Discussion of AC mains isolation transformers

[One and a half]

Another method for checking the phasing is with a battery,  an AA is fine.

 

Allow each winding to be free. Connect the + of the battery to H1. Negative to H2. Measure across H1 with the red lead, and H2 with the black lead and keep the meter in place and have a fixed value on the meter,  say 6V is ok. 

 

Now you will measure 1.4V or a lot less across the winding, allowing for a drop of voltage.

Watch the meter, then let go of the battery. There's a collapse of the field from 1.5V to 0, but the windings will kick back a voltage in the same polarity.

So the voltage will increase positive, which means the winding is phased correctly. 

 

If the meter goes negative, then the winding is azz backwards. This happens very quickly,  often digital meters are too slow and you won't see it. Some meters have a bar graph which is very fast.

The old analog dial meter is perfect for this work.

 

Just to make sure there's no overload, List all the components in the system and tell us the values from the name plates . Just Amp= 300W, CD = 30W, you don't have to give names.

 

Add up all the watts, then divide by 0.6 and that will be your loading in VA that the Transformer needs to be sized at. 

[One and a half]

34 minutes ago, spacexpert said:

Thanks for the tip.  But as soon as I let go of the battery the numbers quickly jump all over the place.  Occasionally I get a negative reading, but it happens randomly on both sides.  I might need more voltage and a better multimeter.  I'll give it another shot tomorrow. 

 

In the mean time here's what it's powering out of a filter-less distribution block.  I'm certain I'm well under for a 750VA.

 

LPS-1 stock energizer

2x B&W ASW608 Subwoofers (listed power consumption 40W each) 

Lavardin IT 55W

Hugo2 (5V iPower)

2x 65W R-Core LPS's (PC+Hard Drive)

 

That's 295W / 0.6 = 491VA, it's the limit, all right.

As Cornan suggested, take off the subs and see what happens. 

 

Still need to make sure the Topaz is correctly setup. 

[One and a half]

7 hours ago, zilch0md said:

 

Which drawing (or flyer) is incorrect?  

 

I haven't seen anything wiring diagrams for the Topaz -31 models that show this incorrect sequence on the H terminals.  (Or maybe I have and didn't notice.)  In any case, I would like to know which resources is in error.

 

Thanks!

 

Mike

Read more  

Sorry my eyes  This post has the correct sequence too the scan is sol old and the terminals are pretty fuzzy to see.

[One and a half]

EEEUWW the CA editor went downhill there.

 

Try : My eyes play up couldn't see the markings that well. The pdf linked in this post is correct. 

[One and a half]

8 minutes ago, spacexpert said:

I plugged in a power meter, it's just a cheap one, but never the less it reads as follows:

Power factor 65 and wattage peaks at 79W but usually hovers around the high 60's.

That's with music playing at my listening level.

 

Regarding the Topaz direction, I grabbed a 24VAC wall wart with bare cable ends and applied it to the H1 + H2 and then the X1 + X2.

Perhaps I need a better multimeter?  As soon as I remove the wall wart leads, the numbers quickly drop, sometimes they spike and I sometimes see a negative symbol, but it happens on both sides.  It's just too random.   

 

I might take it to someone with a higher quality multimeter, see if they can get a better reading?

 

Silly question, the meter is set to AC volts?

[One and a half]

10 minutes ago, spacexpert said:

V with the squiggly line.  I was getting 26.4 on the meter. It's a really cheap mutimeter though, like a $15 job  

It's possible the wall wart is not powerful enough to drive the transformer and the meter can't think too hard. The bar graph meters are upwards of $600, so lets look again at the connections. The load doesn't seem that large, but hang on to the load meter for now. It will come in handy later.

Wire up the transformer to the drawing in this post, and then :

 

Look at the power consumption on the primary of the transformer with no load, then add the audio load on and spot the difference. You don't have a thermometer at all to record the temperature of the case before and after?

Maybe there's an app for that? Although you wouldn't want to melt your phone.

 

One thing I did spot before on the photo, is the earth lead is connected to a case with a spade lug. Slap hand Mr. Topaz. Replace this with a ring crimplug and a star washer, if that screw is loose, at least you have some earth to connect to rather than be completely off.

 

 

Primary Secondary - Final wiring.pdf

[One and a half]

38 minutes ago, spacexpert said:

I plugged in a power meter, it's just a cheap one, but never the less it reads as follows:

Power factor 65 and wattage peaks at 79W but usually hovers around the high 60's.

That's with music playing at my listening level.

