Discussion of AC mains isolation transformers

[diecaster]

It's not odd at all. The Topaz 2.5kVa 91002-31T that I have does not have shield wire terminals and looks just like the photos flkin posted. That wiring looks to be correct based on this:

 

 

[Johnseye]

9 hours ago, flkin said:

I have a Topaz 1.8kW unit rated for 120/240V. But for this unit, I don't see any S1 or S2 shields which I'm more familiar with having done the wiring for another 1kW Topaz. So I'm writing to ask for advice on how I should do this unit. I am trying to wire it for 240V and am following the schematic on the black end plate.

 

Input:

Jumper between H2 & H3

Live wire on H1

Neutral wire on H4

 

Output:

Jumper between X2 & X3

Live wire on X1

Neutral wire on X4

 

 

Having done the above along with earth to chassis, the output multimeter reading is indeed 240V but it seems that live and earth are also connected and this surely isn't correct. Have I made any obvious mistakes?

 

I'm also concerned that the black and white cables (from the pics) from the windings are not attached correctly. Currently they are connected as:

 

Input side: Black H1 and H2, White H3 and H4

Output side: Black X1 and X2, Whilte X3 andX4

 

Can anyone help with the necessary connections?

 

Thanks!  Kin

 

 

 

 

This post helped me tremendously.  The wiring diagram on the back plate isn't exact.

 

 

[wwc]

On 11/15/2016 at 6:16 PM, Em2016 said:

 

Great advice on moving the iso transformer away from the microrendu and DAC though. I hope it helps others the way it helped me

 

 

What about shielding the Topaz from components?   A faraday shield?

[diecaster]

10 minutes ago, Johnseye said:

 

This post helped me tremendously.  The wiring diagram on the back plate isn't exact.

 

 

 

You mean the green wires to the Topaz chassis like in these photos:

 

 

[wwc]

On 11/26/2016 at 7:38 PM, sockpit said:

I The small topaz is barely warm and apparently up to the task. John put the GFCI business to bed. No more little boxes for me, thanks 

 

 

Sent from my iPad using Computer Audiophile

Sockpit,  So the conclusion is no GFCi needed with the transformer?  I somehow missed this....

Thanks

[Cornan]

9 minutes ago, wwc said:

Sockpit,  So the conclusion is no GFCi needed with the transformer?  I somehow missed this....

Thanks

 

Me too actually! However, that reminds me of something that I wondered about before. If removing the IT input fuse and replace it with a pure silver wire (fake fuse) and instead add a RCD/GFCI at the input as well as at the output is a good or really bad idea (both for SQ and safety)? ??

[One and a half]

4 hours ago, wwc said:

What about shielding the Topaz from components?   A faraday shield?

Try and keep a Topaz at least 1 m away from components to be on the safe side only for magnetic influences. A Topaz is that heavy, it is best placed on the floor anyway. My unit is 25 m away from audio components, so there's no magnetic influence at all   There's an intense field when the Topaz is started.

[wwc]

What about building a magnetic shielding case around the Topaz??

[Ralf11]

just rely on the inverse square law with distance

 

 

BTW, if people are terminating wire with the blue crimp-on terminals seen above, there are better crimp-ons available - do a search on "marine grade" crimp on terminals

[One and a half]

1 hour ago, Ralf11 said:

BTW, if people are terminating wire with the blue crimp-on terminals seen above, there are better crimp-ons available - do a search on "marine grade" crimp on terminals

 

A marine grade crimp is made from tin plated copper.

 

A standard industrial crimp is made of tinned plated copper.

 

That crimp is good for 75 Deg and is no good for a Topaz.  This crimp, looks much the same is rated at 105 C in operation and is suitable for Topaz since the minimum temperature rating is for 90 C wire (and crimps). Alternatively a crimp without the colour sleeving is also possible, but clearances need to be observed.

