Discussion of AC mains isolation transformers

[jabbr]

You do not want a separate iso transformer on each amp, that makes the leakage loops through the power amps even worse, not better.

 

Not following this. Let's assume passive speakers (or include the amp + speaker behind a single iso) -- how does that make leakage currents worse?

[Jud]

Hi Gents:

 

Happy Friday. I'm taking it a bit easy today. Was going to go to Fresno to pick tile and counter laminate for the UpTone shop bathroom and kitchenette, but there is a storm rolling in and the big Wilsonart laminate store closes in a couple hours anyway. So I'm just catching up on e-mail and forum stuff--and again procrastinating my year-end accounting.

 

Right now there seems to be a particularly nice selection of Topaz iso trans units on eBay, at surprisingly good prices:

http://www.ebay.com/sch/i.html?_from=R40&_trksid=p2050601.m570.l1313.TR0.TRC0.H0.Xtopaz+transformer.TRS0&_nkw=topaz+transformer&_sacat=0

 

A lot of the eBay sellers fail to indicate the kVA size of the units they are selling, if they were originally line-cord or terminal models, or the input/output voltage options. But they always list--or you can read from the listing photos--the model number.

So here is a link to the only model/spec sheet I have found to-date:

http://www.pacificparts.com/vends/mge/images/t1.pdf It is from MGE UPS Systems, the last firm to own/sell the line before Eaton bought them. I have seen other Topaz models outside of what is on the list, but it is a pretty good sheet for sizing what you are looking at.

 

Also, I recently discovered that Daitron--a Japanese conglomerate that we have known for a while for their ultra-low-noise SMPS units--has for many years (decades even?) been selling isolation transformers that are exactly the Topaz Ultra Isolators.

See: DaitronPower

 

Look at the photos of this one on eBay:

Daitron J 91005-11 Line Noise Suppressing Ultra-Isolator .005pF

 

Plus, on the current Daitron spec sheet, they are listed with interwinding capacitance of 0.0005pf, just like the best of the original Topaz units.

 

Given that they clearly are using the same end castings as the original, and that Daitron generally does its own manufacturing, I am wondering if at some point in the long and twisted history of the Topaz iso trans line (see this post) maybe Diatron bought some rights and tooling.

 

Myself, I am considering a big 5kVA unit to wire up for 240V in/out. My custom studio has a dedicated sub-panel--fed from our main panel by 25 feet of big 3/0 (all cryro-treated, including the bug panel and all home-run runs of Hovland Main Line to the outlets). As part of my new building addition (for the UPF--UpTone Production Facility ), I formed a new utility room (for new house HVAC, water, electric, and future solar control). So that is the perfect place for me to install a big humming iso trans to feed the studio sub-panel!

 

Here are some pics:

 

[ATTACH=CONFIG]31400[/ATTACH]

 

The sub-panel from back when I built the studio in 2004.

[ATTACH=CONFIG]31401[/ATTACH]

 

To prove that I'm crazy, a shot of the room between the 2 layers of 5/8" drywall. 760 4-inch squares (that I hand cut) of PolyDamp ADA viscoelastic constrained-layer damping material (really works!).

[ATTACH=CONFIG]31402[/ATTACH]

 

Lastly, John is coming to visit (work/play) in January, so I just added this very firm queen-size sleeper-sofa to the UPF. No more airbed in the studio! (Though with 2 of our 3 kids off at college, we do also have a regular guest room now.)

[ATTACH=CONFIG]31403[/ATTACH]

 

HAVE A GREAT WEEKEND EVERYONE!

 

--Alex C.

 

Alex, your utility room has a giant pink alien worm in it.

 

When John comes, better feed him to it to keep it friendly.

 

 

Sent from my iPhone using Computer Audiophile

[Superdad]

Alex, your utility room has a giant pink alien worm in it.

 

Yes, it is over 9 feet tall. But far friendlier (energy-wise) than the nasty (and I mean rodent-nasty) 30-year old outdoor packaged AC/heat unit it replaced. The old beast nearly killed me a couple of times when gas built up then ignighted--blasting the steel cover panel 30-feet in my direction!

[One and a half]

It is probably best to have the iso transformer at the back end, but the difference is probably pretty small, if it is easy to put it at the front end of a dedicated line and hard to put it at the back end, then go ahead and put it at the front end.

