[Seekonk]

I'm still pretty sure transformers hum because they don't know the words.

[janoc]

Uh and that is supposed to achieve what exactly?


The diyadio.com forum post seems to be a load of rubbish by folks who don't have an idea about electricity (is that an audiophile forum, by chance?). I think that this gem sums it up pretty well:

Best way to get rid of DC on the trafo is to use an isolation transformer.

 


The "dc-blocker" schematic shows a gizmo that I seriously hope is not meant to be mains powered! 6.3V rated elcos in series with mains???  Someone thinks that those two anti-parallel diodes are going to keep that safe?? That looks pretty lethal (Not to mention that those diodes pass whatever comes in right through, minus the forward voltage of the diode, so it will produce a distorted sinewave if anything - possible EMI problems).

The last one is just a variant on the previous one - an excellent way to set your house on fire or worse, because those elcos are certainly NOT designed to carry so much current, never mind their 35V rating. One bad contact or blown diode and KABOOM the whole contraptions goes. This is like a Darwin award candidate project. 

If your amp is humming with 50/100Hz hum, you need to add/fix the filters on the secondary side of the PSU!

[janoc]

Serious? The caps will never see greater than 1V on both polarity. Diodes usually fail short, so it is not likely to blow the caps. Eben this happens, the caps will blow up before being charged to a lethal voltage.

Totally serious - did you check the posts in that forum, including the construction photos? Those fellows power 1kVA amps from these things. The diodes have no heatsinking and will carry most of the current. All it takes is for one to blow open. Or be poorly soldered.

And even if the diodes hold up, the caps will still be stressed by about 0.7-1V AC voltage. I am not an expert on elco chemistry, but I somehow doubt that those big capacitors will like that longterm.

Oh and never mind such minor thing like safety standards ...

[Richard Crowley]

Good for what? 
Do you actually HAVE a significant DC offset in your mains power?
Is it actually causing some Real World symptom?

It seems like a snake-oil audiophool solution to a "problem" that doesn't exist.
Unless you actually have some evidence that you have not revealed yet?

[grumpydoc]

Not to mention that those diodes pass whatever comes in right through, minus the forward voltage of the diode, so it will produce a distorted sinewave if anything - possible EMI problems.

+1

3roomlab - have you actually measured a DC residual in your domestic mains supply?

[grumpydoc]

I don't understand why they did this either, but the schematics design looks good to me, despite whether it is needed or not is unknown to me.

It's supposed to block DC but the pair of anti-parallel diodes will pass any DC offset greater than their forward voltage so I'm pretty sure that's a fail. 
Edit: On reflection they will just pass the DC offset while distorting the sine wave somewhat.

The caps will pass a fair bit of current - 10,000 uF at 50Hz has an impedance of 0R32, with a 10A load and V_F of 1.0V for t he diode they will take about a quarter of the load current.

Sure, under ideal circumstances the caps won't see much voltage but if one of the diodes fails open.........   

[3roomlab]

I don't understand why they did this either, but the schematics design looks good to me, despite whether it is needed or not is unknown to me.

It's supposed to block DC but the pair of anti-parallel diodes will pass any DC offset greater than their forward voltage so I'm pretty sure that's a fail. 

The caps will pass a fair bit of current - 10,000 uF at 50Hz has an impedance of 0R32, with a 10A load and V_F of 1.0V for t he diode they will take about a quarter of the load current.

Sure, under ideal circumstances the caps won't see much voltage but if one of the diodes fails open.........   

BOOOM ! 

anyway, i went to read up on some diodes, and it seems some may start to conduct as early as 0.3v-0.4v, so i think the calculations on the audio sites based on 0.7v as a starting point should be wrong isnt it?

also, based on the minimum VF, it would mean the capacitors will have to be larger to escape ever turning on the diodes. yes?

*edit* thx for the below post blueskull. i guess i just have to make DC blocker to try it 

[fpliuzzi]

Rod Elliott(ESP) did a fairly indepth write-up on the subject of blocking mains DC offset.

http://sound.westhost.com/articles/xfmr-dc.htm

I used the low-pass filter idea put forth by forum member The Electrician that was mentioned in the post below to measure the DC offset on the mains in my home last year (I made the low-pass filter using a 100K ohm resistor and a 450V AC rated motor run capacitor that I had on hand).