 

Regarding the Topaz direction, I grabbed a 24VAC wall wart with bare cable ends and applied it to the H1 + H2 and then the X1 + X2.

Perhaps I need a better multimeter?  As soon as I remove the wall wart leads, the numbers quickly drop, sometimes they spike and I sometimes see a negative symbol, but it happens on both sides.  It's just too random.   

 

I might take it to someone with a higher quality multimeter, see if they can get a better reading?

 

But when you measure, the voltage is steady, or does it drop all over the place?

[One and a half]

9 minutes ago, spacexpert said:

The voltage is steady, as long as I keep my shaky hands still.

 

I noticed those spade lugs on the earth too, I figured they could pull off.  I'll definitely replace those, thanks for the tip.

 

In the mean time I'll wire this thing up and do some load/no load measurements on both sides.

 

 

 

Make the 24V firm so you don't need to keep holding on, and use hands for measuring as per table above

[One and a half]

44 minutes ago, spacexpert said:

(Primary as X1 X3 X2 X4)  228V output on secondary

No Load        21PF    36W <> 38W  
Load              43PF    78W <> 90W  (averages about 80)

 

(Primary as H1 H3 H2 H4)  246V output on secondary 

 

No Load        22PF    44W <> 45W   
Load              43PF    86W <> 98W   (averages high 80's)

 

 

OK, well done!

 

The Primary is wired properly will always draw the most power, since the efficiency of the transformer is typically 95% (and less) for the Topaz.

 

So this scenario would confirm the Primary is on the H's and makes sense the output voltage is higher, especially on no load.

The pf is pretty bad, but there's no load on and the load that is on is all reactive. This reactive energy does no work at all,  but create heat. What happens is real Watts are needed to create reactive power. Reactive power can be controlled with a capacitor, but where to install, then there's a chance it goes the other way. Another topic.

 

(Primary as H1 H3 H2 H4)  246V output on secondary 

 

No Load        22PF    44W <> 45W   
Load              43PF    86W <> 98W   (averages high 80's)

 

I have to ask, how's the temperature?

[One and a half]

25 minutes ago, spacexpert said:

Fantastic, you're a Godsend.  So the outside labels were right all along.

 

I also had the live and neutral mixed up when I first ran the Topaz.  The above measurements were done with the live on the correct side.

I'm hopping that with the correct direction and phase, that it doesn't heat up too much.

I'll hook it up now and give it a few hours.  I'll report back with some temps.

 

Cheers One and a half. ?  

Let's wait a few hours, some Topaz units just run hot though. Swapping the input phase, is probably important for this type of transformer with its odd magnetic structure. At least there's output now. Did I tell you you need a earth leakage circuit breaker on the output of the Topaz? Details later.

[One and a half]

2 minutes ago, spacexpert said:

Oh boy.  Earth leakage circuit breaker?   Well I'll be careful around this thing in the meantime.

 

I have strain relief cable glands ready to go on it, and I'll need to improve the earth connection.  I look forward to learning about this earth leakage circuit breaker.

 

Currently it's running lukewarm.  Fingers crossed. 

OK I be online for a few hours yet. Lukewarm is good.  BTW has the snap crackle and pop stopped coming out of the speakers?

[One and a half]

2 minutes ago, Cornan said:

 

It is enough with a cheap (USD20-30) portable RCD like this one on the output of the isolation transformer.

 

https://www.testandtagsupplies.com.au/images/PDL951_LRG.jpg

 

Good idea to be carful!

 

Great to hear that it seems to be resolved. A golden star to @One and a half for giving such great support!

 

 

Oh I wish it were that simple. The balanced output has no neutral, so both conductors are live to earth. The ELCB relay needs to be a 2 Pole device. Am thinking to write a separate post for this.

[One and a half]

1 hour ago, Cornan said:

Sorry for the off-topic, but I just want to add a quick word about good RCA cables. IME QED is the manufactorer with the best knowledge about coaxial cables. If you have a look at these pictures you can clearly see how a good constructed coaxial cable should look like. Especially have a look at the shielding and the center core. 

 

Looking at the plugs on QED Performance Audio 40 and Reference Audio 40 you can clearly see how a good RCA plug should look like. Better plugs have less shield contact inside the plugs. Also note the hollow center pin.

 

 

 

Another cable thread about to commence?!?! Cornan,  there's enough interest in star quad to start its own thread, mansr can make usual comments how about? 

[One and a half]

1 hour ago, Cornan said:

 

Oh, that´s good to know! I thought it was the same thing as for isolation transformers with floating secondary! I have a balanced isolation transformer on order that can easily be set to floating or grounded secondary, so I will check your post about this with great interest. Please note that personally I have a central GFCI/RCD installed in my appartment.  

