1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Topaz Low-Capacitance Isolation Transformers - for Affordably Clean Power

1 2 3 4 5 6 7
9 10 11 12 13 14 15 16 17 18
  1. One and a half
    Hi Mike,

    There are a few topics in the post which will be spread over several posts, if that's ok.

    DC blocker
    The main problem in selecting the diodes for the DC blocker which cannot under any circumstances fail, is how to determine the peak current due to the inrush of the Transformer.

    Gut feeling, the Topaz is 'not too bad' as far as inrush is concerned, with a a disclaimer, have not measured.
    However a 120/60 1000VA toroid measures 10 times load current around 80A. Diodes are rated with continuous as well as peak values, which needs usually a lit larger diode. Please check the peak rating of the DC blocker.


    Peak current (amps) = n x VA/V
    n = 10 for power toroid, less for Topaz, but will measure my 1000VA on the weekend.

    The caps in DC blockers also need some thought, usually voltage rating and how much shift above the zero is needed.

    A little bit to consider!

    Did you know.....
    During a short circuit the voltage has a DC offset for AC waveform, the higher the voltage, the more current.
  2. zilch0md
  3. One and a half
    Topaz went through a few changes in ownership, themselves, Square D, then MGE. Square D are still about absorbed by Schneider. The models changed to, Mike, you have a penchant for finding Topaz transformers, I wonder if there are more catalogues out there that describe the model numbers accurately.
    During the Sqaure D time, the 910, 911 series were transported to the MGE model numbers. It is here where the models came with cords and without. It's very likely the corded variety did not have the terminals of the fixed series to save 50c and had looses wires joined by wire nuts.
    If you're lucky the wires are marked, but over time, if Topaz used a wrap around number with sticky goo, it would have fallen off in service by now and on opening the terminal cover, the wire markings would indeed be very ambiguous. In that case, to test requires the use of a very small transformer of 3V or less to check what wires went were. Matter of measuring and working out where the taps are, then with a scope measure the phase, to see where the peaks line up. Not a job of a Saturday afternoon!

    Transformer tips
    Topaz/MGE quote the impedance of the transformers of 3-5%. This value has certain properties, namely:
    - If the impedance is known, the inductive reactance can be worked out. If there's a meter about with some accuracy, and ohms law, the inductance can be calculated. Google can help out there.
    - The 3-5% value is the minimum voltage to flux the transformer, so it starts to work. For a 120V supply, 3% is 3.6V is all you need for the transformer action to work. 5% is 6V. The variance depends on the model and construction. Generally the higher power models like 5kVA are 3%.
    zilch0md likes this.
  4. zilch0md
    Reading your posts, I feel like a sixteen-year-old, who barely knows how to drive the car, much less what makes it go. :wink:

    Xentek is another brand that had the same look as the Topaz/Square D/MGE/Daitron transformers, but until just now, I'd never seen a Xentek with a capacitance rating as low as 0.0005 pF, as with the -31 and -32 Topaz and MGE models.

    But check this out!

    Here's a 5kVA Xentek with a label that rates its capacitance at
    0.0001 pF



    They want a pretty penny for it.
    Last edited: Oct 5, 2017
  5. pctazhp
    I feel exactly the same way. This thread is certainly not for the faint at hear, particularly for someone like me whose eyes glaze over at most of the posts. I'm moving to Europe before the end of the year and want a transformer that will step down 240 V to 120. Any SQ improvement would be welcome but mainly I don't want a transformer that will degrade SQ. Even more important, I don't want a transformer that will damage my gear, and have no way to know how to make sure that doesn't happen.
  6. oneguy
    I found one a couple of weeks ago but figure it was a misprint since I couldn’t find any other corroborating data. I guess it may not have been.

  7. zilch0md
    This will only burden @One and a half , but he's at least thinking about posting some steps for testing these critters when they are unboxed.

    It is dubious when you consider that every other Xentek label I've seen that actually has a capacitance rating (many Xentek's don't) is rated at 0.001 pF, not 0.0001 pF. Just a typo?

    Here you go -- fixed it!