I only measured 3mV of DC on my mains using this filter and a Fluke 27/FM. This test was done purely out of curiosity after reading Rod's interesting article.

Post by The Electrician (in another EEVBlog thread on this very subject).

www.eevblog.com/forum/beginners/mains-voltage-with-non-zero-dc-how-bad-is-it/msg396249/#msg396249

[3roomlab]

ok so i took a video of the DC drifting about. it is there all the time, but under 0.5v . the UNI-T is set to measure 220.00vDC. it seems to oscillate by itself (a harmonic of 50Hz?), i am surrounded by over 300 households ... with alot of aircon compressors 

youtube.com/watch?v=7W_jOXeelX0[/url] (update :: this measurement i have done is wrong, -- > as suggested a [L]--680k-0.47u--[N] arrangement is made, which reveals a DC bias running around wild @ 1 to 4volts instead!) my math is wrong !


https://www.eevblog.com/forum/beginners/mains-voltage-with-non-zero-dc-how-bad-is-it/msg396249/#msg396249
good thread !

[3roomlab]

i wired a 680k and 0.47uF in series to read some voltage. am i right to say, the voltage i am seeing across the cap is 1:101 in ratio? (0.47uF approx = 6.8k, making the divider as 1:101?)

[coppice]

I'd use an isolation transformer to block any DC. They are available at any local sub-station.

[madires]

i wired a 680k and 0.47uF in series to read some voltage. am i right to say, the voltage i am seeing across the cap is 1:101 in ratio? (0.47uF approx = 6.8k, making the divider as 1:101?)

Have you wired L --- resistor --- cap --- N?

[Kalvin]

Run it through UPS with real sinewave inverter.

[3roomlab]

i wired a 680k and 0.47uF in series to read some voltage. am i right to say, the voltage i am seeing across the cap is 1:101 in ratio? (0.47uF approx = 6.8k, making the divider as 1:101?)

Have you wired L --- resistor --- cap --- N?

yep

[madires]

i wired a 680k and 0.47uF in series to read some voltage. am i right to say, the voltage i am seeing across the cap is 1:101 in ratio? (0.47uF approx = 6.8k, making the divider as 1:101?)

Have you wired L --- resistor --- cap --- N?

yep

In this case your calculation is correct for 50Hz.

[janoc]

Audii grade toroidal transformers have extremely low resistance, hence are easy to saturate, especially considering they have no air gap at all. Try to dump 1v dc into a well built toroidal xfmr and see how much current it will draw.

And that is relevant how? At worst the transformer will heat up more, because a saturated core has lower efficiency and you will not get the full rated power through it. However, with the practical DC offsets on mains you have no chance to saturate a transformer core of any reasonable mains transformer. And any heating effect will be totally dwarfed by things like mains fluctuations due to nearby appliances being turned on and off, noise from motors and such.

Dumping 1V DC into a transformer (any transformer, the fact that it is toroidal makes no difference) is supposed to test what? At DC the winding of the transformer is basically purely a wirewound resistor. So yes, of course, it will draw current - it is a solenoid!


The forum posts actually talk some nonsense about mains hum being audible in their amplifiers if powered from transformers having DC on their primaries. 

[3roomlab]

i ended up using a 0.22uF. so effective impedance should be 14.47k. making the divider 1:46.99.

this gives 5.02v AC measurement across the cap, which = 236v AC
and translating the 0.04v DC fluctuation into about 1.9v of DC ... which seems rather excessive 

[Seekonk]

The DC value is not multiplied.  That is the DC component though I am not sure how you measured it with that much AC.

About a year ago I did some experimenting with random triac firing into a transformer.  It was a small transformer only about 5W and was using  the 277V winding on 120V.  On the table I had a number of AAA batteries.  They began rolling around even though quite a distance from transformer.  It was like a magic act.

[f5r5e5d]

there seems to be some overeager "debunkers" here who lack a few bits of the picture - read the ESP article - Rod Elliot is generally on the "debunker" side of audio electronics discussion

DC on mains does happen - SCR controllers chopping the AC line directly as in some lighting, heating, cooling equipment can be an issue if you are in a large apartment building

the DC can cause problems, especially with toroids as already mentioned

if you don't have DC then a Blocker won't help - and yes, audiophiles may often misdiagnose the situation, even report subjective improvement where there was no measurable change


but dismissing everything just because some audiophiles sometime, somewhere got it wrong isn't a good reflection on your commitment to hard headed engineering

[3roomlab]

The DC value is not multiplied.  That is the DC component though I am not sure how you measured it with that much AC.