 

I guess it means that it has to be something like this portable RCD then?

 

http://uk.rs-online.com/web/p/mains-rcd-connectors/0385042/

Yes something like that, did you see the price,  usually triple,  anyway more to it to come.  Also the post will cover protection for the Transformer  in case of overloads and safely shut the power off if it does. Optional overvoltage / undervoltage. The Topaz won't be around forever, best to protect them.

[One and a half]

7 hours ago, spacexpert said:

I think so too.  I will have to buy another one, shipping is a nightmare.    

The Hammond guide posted is a good link. The hotter than Melania transformer is producing heat for the wrong reasons.  Perhaps a winding is shorted internally through old age or a drop to the floor years ago. This means dismantling and costly repair.

 

What really needs is the Topaz to be Insulation Resistance tested at 500V for one minute, winding impedance test, plus a current measurement taken with a scope on the voltage to determine distortions.  That's about two hours work for technician and in my country that cost will pay for the freight on a new transformer. 

 

While it's there, leave it on and have it power the system and listen for clicks as well as music. If you don't hear any clicks,  then the Isolation transformer concept works.  I would not leave it on overnight and have a ceramic tile or brick underneath it to insulate from the floor or carpet. 

 

I'm sorry to say and given the effort to test for the obvious,  the Transformer 's heat will just lead to more problems. 

 

The search would be on to find another one. I'm looking for a 2-3 KVA unit and you're welcome to engage in trading on my 1KVA Topaz. It's sad that we have to buy old gear, let's stir a manufacturer to come up with something similar with a warranty.

[One and a half]

13 hours ago, unbalanced output said:

However, dynamic response is quite a concern - Naim recommends to be very careful when putting anything upstream of their power supplies. 

 

This is a myth and I'm not criticsising you, one of those stories that hasn't a foundation.. A Topaz for example has worst case 5% impedance typically 3%. It's short circuit current is (100/5%) x Full load amps of the transformer.

 

Example 2000VA ISO transformer at 120V AC has full load capability of 16.6A. If a dynamic is so strong, that it must demand 20 x 16.6A which is a dead short on the transformer, that's 333A. The transformer can't supply that current for too long, like minutes, but a 60 ms transient....walk in the park. The volts may sag, even if it went to 10%, that's 333A x 12V = 4,000W. Most if not all output transistors will fry first, most likely due to clipping.

 

Rule of thumb : Load watts / 0.6 = VA of the transformer.
 

[One and a half]

1 minute ago, frederick184 said:

I just picked up a Topaz 91095-32 500 Va isolation transformer on eBay.  My system is split into two parts.  My Comcast modem, wireless router, dedicated music PC, NAS, fiber switch and FMC's are all in an upstairs room.  The FMC's are connected to the NAS and the wireless router, and the PC has a fiber NIC connected to the fiber switch.  All devices are plugged into a single Wiremold power strip with no filtering.

 

Downstairs there is another wireless router connected to the master one upstairs in a mesh network.  An Ethernet cable connects to my Ultra Rendu powered by LPS-1.  This in turn is plugged either into my Oppo UDP-205 players USB DAC by a Lush cable or to my LH Labs Infinity X DAC.  The signal goes from there to my Parasound A21 amp and then to my Magnepan MG 3.7 speakers.  All devices are plugged into an Audience AR6-PDC with no filtering.  This part of the system is fed by a dedicated 20 amp circuit.  The upstairs part of the system is not.

 

I intended to buy an isolation transformer for both, but I'm wondering in the meantime which part of the system would benefit from it most?

 

Colin

The Topaz is primarily a protection device. Since it's possible to hear clicks and plops, the audio system has priority for the Topaz to be fitted. 

[One and a half]

28 minutes ago, spacexpert said:

Just been checking out other IT options in the hope of finding one that will run cooler.

I note that some have a sticker on the black part that reads "caution hot surface, avoid contact" 

 

So I guess these things can get hot enough to warn you not to touch.  That's pretty hot.

 

It wouldn't by chance be a combination of 240V + high capacitance .0005 that is a factor?

Regardless I guess it should run cool with low load...    

Hmm. How hot is hot? The Topaz can work from 0-50C ambient temperature.  It could have Class F Winding insulation, which can handle 130 C. @zilch0md quoted 40.6C at an ambient of 25C  his transformer wasn't fully loaded....that's 15.6 C above ambient. Perhaps double that for full load, that 80C case temperature, still OK for Class F. 80 C is really hot, you don't get even near it!