    Last edited: Oct 5, 2017
  8. mwhitak
    Thanks Mike and @One and a half ! Great thread and info. Learned a lot from you guys and John S. I am looking forward to testing procedures and recommended protection ideas. Topaz 2.5 kva 91002-31 on the way. Listed on ebay as 500 VA but probably an error. I think you mentioned it in an earlier post Mike. Got great price. Figured it was worth the chance since the model number and pictures seem to match the 2.5 kva unit. Trying to decide which way to wire the secondary, grounded neutral, floating neutral, or balanced. Lots of opinions on this. I'm leaning to try balanced 60-0-60 with equipment grounds wired/referenced to center tap of secondary (only), then running power output through a 2 pole rcbo. A fault (to chassis) by either power conductor on the secondary should trip the rcbo., but the secondary side equipment grounds will stay isolated from the primary/house side equipment ground and potential nastiness. My thinking is get rid of the equipment grounds, if possible, safely. They are only present on equipment to provide an alternate pathway for current in the event of a fault condition, that trips the overcurrent protection, thus making the circuit safe. However, in the US, current NEC code allows for the elimination of equipment grounding on appliance type (receptacle) loads that are protected by a GFI. This was done to allow the safe use of 3 prong plugs in older houses with no equipment ground conductor present in the original wiring. A secondary wiring setup that allows tripping of the overcurrent device (on the entire secondary), in the event of a fault to chassis by either current carrying conductor, would seem to meet the spirit of the code, even if/though the equipment ground is only connected to the center tap of the secondary (and not the transformer chassis/primary side equipment ground). I would also mention that it's technically a code violation to use balanced power set ups in a residential environment, just so everyone is clear. The reason is it's atypical from the norm, and would confuse people (read homeowners). Proceed with caution. Also waiting for delivery of Eddy Current Aficionado SET to pair with current HD800S. I'm thinking my PC/file server on UPS (networked to my MicroRendu with FMC's right before the Rendu) would probably not need to be plugged into the Topaz power strip as it's sufficiently isolated already with the FMC's in line on the ethernet connection to the Rendu....and....the UPS and PC smpt's will spew a lot of crap out ? I'm thinking another dedicated circuit, maybe with a smaller, second topaz or B&K isolation transformer for the PC. Is there a minimum amount of loading the topaz would like to see ? I know total actual load should be no more than about 50 or 60 % of total rated VA, but what about a minimum amount ? The flea watt SET will not present a big load, so I'm hoping the 2.5 kva isn't too big...couldn't pass it up though for the price and future expansion possibilities. Cheers and many thanks guys !
    Last edited: Oct 7, 2017
  9. zilch0md
    Thanks for the many good points you've made in your first post, mwhitak!

    Regarding balanced power, you'll not find me promoting it anywhere. I'm certainly tolerant of other people trying it, but I'm not interested in doing so, myself. In addition to introducing some inherent risks, it seems to be a "novelty" solution looking for a problem that doesn't exist. I can't say that I've done any comparison testing, but I'm content in the knowledge that my "unbalanced" Topaz is providing -146 dB of CMNR and -65 dB of DMNR. I've come across a couple of references where it's said that balanced power can reduce CMNR by another -20 dB, but no one has ever been able to tell me where they got that figure or explain how that's possible. In the end, even if that's true, I have to ask myself, could I possibly hear a difference between -146 dB and -166 dB of CMNR?

    Here's hoping that your purchase does indeed turn out to be a very good deal.

    By the way, I lived in Novato, CA for a year, at age nine - went to Lu Sutton Elementary when my dad was stationed at Hamilton AFB - which was closed, long ago - then we moved to Homestead AFB, south of Miami, then to... then to... :)