 

[Ralf11]

https://www.molex.com/molex/products/family?channel=products&chanName=family&key=heat_sealable_terminals

[One and a half]

22 minutes ago, Ralf11 said:

https://www.molex.com/molex/products/family?channel=products&chanName=family&key=heat_sealable_terminals

Just the ideal connector in case a connector is exposed to battery acid or diesel, so who listens to their audio system, next to a rotating engine @Ralf11 ?

[wwc]

Finally a response from Eaton on their Power Suppress 100:

 

We do have testing on our 1kVA T100 and the interwinding capacitance is about 0.90pF.  This is greater than the 0.005pF that is referenced below, but with the output filter network in play, the apparent noise attenuation is -136dB.

[Solstice380]

10 hours ago, One and a half said:

That crimp is good for 75 Deg and is no good for a Topaz.  This crimp, looks much the same is rated at 105 C in operation and is suitable for Topaz since the minimum temperature rating is for 90 C wire (and crimps). Alternatively a crimp without the colour sleeving is also possible, but clearances need to be observed.

 

That is probably a better connector than the marine one.  The marine connectors just have some goo in them to prevent water intrusion.  Typically not a problem in your listening room!  

[Solstice380]

1 hour ago, wwc said:

Finally a response from Eaton on their Power Suppress 100:

 

We do have testing on our 1kVA T100 and the interwinding capacitance is about 0.90pF.  This is greater than the 0.005pF that is referenced below, but with the output filter network in play, the apparent noise attenuation is -136dB.

 

Well that clears things up.... NOT!

[Ralf11]

that "goo" will also exclude air, hence slow oxidation

 

but you kids do what you want

[Solstice380]

14 minutes ago, Ralf11 said:

that "goo" will also exclude air, hence slow oxidation

 

My life in't long enough to worry about that!

[sockpit]

On 4/9/2018 at 2:05 PM, wwc said:

Sockpit,  So the conclusion is no GFCi needed with the transformer?  I somehow missed this....

Thanks

Somewhere, probably far above, this got hashed out.  Someone was insisting on the need for a gfci on the output side of a normal Topaz. I asked JS either publically or in a PM about it.  He assured me it was perfectly safe by itself. 

 

Now if you’re  into modifying these things endlessly, floating them, unfloating them, balancing them, turbocharging them, etc. in the quest for a nanoincrement of SQ, you may need to add a gfci, wrap it in rubber, sell your house, and move next store to the local fire  department.  

 

But if you just buy a topaz and are lucky enough to be able to plug it in and use it as designed, you can relax and enjoy the music ?

[Speedskater]

On 4/10/2018 at 9:48 AM, wwc said:

Finally a response from Eaton on their Power Suppress 100:

We do have testing on our 1kVA T100 and the interwinding capacitance is about 0.90pF.  This is greater than the 0.005pF that is referenced below, but with the output filter network in play, the apparent noise attenuation is -136dB.

Measuring anything as big as a transformer at less than 10pF is not easy. Different test setups and different meters will give different results. At the end-of-the-day it doesn't matter.

[Abtr]

14 hours ago, sockpit said:

Somewhere, probably far above, this got hashed out.  Someone was insisting on the need for a gfci on the output side of a normal Topaz. I asked JS either publically or in a PM about it.  He assured me it was perfectly safe by itself. 

 

Now if you’re  into modifying these things endlessly, floating them, unfloating them, balancing them, turbocharging them, etc. in the quest for a nanoincrement of SQ, you may need to add a gfci, wrap it in rubber, sell your house, and move next store to the local fire  department.  

 

But if you just buy a topaz and are lucky enough to be able to plug it in and use it as designed, you can relax and enjoy the music ?

You only need a (2 pole) GFCI on the output of a mains IT for safety (protection against electrocution), if you *float* the secondary. Nevertheless, IME shorting phase and Earth of a non-floating secondary output will violently blow a high current mains fuse; it doesn't gently trip the central GFCI. So even if the secondary isn't floating you may prefer a GFCI (RCD) on the output of your IT.
 

[Abtr]

On ‎7‎-‎4‎-‎2018 at 11:45 PM, One and a half said:

If the earth is lifted from the centre point, the transformer doesn't do any work at absorbing unbalanced crud ...  