 

Class A insulation has a recommended temperature limit of 105 Degrees C or 221 Degrees F. 

Class B goes to 130 Degrees C or 226 Degrees F, 

Class F to 155 Degrees C or 311 Degrees F, and 

Class H to 180 Degrees C or 356 Degrees F.

 

The 0.00005 pf is an ability to block high frequencies, the lower the pf, the higher the opposition to that noise creeping in.

 

I hope to publish some figures on the 1000VA unit, surface temperature of the laminations and loading this weekend.

[One and a half]

Temperature and loading results 1000VA 0.001pf 91001-21 transformer.

 

Ambient : 22.7C 

Idle :

Consists of energising 1000VA Equitech transformer and MPD-3 DAC which is on 24/7 . Primary current 0.47A, 239V input 112.33VA.

Lamination temperature : 37.5 C

 

On Load : Ssytem 1 and System 2 on, low level music, maybe 1 Watt output

1.43A, 239V, 342VA input. 

 

Left on load for an hour, Lamination temperature : 38.3 C, Ambient Temperature 24.4 C

 

Temperature readings with a contact K type thermocouple. 

[One and a half]

4 hours ago, spacexpert said:

I've noticed that at certain times of the day my amp hums quite a bit, then other times it is quiet.  Is that related to DC offset?

Does the hum change pitch?

[One and a half]

1 minute ago, spacexpert said:

No, it is a steady pitch coming from the toroidal.  Interestingly, when it hums, if I put my ear to my speakers I hear the same sound through them.  It might last for 30 minutes then die down for several hours.  

I don't hear it from my listening position, so it doesn't bother me.

 

In the mean time my best offer was accepted for another Topaz.  Same spec as the last one.  So... ?    

Lots of Luck!

[One and a half]

1 hour ago, spacexpert said:

No, it is a steady pitch coming from the toroidal.  Interestingly, when it hums, if I put my ear to my speakers I hear the same sound through them.  It might last for 30 minutes then die down for several hours.  

I don't hear it from my listening position, so it doesn't bother me.

 

 

There's a ground loop there trying to break through..... quite possibly a loose connector in the signal chain somewhere. Any clicks?

[One and a half]

1 minute ago, spacexpert said:

 

Another weird thing about the Lavardin amp... it likes the hot wire on the wrong side of the IEC, so I have to make my own power cords for it.  Seems strange, but I do what they say. ?  

With the wiring of the Topaz in balanced mode, there's no wrong side, both conductors at the amp's plug are live to earth. 

[One and a half]

2 hours ago, Cornan said:

 

It is probably correct to use another type of power cable on high power devices such as powerful amps, powerful active speakers, powerful subs and powerful integrated amps. These devices might need thick gauge wires that a starquad simply is not suitable for due to its construction. But I think we are talking about really high current devices here.

 

What's the voltage rating on the star quad?

[One and a half]

1 hour ago, zilch0md said:

 

My 500VA Topaz 19095-32 is also pretty close to silent.  I have to put my ear close to it to hear anything.  It is not a toroidal transformer, though.  Then again, there are some people out there with noisy Topaz transformers (i.e. John S.) 

 

I really think the quality of the incoming AC mains power is a big factor, no matter what type of transformer you are using - and that incoming quality can fluctuate throughout the day.  

 

There are are a lot of people out there who have silenced noisy transformers that are inside their various audio components, just by inserting a DC blocker ahead of the their A/V systems - here in the States, too, not just elsewhere.  See the reviews of the relatively affordable Emotiva CMX-2, which is more than just a DC blocker.  It's the people who had noisy transformers who are writing the most positive reviews of the CMX-2 - so, I'm pretty sure it does a good job of DC blocking, even if it only does a fair job of CM or DM noise reduction.

 

If I had a noisy transformer, even a noisy Topaz, I would try putting the CMX-2 in front of it.  I definitely would not put anything like this (containing MOVs for surge protection) in between my Topaz and the power strip into which all my gear is plugged.  Surge protectors, if really needed, should be placed ahead of the Topaz - which is exactly where you want a DC blocker, too.

The Topaz/Elgar/Xentek is an effective surge suppression in itself. Absorbs a 6000V spike to some mV value, so any MOV or device like it just adds nuisance currents and is redundant.

 

Noisy transformers fix often is to tighten the four large bolts on the frame when the Transformer is warm to hot. These devices are years old and with transport the nuts loosen.  Check this by examining the paint on the nuts, if it's cracked, then they are loose.