    Revealing my age, one weekend, my mother wanted to drive into San Francisco to see "the hippies." While stopped at a red light behind some other cars, at the notorious corner of Haight and Ashbury, the sidewalks were covered with people painting flowers on their faces, sitting on mattresses in the lotus position, etc . It was everything we had seen on TV and more. My mother's disgust and disdain must have been plainly visible on her face, as a white guy with a perm, bell bottom jeans and a tie-dyed sweatshirt walked right up to her side window, making crazy faces and even started licking the glass - giving her exactly what she came to see. My mother started freaking out, telling my dad to go, but we were stuck behind traffic, so my dad was telling her to just sit still and stop making eye contact. When we finally started to pull away, the "hippie" saw me smiling at him from the backseat, stood erect and winked at me, making a peace sign with his outstretched arm. He knew that I knew it was all an act. Somehow, my mother didn't get it. That happens to be my most vivid memory of living in Novato. LOL
    mourip likes this.
  10. mwhitak
    Great story, thanks for sharing ! Hamilton AFB in Novato is now a residential neighborhood, and the hippie thing could never happen now in SF, housing costs are astronomical and you can barely scrap by there on a 6 figure income. The whole Bay Area is just nuts for housing. You should see how many people are in Novato now....with everyone renting out rooms in their houses...the population has tripled in the last 10 years !

    Your point is well taken on the CMNR. The Topaz has arrived, and it is indeed 2.5 kva !
  11. uncola
  12. mwhitak
    Thanks uncola...very interesting ! Does anyone know the way to actually measure the interwinding capacitance ? I think I read in the CA thread that someone measured a topaz and got a different reading than the spec. But no one challenged it or questioned it further. Also interesting is the differing language on the topaz units. Some say "effective capacitance" and some just say "capacitance". I wonder if this is just semantics, or if there is a real difference.
  13. zilch0md
    Yeah, congratulations on that really good price for the 2.5 kVA Topaz!

    And I hear you regarding the housing costs. Out of sight! My daughter has a friend who's paying $2800/month for an 800 sq. ft. apartment in the Mission district. Housing costs are so crazy out there, with people suffering as much as four hour commutes from "affordable" housing, that there are homeowners here in the Dallas area and Austin who rent out upstairs rooms to tenants who fly to San Jose daily. A room and private bath in a house in Grapevine, Texas, near DFW airport goes for about $500 a month. It's really hard to believe that expense plus the cost of five round-trip flights a week actually allows them to save money, but they sometimes overnight at a friends' place in San Jose, to avoid flying. And... they officially live in Texas - with no state income tax - only Federal. That's surely a big plus, right there. I wonder if their travel expenses are deductible. Sheesh!

    I suspect @One and a half can answer your question, but I also suspect these old transformers can deviate from their factory fresh specifications. :-|
  14. mwhitak
    Thanks Mike ! Man-o-man that's crazy to fly in from TX for a commute. What won't people do for a job ? To me, if you can't live relatively near where you work, there's a fundamental breakdown in the 'social contract'. I do commercial electric, and I always get offers from companies in silicon valley which I turn down (Tesla recently). I refuse to help them since I could never afford to live in that community. They have driven up the cost of living here. And throwing money away to rent like that...when you could be throwing money away on audiophile pursuits lol ! So I work in hospitals mostly. Thought of leaving for TX, then I heard lots of californians where going there...so scratch that idea...lol. The occupational licensing makes it hard to move from state to state for me, and there's very little reciprocity between states.

    My microrendu has returned with the 1.4 board upgrade ! It was hard to have it gone for a while. Thinking about a DAC upgrade. Musical Paradise MP D2 seems a great value. Can't afford a totaldac or lampizator.
    Last edited: Oct 8, 2017
    zilch0md likes this.
  15. One and a half
    Not a trivial task!
    Some digital handheld meters can measure a few thousand pf, are accurate to 1%. Any better and bench models take the prize, worth more than 6 Topaz transformers. The AC current method is a better for digital hand helds, quite a few can measure down to 1uA. There's a but. The transformers under discussion here draw a lot lower current about 20pA for the 0.0005 pf models, that's 20 x 10 ^-6 uA.. I think you'd have to trust the nameplate data. Effective capacitance takes into account a calculated measurement, could be based on the current method.
    Have a look here how to calculate these values.
    zilch0md likes this.
1 2 3 4 5 6 7
9 10 11 12 13 14 15 16 17 18

Share This Page