Well, I solved my mains earth problem. I reported earlier that connecting the secondary center tap of my balanced IT to mains earth produced unacceptable audible distortion in my system. I found the cause of this distortion. It is a significant DC component/offset in my AC mains and earth.

 

I already used a DC-blocker in the live mains connection to my IT, but not in the earth connection. Connecting mains earth to the secondary center tap injected about 1V DC into the balanced zero reference. So I put a single capacitor (400V 18uF MKP type) as a DC-blocker in between mains earth and the secondary center tap, which solved the problem.

[nbpf]

It took me a while but I have now managed to read through this thread from the beginning to the end: many thanks to all contributors!

 

I understand that 1) the main reason for having all devices plugged into a single unfiltered, unprotected distribution block is to lower the noise generated by leakage loops and 2) putting an IT between the (unfiltered, unprotected) distribution block and a wall socket is an effective means for protecting the connected devices from power surges while at the same time preventing high frequency noise from the mains to enter them.  

 

I have all the audio devices that are interconnected and that need power (a Naim SN2 + a Naim DAC, an Oppo 203, two Teddy Pardo LPSU and a 5V/3A ifi PS) connected to an unfiltered, unprotected 6 plugs power strip. I have never experienced any hum from the SN2 or from any other device. My power strip is directly plugged into a wall socket which is not on a dedicated mains spur. I have never tried to assess the quality of this power supply, I would not know how to do so. 

 

Just to put what I have learned in this thread in a perspective: How do the problems (and solutions) discussed here relate to the problems addressed by PS Audio power plant devices? And is there a way of estimating the possible benefits of inserting an IT between the power strip and the wall socket without actually doing so? Thanks, nbpf

[drjimwillie]

 I have two dedicated 10 gauge, 20 amp lines from my system.  I have my 250 W CJ amp plug directly into one of them.  Everything else is on the other one. 

 I know that John S. has said to plug everything into one IT. 

 There has been discussion, as to whether this is best or not for my type of set up. But, has there been any consensus on whether I should plug everything into one IT or if I should get  to IT's, one for each circuit?

 Are used a kill-a-watt and my total consumption is 355 W. 

I have been pricing ITs and a 750 or 1k VA, that would power the entire system cost 3x more than a 500 VA for the non-amp circuit. 

I use the dedicated circuit for the instantaneous current draw opportunity for the amp. 

I look forward to your experience and advice.   Thank you

[look&listen]

11 hours ago, drjimwillie said:

But, has there been any consensus on whether I should plug everything into one IT or if I should get  to IT's, one for each circuit?

No consensus I see (with audiophiles??) . Just try both ways and see which best for you.

 

11 hours ago, drjimwillie said:

Are used a kill-a-watt and my total consumption is 355 W. 

I have been pricing ITs and a 750 or 1k VA, that would power the entire system cost 3x more than a 500 VA for the non-amp circuit.

My kill-a-watt measure similar, but little higher with HT and all on. Choose to get 750VA iso-transformer, but not install yet.

If Topaz be sure to get 91xxx.3x for best spec. No good 1KVA IT's on eBay for long time, they very popular and expensive now. 750VA not so popular so more available and cheaper. Good Luck.

[drjimwillie]

 Thank you for your advice. 

1.  I could buy one 750 and I could  plug the whole system into it and see which sounds better. The amp  on the same circuit   Or the camp on the separate circuit without an IT. ( no 750s available now  at a reasonable price )

2.  I could buy one 500 VA and to the same experiment as above.  But I will know that it does not have the same overhead, and I would worry that I'm losing some of the dynamics because I don't have the instantaneous current. ( I saw a 500 for $125).  I would worry that I should buy a 750 so that I could have the suggested overhead. 

3.   I could buy two 500s.   One for each circuit. 

 

 I still wonder if anyone who has two dedicated circuits is using ITs and what they have done. 

 

 Things are always slightly more complicated. I have a small hum or buzz and when I connect a ground wire between the chassis of the amp and preamp the bus gets louder.  This is what I'm trying to fix