 

John S.

 

Out of choice, I needed to install the 2kVA ISO at the panel downstairs at the source cause lugging a 26kg piece of iron upstairs wasn't going to happen... One other advantage I can think of is the common mode rejection works with the length of wire to make common mode noise that little bit extra difficult to transmit.

 

I still need to install the RC filter on the secondary of the ISO and also the balanced transformer. Now currently looking at differential mode protection on the primary of the ISO, seems to be a lot of MOVs and Gas discharge devices on commercial equipment. A voltage protection relay might be be quick enough to detect a fast rising voltage in time to protect the load.

[JohnSwenson]

Not following this. Let's assume passive speakers (or include the amp + speaker behind a single iso) -- how does that make leakage currents worse?

 

My understanding of he topology in question is all the source including preamp is on one strip plugged into one iso transformer. Then each power amp would be on its own iso transformer.

 

The leakage loop from poweramp PS to preamp PS has to go through two iso transformers each of which adds impedance to the loop increasing the voltage developed around the loop. In addition there is a loop from poweramp PS to the other poweramp PS which goes through the interconnects connected back at the preamp.

 

John S.

[asdf1000]

Hi John and all

 

I was just reading this comment over on the PS Audio Forum:

 

"I don’t “hate” isolation transformers, but I surely don’t like them nor do I support their use. They typically make things worse. Adding these products in series with the AC line does exactly the opposite of what we hope to do: lower output impedance of the AC line and regulate the power. Yes, the isolate, but that’s not important sonically. What IS important is the aforementioned impedance drop. If you place a large lower power factor device on line—say it has a PF of 0.7 as most amplifiers and equipment do—and then draw high peak current, you will cause a divot to be sucked out of the incoming AC. This is because the resistance (impedance) of the wires connecting your home to the power transformer on the utility pole are relatively high. It’s simply Ohm’s law. Draw too much current through a resistor and you get a voltage drop. We see it all the time (which is why we built an oscilloscope on the Power Plant’s front panel) and want people to understand what is happening. When you then insert an isolation transformer into this soup, the voltage loss can only get worse. Flat topping and voltage divots can’t get better with an isolation transformer, only worse. They store no energy and cannot make up for losses. Only a regenerator can."

 

Just sticking to the technical reasoning involved (and leaving out any possible commercial interests and factors in the above comment) what are your thoughts?

 

The link to the post here: In Praise of Isolation Transformers | Power products | ForumsPS Audio

 

Cheers

[jabbr]

The leakage loop from poweramp PS to preamp PS has to go through two iso transformers each of which adds impedance to the loop increasing the voltage developed around the loop. In addition there is a loop from poweramp PS to the other poweramp PS which goes through the interconnects connected back at the preamp.

 

Ok, perhaps I'm really confused, but isn't the point behind "breaking" the leakage loop to increase the impedance i.e. infinite impedance would mean no loop??

 

What fixes the current? Otherwise why does the voltage increase? I think I must be missing something very basic...

[Superdad]

Ok, perhaps I'm really confused, but isn't the point behind "breaking" the leakage loop to increase the impedance i.e. infinite impedance would mean no loop??

 

What fixes the current? Otherwise why does the voltage increase? I think I must be missing something very basic...

 

No offense jabbr, but you really seem to be going round and round with John on this topic. As he detailed quite completely here, his purpose in suggesting using an isolation trans is not to block leakage currents, but to give some quality noise and surge protection so that people can then safely plug all their gear into one filter-element-free heavy power strip or outlet box, thus reducing the impedance between component power supplies and reducing the problems caused by the unavoidable leakage currents between them.

 

And since you were also talking about medical usage and how iso trans units are used to block leakage to humans, he further clarified his points in this post.

 

Rereading those posts may help clear up your confusion. Or you may then have a more focused question. I know you are really smart about this stuff, so it may be that you are ascribing more to John's suggested use of iso trans than what he intends.

 

It's funny, I've had two 250kVA custom 1:1 trans (with floated secondary) in my system for over 10 years (the old Hovland VoltAire units), but it has become clear to me just this year that the positive action they are having is not from the mechanism I always thought.

 

Ciao,

 

--Alex C.

