Transformers Connected In Series

DO learn and understand BEFORE doing these things.

DONT do them without learning and understanding these things.

PAY attention to safety or you will not last long.

Square waves, ouch.

What are you doing exactly, and have you oversized the transformers and diode bridge to take into account the harsh waveform you are using (transformers do not like square waves).

What is the application and what are the transformers? Please tell me you are not using standard 50Hz transformers with iron laminate cores rather than high frequency ferrite cores.

<cringe>

Has anybody, electronic engineers or otherwise, seen any point in the series connection?

Sounds like uweka is building an autotransformer out of available parts or to save money rather than buying the correct transformer and has been caught out by the unexpected 50/60Hz distribution-sized transformer response to a non-sinusoidal waveform.

Hopefully more information will be provided to help us answer the original question.

I can't see the point of it but that doesn't mean that one doesn't exist...

The comments thus far regarding pushing a square wave through a trafo should be heeded by the OP.

Please tell us what this is for OP?

It's not the series connection. It's the square wave causing the spikes. That's how an ignition coil works to produce a 15kv spark in a car engine.

Horace, If you spike a transformer, then put another transformer in series, the spike on the second transformer will be greater. Even though these are step down transformers 2:1, the spike in current on the primary will create a very large and fast changing magnetic field will impart larger current spikes on the secondary.

I'm still trying to figure out what the op is trying to accomplish. Three 2:1 step down transformers in series with an 8khz rectangle wave and a lot of power? The output of the third transformer is going to be a huge spike.

Uweka,

Some of the comments above are partially correct. Don't take them at 100% truth. Few people have real experience of power electronics at the 150kW level.

It is very difficult to tell you the exact cause of the problem because there aren't enough details in your post to analyse the situation. In general it should work as I have done it but the details are important.

Here are a few things to look for.

1- Make sure that the transformers can sustain the volt * seconds you are applying without saturation.

2- Make sure that the transformers' polarities are correct. This could cause one of them to see higher voltage and saturate or produce a higher voltage at the secondary.

3- Leakage inductances (primary and secondary) can cause enormous spikes during switching transients. Control them and add clippers / snubbers if needed. Reduce the di/dt of the IGBTs if possible.

4- Try it at lower voltage first, then increase it so that you can monitor the parasitic effects progressively.

Good luck, and play safe.

Assuming the BH curves of the transformer cores used in your transformers can properly respond to your operating frequency, connecting the transformers in series is should work as expected!

Since there are two possible ways of connecting transformers in series,

1) series aiding

2) series opposing

be sure that they are all properly connected..

Same connections should be true or applies to both the secondary and the primary windings of the transformers.

Your description is fuzzy. A schematic would be a lot better. A rectangular shape from a full bridge? I would expect DC. Are the transformers in series to increase the voltage rating? the power rating? Are they designed for 8kHz? Were the outer secondaries measured? How much voltage was applied to the primaries?

I've been thinking about this, trying to fit the pieces together in my head.

Full-bridge rectified square wave at 8kHz? Sounds like a switching power supply is being built, those are usually used for sensitive electronics as part of an Uninterruptible Power Supply (UPS).

Voltage step down from 1.5kV to 750V? That's enough 'Electromotive pressure' to destroy anything that could be described as 'sensitive electronics.'

Here in Chiocagoland, we have an electric train system for mass transit ('The L,' short for 'Elevated Train'), and that runs on 800VDC, but there we are talking about electric motors, which do not require 'pure, clean' DC from a switching supply to run them, and in fact, big electric motors tend to 'pollute' the DC they run on because of the power dips and spikes when they start and stop.

I would recommend stepping the voltage down from 1.5kV to 750V while it is still sinusoidal, as that is the best 'flavor' of power to run through transformers, and once you have the voltage at the right level, THEN convert it into an 8kHz square wave and rectify it into high-frequency pulsed DC. After that one RC network to act as a 'low pass filter' should squelch the spikes and dips, leaving you with DC as clear and clean as glacial water off the Bavarian Alps.