Transformer schemes 2

[itsmoked]

I have three of these laying around:

I've have 208V 3ph and I need 460V 3ph.

I initially thought I'd use use the 0V and the 208V taps daisy-chaining each H3 to the next transformer's H4. Supply 208 to each H3 and use the three H1s for 460V but on closer examination that seems more likely to get me 600V H1-H1-H1.

Suggestions?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[davidbeach]

What are you doing with the X side? Without current in the X winding there's no current in the H winding.

 

[wroggent]

I would suggest trying to connect H1 to X2 and H4 to X3 and putting your 208 across H3 and H4, per transformer.

 

[wroggent]

I suppose you'd also need to connect H3 of 'transformer 1' to H4 of 'transformer 2', H3 of 'transformer 2' to H4 of 'transformer 3' and H3 of 'transformer 3' to H4 of 'transformer 1', as suggested in OP.

 

[davidbeach]

A wye connection of the X windings X0-X1 and a delta connection of H1-H4 will give 440. I'm not seeing anyway of getting that last 20 volts.

Bill will come along and have some esoteric connection that won't get 460 but will be within 5 or six volts. Just gotta wait for him...

 

[waross]

You connection should work if you can live with 440 Volts instead of 460 volts.
You are ignoring the secondary windings and using the primaries as autotransformers.
Draw a delta with H4 and H3. Then extend each side to H1.
There is a phase shift. The amount of phase shift depends on the ratio of primary to secondary voltages of the auto-transformer connection. (The X1, X2, X3 secondary is not used and has no effect.)
I have some old text books that show a similar connection (But with much smaller step-up ratios) used on power transformers to intentionally introduce a small phase shift so as to vary the load sharing between transmission lines.
Is that technique still in use, David?
Thank you for your confidence in me, David.
As I was looking at this and composing an answer I kept thinking;
"Get close to 460 Volts with those available voltages? No way.
440 volts is easy, but 460 Volts...?"
Then the light went on.
Think single phase.
On each transformer if you connect the 24 Volt secondary as an autotransformer boost, you will have 464 Volts. That is within the 5 or 6 Volt limit suggested by David.
And further, our standard voltage in North America is 480 Volts and I consider anything between 440 Volts and 480 Volts to be fair game.
Your connection will be 208 Volts in at H4, H3. Connect H1 to X3.
464 Volts will be developed across H4, X1.
Now we have three 208 Volt to 464 Volt autotransformers.
Connect the H4, H3 windings in delta across 208 Volts.
Take 464 Volts from H4, X1.
Ignore the phase shift.
The voltages will still be symmetrical with respect to the system neutral.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[HamburgerHelper]

What old book shows this connection?

 

[waross]

I have some old text books that show a similar connection

Not exactly the same. My memory is a little rusty.
The connection proposed for Keith produces mostly voltage change with some phase angle change.There is nothing wrong with a phase angle change in a single stand-alone unit.

For power flow control, the added winding is taken from a different phase.
The result is mostly phase angle change with some voltage change.
The principle for power flow control and some various possible connections for power flow control:

http://pjm.com/~/media/committees-g...jm-phase-shifting-transformer-principles.ashx

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[itsmoked]

First off, davidbeach, that was exactly my thinking! I figured Bill would have me configuring them as some sort of zig-zag Scott-T mobius.

I just need to run a 460V motor blower that draws 250mA so somewhere around 460 will likely work fine.

You are ignoring the secondary windings and using the primaries as autotransformers.

Yes that was my hope so these 100VAs would do the trick. Nothing needs the LV sides.

Draw a delta with H4 and H3. Then extend each side to H1.

There is a phase shift.

Don't care in the least.

I must confess that I'm forced to test this scheme on bonified 240V delta power.

So I've hooked them up using H4 and H2 instead of H3. When I do that I get 461V between H4 and H1 measured across each xfrmr and 602V between any H1s. This is without the added X1 based boost.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Hi Keith.
Would your 240 Volts actually be 230 Volts? That would explain the 461 Volts.
As for the 602 Volts. Memory is getting rusty.
Usually I am boosting much less than the 208 Volt to 440 Volt jump that you are making. With that ratio of boost, the phase shift adds more voltage than I had anticipated. I have never used a full autotransformer delta boost.
I use either full wye autotransformer or open delta autotransformer. That's what I see in the field also. Please forgive my for neglecting the added phase shift voltage.
How about an open delta autotransformer boost. Two transformers and 440 Volts. Easy to add the extra 24 Volts if you need it.
What is the VA rating of the transformers?

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[itsmoked]

Hi Bill,

For some at-the-scene reporting!

These are 100VA control transformers.

This is the voltage I've got to test with. Notice I'm using H4 and H2 (not 208V H3)

This is what we get across a single transformer wired up in our full 3 transformer delta scheme.

But, this is what is seen across the delta of the three H1 terminals between any two transformers.

And another bit, is the current demanded by this "unloaded" micro bank of control transformers. .15A @ ~230V. Seems high but maybe not.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[111R]

I think the 600V+ measured from H1 to H1 on the next phase is because you have a 120° phase shift and resultant sqrt(3) factor on the H1-H2 section of the winding between phases.