[YashN]

I was just reading this comment over on the PS Audio Forum:

The link to the post here: In Praise of Isolation Transformers | Power products | ForumsPS Audio

From the Ted Smith post at that thread, the info on PurePower's site is quite interesting. Hadn't heard of them despite being local.

[YashN]

Lastly, John is coming to visit (work/play) in January, so I just added this very firm queen-size sleeper-sofa to the UPF. No more airbed in the studio!

Sounds like a ploy to keep John indoors so that he can get some new technology from his mind into a worthy audiophile contraption.

 

Good plan.

[Middy]

That make of sofa bed is really good..

It comes with a multimeter and solder station...

 

 

Sent from my SM-G900F using Computer Audiophile mobile app

[One and a half]

Hi John and all

 

I was just reading this comment over on the PS Audio Forum:

 

"I don’t “hate” isolation transformers, but I surely don’t like them nor do I support their use. They typically make things worse. Adding these products in series with the AC line does exactly the opposite of what we hope to do: lower output impedance of the AC line and regulate the power. Yes, the isolate, but that’s not important sonically. What IS important is the aforementioned impedance drop. If you place a large lower power factor device on line—say it has a PF of 0.7 as most amplifiers and equipment do—and then draw high peak current, you will cause a divot to be sucked out of the incoming AC. This is because the resistance (impedance) of the wires connecting your home to the power transformer on the utility pole are relatively high. It’s simply Ohm’s law. Draw too much current through a resistor and you get a voltage drop. We see it all the time (which is why we built an oscilloscope on the Power Plant’s front panel) and want people to understand what is happening. When you then insert an isolation transformer into this soup, the voltage loss can only get worse. Flat topping and voltage divots can’t get better with an isolation transformer, only worse. They store no energy and cannot make up for losses. Only a regenerator can."

 

Just sticking to the technical reasoning involved (and leaving out any possible commercial interests and factors in the above comment) what are your thoughts?

 

The link to the post here: In Praise of Isolation Transformers | Power products | ForumsPS Audio

 

Cheers

 

Where do you start with this prose.

 

Impedance drops. *YES* there is an impedance drop if your 400W amplifiers are connected to a 100VA Isolated transformer. If an Isolation transformer is sized correctly and has better than 2% regulation, the rest of the peak current draw reasoning is out the window.

 

Voltage divots? This is gross harmonic distortion, certainly not from a utility supplying power to a house or apartment building, there are limits for these distortions. The relatively high impedance from the street supply..... put a dead short on that, and it's got strength behind it. The reason the fault level is low (impedance is 'high'), is because a) there's no need to domestic situations to have a high fault level, b) high fault level breakers are super expensive, c) which leads to expensive protection mitigation systems also costing thousands.

 

An isolation transformer never claims to regulate power, so what is the point mentioning?

 

That the ISO transformer is a naturally high surge suppression device and common mode noise killer is not mentioned. Any inverter is a leakage current master, the PSAUDIO Regenerator is an inverter, so draw your own conclusions.

 

In fact, an isolation transformer is often included on UPS internally to keep the common mode noise in check as well as providing a smooth sine wave output.

[One and a half]

From the Ted Smith post at that thread, the info on PurePower's site is quite interesting. Hadn't heard of them despite being local.

 

PurePower topology is a double conversion inverter with or without the battery. There's dozens of UPS manufacturers with 5 or 6 nines reliability that churn out the same if not better specs.

[jabbr]

No offense jabbr, but you really seem to be going round and round with John on this topic. As he detailed quite completely here, his purpose in suggesting using an isolation trans is not to block leakage currents, but to give some quality noise and surge protection so that people can then safely plug all their gear into one filter-element-free heavy power strip or outlet box, thus reducing the impedance between component power supplies and reducing the problems caused by the unavoidable leakage currents between them.

And since you were also talking about medical usage and how iso trans units are used to block leakage to humans, he further clarified his points in this post.

 

Well I remain confused about these issues despite the attempt to clarify. From @JohnSwenson 's clarification:

What I am talking about is very low impedances, small voltages and small currents. It has nothing to do with earth ground, safety ground etc. It is formed by capacitances in the PS between AC line and DC output generating a voltage between AC line and DC output. When a low impedance connection is formed between the DC output of one PS and the DC output of another PS these voltages form a current flow. Since the loop (I'm calling this a leakage loop) goes through TWO power supplies the waveform of this leakage current depends on the characteristics of both supplies. Thus it is hard to define the characteristics of a single supply, unless you define a "standard" supply to reference all others to.

