Ferroresonate Transformer Application

[tt90lrb]

Looking for some technical insight or a pointer to better understand a problem recently.

A customer has a UPS (on off-line ferroresonate type) providing backup power for a control system. Because of power quality issues at another customer site, they have installed a power conditioning transformer at the input of this UPS (the transformer made by SOLA/Hevi-Duti and is also a ferroresonate type device).

The customer has experienced noise/spikes/harmonics between the UPS and the xfmer causing the UPS to switch to battery power constantly. Basically copied a solution at one site to this application without fully investigating the equipment involved.

Both the UPS manufacturer and the xfmer techs stated that a power conditioning transformer should not feed a UPS of this type. The combination of a high impedance device feeding another high impedance device was causing the issue.

The power conditioning xfmer was removed (replace by a general purpose xfmer) and no problems have been detected.

Neither vendor could supply any tech data on why this application doesn't work. Could someone elaborate?

 

[jbartos]

Comment: The Ferroresonant Voltage Regulator Transformer (definition is in IEEE Std 100 "Dictionary") is used where the stabilized voltage output is required. It uses ferroresonant regulator circuits that better be well engineered and designed not to cause any interference.

 

[jbartos]

Suggestion: Visit

http://www.elect-spec.com/klr_z.htm

or

http://www.thomasregister.com

and type Transformers: Ferroresonant under Product or Service, which will return 80 companies to approach to

 

[tt90lrb]

Thanks jbartos - I'm really not looking for product info on ferroresonant voltage regulator transformers, but the problems involved in having two in series (feeding one from the other). It's my understanding that this type of xfmer produces a non-liner current waveform. When load increases, they tend to fight each other and produce noise/harmonics on the connections inbetween.

Looking for someone to provide to tech detail on what might be happening.

 

[busbar]

You may be hard pressed to find published documentation that deals with “cascaded” ferroresonant AC-power devices. The problem may be better described as a higher-impedance source—the SOLA—not working well serving as what it sees as a lower-impedance load—the UPS.

The UPS may have two problems with this arrangement. The lower impedance may be a “nonlinear” load seen by the SOLA varies at difference points on the upstream [imperfect] sinusoidal voltage the SOLA presents. The UPS’ power-handling circuitry may not work well with a limp source, or it may be [possibly on a sub-cycle basis] its integral voltage-sensing circuitry interprets the source waveform [or a part of it] as being out of range of its acceptable limits.

At any rate, ferroresonant-AC devices generally are happiest seeing a stiff source, while serving a much-less-stiff load. In some cases, the “weak spot” is voltage crossover distortion on the upstream-transformer secondary, that is sensed as something like a frequency deviation by the UPS, even though RMS-AC measurements with a DMM indicate everything is fine.

Sometimes a UPS can report failure information that could provide clues.

 

[jbartos]

Comment on the previous posting: Actually, the ferroresonant power supply manufacturer should provide tech support, including literature, for its product paralleling. Visit

http://www.lamarchemfg.com/PDF/A36D.pdf

where load sharing/paralleling is addressed
The reason for this is the output quality can vary and appropriate filtering is required to minimize the mutual interference and the harmonic waveform synthesis on the output. This is however not all if there are nonlinear loads. Then, additional filtering or load caused harmonic mitigation will be required. Certain technological aspects and methodology the manufacturer uses may be proprietary.

 

[ScottyUK]

jB,

The issue isn't paralleling, it is 'series-ing' where one device is feeding the other. The situation is rather different.



-----------------------------------

Start each new day with a smile.

Get it over with.

 

[jbartos]

Comment: The "series'ing" is included under my comment covering the nonlinear load. In fact, the paralleling may happen to if the units come in the modular version. Modularity is a very attractive solution; especially, when it comes to shipboard applications, and larger systems.

 

[tt90lrb]

I believe busbar has described the original problem correctly. Most of the RMS-AC readings didn’t show any problems. However, the on site personnel was reporting a frequency jump in the 150+ range when load was increasing and the UPS unit switched to battery power. I do believe it was a distorted wave or harmonic feedback that was causing the erroneous frequency measurement – but it was still read by the UPS as a power quality issue.

With out the SOLA power conditioning transformer feeding the UPS, the unit seems to be operating fine under full load. Only complaint is flatness in the output voltage waveform of the UPS at zero crossing. I believe this is typical of this type of UPS.

I appricate the feedback.

 

[jbartos]

Comment: SOLA/Heavy duty does seem to offer "power conditioning transformer" as stated in the original posting. It would be good if the nameplate data are posted to be able to identify the hardware. Visit

http://www.sola-hevi-duty.com/

for product categories
It appears that there are:
1. Sola UPSs
2. Sola Power Conditioners
3. Sola Transformers
The above postings suggest that there might have been interaction between nonlinear elements within the UPS input and within "so-called Power Conditioning Transformer" output side on the UPS input side.

 

[jraef]

Of interest to me in this thread is that Sola could not offer you a good explaination. Frank Sola INVENTED the ferroresonant transformer in 1936 and the company that bears his name had an exclusive patent on the technology until 1986. I worked for Sola for a short time 15 years ago and there were many very good engineers who could have produced excellent written expainations. Now that they have competition they must have lost the desire to maintain expertise of the products. Sad really.

By the way, Busbar is right on the money. Not a lot of study has been done because there would be little reason to do what they did. Except of course, as in this case, it accidentally ended up that way.

"Venditori de oleum-vipera non vigere excordis populi"

 

[jbartos]

Suggestion: Perhaps a little modeling of the setup may reveal the problems; including harmonics, resonances, spikes, etc. Power supply quality monitoring is also useful.

 

[Jasonarc]

tt,
I had a similar result with a 5kVA ferro UPS feeding a 1.5 kVA on line UPS. I was told that tank circuit (ie capacitors that run the ferro xformer at full saturation) cause the output of the ferro xformer to resonate when it's supplying an inductive load (ie on-line UPS). This causes the on line UPS to transfer to battery operation and effectively disconnecting itself from the output of ferro-xformer.
Guess what happens to the ferro UPS output next... it stabilizes and therefore the on line UPS tries to return online only to repeat the whole scenario again.

Put a scope on the output of your ferro line conditioner and watch the funny things happen.

 

[tt90lrb]

At first glance, the waveforms don't look that bad. You have to realize that I'm working this through e-mail (remote off-shore site) and the data doesn't have a real high resolution.

Funny thing ... different rig with a slightly different application. Same ferro xfmer but feeding an on-line type UPS (rectifier-inverter scheme). The opposite problems are happening. Under light load, the UPS will not sync with the incomming feed. Once the unit is loaded above 30%, the alarms go away and all is running fine.

My thinking is that the rectifier isn't providing the ferro enough current draw (inverter chopping the waveform) for it to regulate properly.

 

[djs]

I had a slightly different problem on a UPS with a ferroresonat transformer. It was feeding two server type computers with redundant power supplies.

The UPS would trip out due to either under or overvoltage. The manufacturer had a fix, replace a voltage regulator board in the UPS.

 

[jbartos]

Comment on the original posting marked ///\\The combination of a high impedance device feeding another high impedance device was causing the issue.
///The high impedance source would cause a high voltage drop. Is that the case? This is what can be measured.\\