 

[itsmoked]

Whoa 111R. So you're saying:

The extra winding between the 220 tap H2 and H1 the 440 tap, which happens to be 220V too, is contributing 220V x √3 = 381V

Adding 220 + 381 = 601V : (H2-H4) + (H1-H2)*√3 = 601V.

Did I get this right? If I did that's interesting news to me.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[prc]

I think the original Dy suggested by Davidbeach is the only possibility. A single stack winding cannot be made in to an auto-transformer. Voltage will appear but you will not be able to draw current! For transferring current from one winding to another, the leakage impedance between windings should be low. Consider a two limb core. One limb has two concentric windings H&X . This will work as a HV/LV single phase transformer. If you put the H and X separately on the two limbs( typical representation of transformer in text books), it will not work. You will get no-load voltage correctly. But you will not be able to draw current.

 

[waross]

Hi Keith.
Draw a sketch and you will see that the voltages are as they should be.
Part of the problem is that your first hand drawn sketch is grossly not to scale.
You show H1, H3 as a little tag.
It is actually greater than H3, H4.
H3, H4 is 208 volts. H1, H3 is 232 Volts.
To make it simple use your present connections.
Then both the energized winding and the extension will both be 220 Volts.
Your 240 Volts delta seems to be only 227.9 Volts.
227.9 Volts x 2 = 455.8 Volts. Measured voltage = 455 Volts.
Pythagoras tells me that the voltage from H1 to H1 will be √7 x 227.9 Volts.
Measured = 604 Volts. Calculated = 602.97 Volts.
More tomorrow.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[prc]

The connection proposed by waross is extended delta auto connection. The relation ship for voltages is as below:
Let E1 =voltage acrossH1 E2= Voltage across the delta Em= extended delta voltage. In the case considered all of these are 220V.
E1 is the vectorial sum of (E2+Em-Em(120 degrees shifted)) So the relationship is square E1= 3 Square Em + 3EmE2 + Square E2 (Please see page 267 of Blume's book on transformers)
This gives square root7 x227.9=603V as concluded by waross. But as far as current loading is considered, my earlier statement holds good.

 

[waross]

prc; Thank you for expanding on the origin of the square root 7 factor.
I accept your information that some transformer configurations are not suitable for buck boost operation.
However, for these connections we are using different taps on the same winding. If that is a problem, then any tapped transformer would have a problem.
I am aware that some winding configurations of small transformers are not suitable for buck-boost operation, but most transformers are suitable for buck boost use subject to voltage and insulation level issues.
I have used many conventional transformers for buck or boost service and I have seen many more in operation with no issues.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[waross]

The safest connection will be two transformers in autotransformer boost configuration. That will give you 464 Volts. This depends on your willingness to stress the 24-120 Volt winding with 464 Volts to ground.

However there may be another issue.
100 VA / 440 Volts = 0.227 Amps. That's cutting it close (actually cutting it on the short side of the line) for a 0.250 Amp motor.
The open delta autotransformer boost connection is a standard connection in some industries. I see it a lot in the field.
I am hesitant concerning the 120 Volt wye to 440 Volt delta connection. If you use this connection consider floating the wye point.
Given the unique properties of a four wire wye:delta connection, it won't take much unbalance on the supply system to fry those transformers.
The note allowing use of 115 Volts from the 120 Volt winding, raises a little concern about the origin of these transformers.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[itsmoked]

That may be the 'safest' Bill but it won't work with what I have.

However.

Your mention of 120V/115 in that screwy schematic had me hunt down the manufacturer's data sheet. This clarified that the the secondary was really 115V and the primary side is actually 208/230/460


Here's what works:

Auto-transformer using everything.
The secondary 115V winding is added to the bottom of the primary HV part of the transformer which is the first 230V half.

This gives 115 + 230 = 345 (input ratio)


Then this 345 has added the rest of the original primary which is another 230:
345 + 230 = 575 (output ratio)


Our turn-up ratio then is 575/345 = 1.67

Those are the hard winding ratios.

Now we supply merely 208V to the lower half and get the ratio'd result:
208V x 1.67 = 347.4V out of each transformer's H1.

Referring back to your handy √7 Pythagoras rule we see that 460 Volts, the ultimate delta voltage sought for, would be present with interim voltages:
460V/√7 = 173.86V which when doubled:
173.86V x 2 = 347.7V (All from your earlier post)

As for VA it seems roughly, 100VA/230V(transformed part) = 0.43A, needed: 0.25A.

Here's a better balanced diagram:
(NOTE: The circled values are actual applied/resulting voltages while the uncircled are winding ratios.)

Keith Cress
kcress -

http://www.flaminsystems.com

 

[prc]

Let us make things simple:
1) You require 460V 3 phase ( 265.6 V single phase) with 208 V 3 phase ( 120V single phase) input.
Put X winding in star with X3 earthed, H winding in star with 208V tap earthed.
Supply 3phase 208 V to X1. So it is overexcited by 120/115 pu.
At H1 you will get (460-208) x(120/115)x1.732 =455.5 V, nearest to your requirement.
2) Waross, you can use tappings in a winding freely, provided power input is in the primary winding. Generally you cannot make a tapped single winding in to auto-transformer, with primary open. Exception is small control transformers where the entire windings will be layer type, each tapping section forming a layer. So each layer will act as a concentric winding and then you can draw current from auto connection. Please try the winding connection in my first posting. You will get voltage, but you will be able draw only negligible current.