 

I'm having difficulty understanding how, if the power supplies for the amps use low-leakage transformers, of if preceded by a low interwinding capacitance transformer, how can a "low impendance" loop exist between two amps? Even if from the pre-amp out along the interconnect, what is the return loop?

 

Specifically I am asking in the situation where we have a very low capacitance between the AC and the DC. Bill Whitlock has been making the point of the importance of minimizing so-called "parasitic capacitance" ** for decades e.g. http://web.mit.edu/jhawk/tmp/p/EST016_Ground_Loops_handout.pdf ** he uses the term "parasitic capacitance" synonymous with "interwinding capacitance"

 

Rereading those posts may help clear up your confusion. Or you may then have a more focused question. I know you are really smart about this stuff, so it may be that you are ascribing more to John's suggested use of iso trans than what he intends.

 

Perhaps I'm not being clear: I am not at all concerned about the term "iso", I am trying to understand the basic electronics behind this. I would understand this better with a schematic than descriptive prose such as "iso" and "leakage". Perhaps that isn't appropriate in this vendor subforum. I thought the point of breaking off this thread from the lps-1 was to allow some non-product based discussion but again as we are still in the Uptone sponsored area I'll back off pushing this issue. I'm not trying to be argumentative, really just trying to understand, and I might not be so smart, but asking for a schematic might understandably be too much.

[Superdad]

I'm having difficulty understanding how, if the power supplies for the amps use low-leakage transformers, of if preceded by a low interwinding capacitance transformer, how can a "low impendance" loop exist between two amps? Even if from the pre-amp out along the interconnect, what is the return loop?

 

Specifically I am asking in the situation where we have a very low capacitance between the AC and the DC.

 

Good question and I'm sure John will attempt to answer.

 

Perhaps that isn't appropriate in this vendor subforum. I thought the point of breaking off this thread from the lps-1 was to allow some non-product based discussion but again as we are still in the Uptone sponsored area I'll back off pushing this issue. I'm not trying to be argumentative, really just trying to understand, and I might not be so smart, but asking for a schematic might understandably be too much.

 

Don't worry at all! Your questions are COMPLETELY appropriate for this thread. We are all here to learn.

[joelha]

I finally hooked up my power amps to my Topaz iso unit and then I could hear the hum more clearly. Obviously a bigger load on the iso device is causing the louder hum.

 

My question is, as I'm using a 1KVa unit, will substituting a larger unit, say 2KVa, create more of a hum because of it's greater capacity or less of a hum because it will not be taxed as much by my equipment?

 

Joel

[Middy]

http://www.computeraudiophile.com/showthread.php?t=24916

AC Filtering, Grounding Boxes, Linear PSU and Balanced Power.

 

If John is kind enough to answer this and Joelha.....

 

 

This is half the question I wanted to know with transformer hum....

I know the different plates can vibrate more after time and are bonded with conformal coating/ Varnish.. But like Joelha issue is it load or dc offset causing the lions share...

 

Just trying to figure out stuff.... Sorry for the bother...

 

Sent from my SM-G900F using Computer Audiophile mobile app

[jabbr]

Haha I asked for a schematic, so here's one illustrating my point:

 

Admittedly this schematic is simple but suppose we consider the "leakage" loop between only the amp PSUs ... that's even simpler.

[jabbr]

...people can then safely plug all their gear into one filter-element-free heavy power strip or outlet box, thus reducing the impedance between component power supplies and reducing the problems caused by the unavoidable leakage currents between them.

 

Superdad, consider this: the LPS-1 has effectively infinite impedance between power supplies! So simply reducing impedance is not necessarily a goal, right?

 

Also aren't any low frequency currents which might be present on the AC mains filtered out by the hopefully massive PSU capacitors in a good amp?

 

Now, my impression is that the impedance we do want to reduce is that of the circuit ground.

[One and a half]

Low impedance is not 'low' enough, it's larger than zero being the ideal, but can never get there.