About a year ago I did some experimenting with random triac firing into a transformer.  It was a small transformer only about 5W and was using  the 277V winding on 120V.  On the table I had a number of AAA batteries.  They began rolling around even though quite a distance from transformer.  It was like a magic act.

that much AC? its normal
hmmm ok, i try another approach
since the fluctuation comes and goes. assuming they happen at 0.5Hz
it will make the 0.22uF look like a 1.45Mohm, which approximates to a ratio of 1:2
=0.06v DC  ... hmmm

i think i should log it someday, that leaves me back in square 1. must be lotsa ladies turning on the hairdryers ... making the DMM chasis vibrate

[grumpydoc]

Rod Elliott(ESP) did a fairly indepth write-up on the subject of blocking mains DC offset.

http://sound.westhost.com/articles/xfmr-dc.htm

Now that seems to be quite a reasonable article.

[fpliuzzi]

Rod Elliott(ESP) did a fairly indepth write-up on the subject of blocking mains DC offset.

http://sound.westhost.com/articles/xfmr-dc.htm

Now that seems to be quite a reasonable article.

I thought that Rod's article would be of interest here because he always supplies useful data to support whatever subject he's discussing. I've spent quite a few hours at his site over the years checking out his projects and articles.

[ConKbot]

there seems to be some overeager "debunkers" here who lack a few bits of the picture - read the ESP article - Rod Elliot is generally on the "debunker" side of audio electronics discussion

DC on mains does happen - SCR controllers chopping the AC line directly as in some lighting, heating, cooling equipment can be an issue if you are in a large apartment building

the DC can cause problems, especially with toroids as already mentioned

if you don't have DC then a Blocker won't help - and yes, audiophiles may often misdiagnose the situation, even report subjective improvement where there was no measurable change


but dismissing everything just because some audiophiles sometime, somewhere got it wrong isn't a good reflection on your commitment to hard headed engineering

https://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect runs rampant around here. Lots of people failing to realize that most audiophile shenanigans originated from something, and rather than trying to dig up fact and origin and maybe learn something about a new discipline or area of electronics, just like to blindly be like "lol audiophiles so dumb, we're so awesome" 

[macboy]

Audii grade toroidal transformers have extremely low resistance, hence are easy to saturate, especially considering they have no air gap at all. Try to dump 1v dc into a well built toroidal xfmr and see how much current it will draw.

And that is relevant how? At worst the transformer will heat up more, because a saturated core has lower efficiency and you will not get the full rated power through it. However, with the practical DC offsets on mains you have no chance to saturate a transformer core of any reasonable mains transformer. And any heating effect will be totally dwarfed by things like mains fluctuations due to nearby appliances being turned on and off, noise from motors and such.

Dumping 1V DC into a transformer (any transformer, the fact that it is toroidal makes no difference) is supposed to test what? At DC the winding of the transformer is basically purely a wirewound resistor. So yes, of course, it will draw current - it is a solenoid!


The forum posts actually talk some nonsense about mains hum being audible in their amplifiers if powered from transformers having DC on their primaries. 

You misunderstand; the hum mentioned is due to physical vibration of the transformer, not coming through the amplifier into the speakers. This vibration is reduced when the DC offset is removed.  Physical humming of transformers when DC offset is present is quite well known. It can be a serious problem for utility companies whose substations can hum loudly enough to generate complaints from neighbors. They can also run into saturation issues.

1 VDC with 120 VAC superimposed still has the same effect (same DC current generated) as 1 VDC without the AC.

Don't assume that a little DC on a transformer primary isn't an issue just because you hadn't thought about it or encountered it before. It is a real issue.

Start with this: http://www.ewh.ieee.org/r6/san_francisco/pes/pes_pdf/PGE_Presentation_TransformerHum.pdf

[janoc]

You misunderstand; the hum mentioned is due to physical vibration of the transformer, not coming through the amplifier into the speakers. This vibration is reduced when the DC offset is removed.  Physical humming of transformers when DC offset is present is quite well known. It can be a serious problem for utility companies whose substations can hum loudly enough to generate complaints from neighbors. They can also run into saturation issues.