 

What is desirable is to attract leakage currents caused by conducted noise away from its source, typically if there is a low impedance, then this noise can be attracted to it.

A low impedance is a null on the circuit and can be a filter. Since conducted noise is wideband 100kHz to 1-2 MHz, it's not easy to design such a filter, complicated by the construction of the filter itself which may cause resonance in the circuit far worse than the noise.

 

Having all the AC power components plugged into the one power strip keeps the voltages at one source potential.

[One and a half]

Transformers hum and can rattle. They can also play tunes of different frequencies. If you hear tunes on the audio systems iso transformer, there's real problems.

Rattle is caused by loose laminations caused by hum.

 

Hum will go away if the Transformer has the right load on it, say 80%. Unloaded transformers rattle more.

 

For balanced transformer, rattle occurs on the primary side due to the transfer of the secondaries harmonics which causes lumps of current to develop. I use an iso transformer to smooth out the noise and this is effective. This practice is recommended by Equi-tech in their manual for the model Q.

[YashN]

I am trying to understand the basic electronics behind this. I would understand this better with a schematic than descriptive prose such as "iso" and "leakage". Perhaps that isn't appropriate in this vendor subforum. I thought the point of breaking off this thread from the lps-1 was to allow some non-product based discussion

 

I want to understand it too, so perhaps the better venue for discussing it is the AC Filter, etc. thread where you can discuss it unfettered. For instance, I want to understand whether the breakage of that 'leakage current' also prevents the original issue of 'hash injection' from interfering with SQ with other equipment, asked a question but now I can't even find my post. I thought it was here, but with all the moving of posts and so on, I can't even follow a discussion.

 

Something been on my mind about the difficulty in simulating it, so a post is forthcoming on that but over there.

[joelha]

Transformers hum and can rattle. They can also play tunes of different frequencies. If you hear tunes on the audio systems iso transformer, there's real problems.

Rattle is caused by loose laminations caused by hum.

 

Hum will go away if the Transformer has the right load on it, say 80%. Unloaded transformers rattle more.

 

For balanced transformer, rattle occurs on the primary side due to the transfer of the secondaries harmonics which causes lumps of current to develop. I use an iso transformer to smooth out the noise and this is effective. This practice is recommended by Equi-tech in their manual for the model Q.

 

Assuming this was in response to my question about an increased level of hum on my Topaz, the hum I'm hearing is not being heard through the music. The hum is being heard only from the Topaz. And the hum was virtually inaudible before I hooked my power amps to it. Then the hum became much more prominent.

 

Just out of curiosity, what happens if an iso unit is near or even exceeds its limit? Will it overheat, kick off, something else?

 

Joel

[Middy]

My question sort of was to see if a DC BIAS/ OFFSET FILTER WOULD BENEFIT THE TOPAZ OWNERS. THAT THEY HAVE THE SAME ISSUE AS TORIODAL LIKE MINE JOELHA? ? Apart from load hum increases filtering.....

 

Sorry my phone just loaded upper case not retyping again...stupid samsung...

 

Sent from my SM-G900F using Computer Audiophile mobile app

[Cornan]

My question sort of was to see if a DC BIAS/ OFFSET FILTER WOULD BENEFIT THE TOPAZ OWNERS. THAT THEY HAVE THE SAME ISSUE AS TORIODAL LIKE MINE JOELHA? ? Apart from load hum increases filtering.....

 

Sorry my phone just loaded upper case not retyping again...stupid samsung...

 

Sent from my SM-G900F using Computer Audiophile mobile app

AFAIK all types of Isolation transformers are sensitive to DC currents in the AC mains. If you experiance audioble hum you will probably gain using a DC blocker. With that said I have ordered one as you know (still waiting for delivery notice) even though my Peaktech 2240 is as good as silent...meaning that in a totally silent room I can barely hear it at a 20 cm distance. All my DC devices in the audio chain are powered with battery supplies disconnected from the AC mains...so it will be very interesting to test if it will still make a difference. I'll keep you posted when I have got it & tried it with my tech type IT!

 

Also still waiting for %! hifi grade fuse for my Peaktech...but just got a delivery notice...and will hopefully see if it improves thing in a couple of days.

 

Sent from my Nexus 7 using Computer Audiophile mobile app