1 VDC with 120 VAC superimposed still has the same effect (same DC current generated) as 1 VDC without the AC.

Don't assume that a little DC on a transformer primary isn't an issue just because you hadn't thought about it or encountered it before. It is a real issue.

Start with this: http://www.ewh.ieee.org/r6/san_francisco/pes/pes_pdf/PGE_Presentation_TransformerHum.pdf

OK, I didn't know about this issue. I am not a sparky or high energy guy. Interesting, thanks for the link.

On the other hand how much DC would you need to push a normal transformer in a consumer electronics product into a (partial) saturation? I would think that it would be very difficult to find so much DC available from a domestic outlet and the transformer would have to be nearing saturation already - that would mean operating that amplifier close to the maximum of the possible PSU current. Aren't transformers derated to have a reserve so that they are never being operated close to saturation?

Something doesn't compute here for me. 

[3roomlab]

that page 6 data, just 1A DC forces such a large transformer to give out 82dBA. that is loud

[Seekonk]

Look at the current waveform for that, the transformer is going into magnetic saturation.  So if >500W transformers have too little resistance then add some.  I see a half ohm series resistor with anti parallel diodes across it.  No noise for the quiet passages and the diodes conduct for the louder times.  If at 500W it is unlikely that you will hear the transformer.  In the circuits shown, extra resistance is probably all that is added.

It is a solution looking for a problem.  If this is to be taken seriously there has to be four times the capacitors shown. The capacitors are what makes the magic.

[3roomlab]

Look at the current waveform for that, the transformer is going into magnetic saturation.  So if >500W transformers have too little resistance then add some.  I see a half ohm series resistor with anti parallel diodes across it.  No noise for the quiet passages and the diodes conduct for the louder times.  If at 500W it is unlikely that you will hear the transformer.  In the circuits shown, extra resistance is probably all that is added.

It is a solution looking for a problem.  If this is to be taken seriously there has to be four times the capacitors shown. The capacitors are what makes the magic.

sorry i didnt catch, what does it mean add resistance?

found this nice video about GIC

[janoc]

that page 6 data, just 1A DC forces such a large transformer to give out 82dBA. that is loud

Yes and we are talking about an enormous substation transformer with oil cooling. Or you want to say that you have 1A DC available from your domestic outlet?  This doesn't scale linearly - i.e. 1A on big transformer = noise doesn't mean that a few microamps DC will cause comparable noise from a small transformer. The core is likely different, the loads on the transformers are very different, etc.

Careful with the analogies.

[The Magic Rabbit]

A well designed HiFi amp should be able to cope with a slight DC offset anyway. They're usually able to cope with a fairly large amount of crap in the AC supply. It's audiophoolery at work IMHO.

Sent from my XT1039 using Tapatalk

[Seekonk]

Even if you accept there is a DC offset I think the circuit only works as designed at low power levels.  And as I said before the added crap only serves to add more resistance which also solves the problem. More capacitors in parallel are needed to keep the sag  less than 100mv at peak current.

[3roomlab]

why bother with cap banks .... with element14 to the rescue ...
ONE CAPACITOR to RULE them ALL !!! 47000uF !    100v !!!

[Seekonk]

Now that is gettiNg your two cents worth!

[janoc]

Audio grade transformers (thick copper wires on toroidal cores) are VERY easy to saturate. Toroidal cores don't have air gap to store energy, while thick wires allow significant current at low DC voltage.

Um, what is an "audio grade" transformer? (as opposed to "non-audio grade").

If this is a real issue (I am not disputing the physics but the chance of this occurring in real installations), then I would say that power supply is poorly designed. One cannot expect squeaky clean AC from the mains.

I have worked in recording studios and with quite a bit of audio gear and DC offset on mains was never an issue anywhere (and some places had really nasty dirty AC!). We certainly didn't have any "DC blockers" anywhere.

Also, with a 500VA+ PSU I would probably expect it to be a well filtered, shielded switching PSU than a huge toroidal trafo these days, even in studio gear today, but I guess the affinity for linear supplies goes with the audiophile territory ...

[3roomlab]

comparatively

E-core, toroid, R-core, etc ... which are generally more saturation tolerant?