Reverse engineering a transformer without knowing the secondary turns

Thread Starter


Joined Dec 22, 2018
This may be a long shot.

I have a high voltage switching ferrite transformer that has failed. Due to the construction of the transformer I could not retrieve all its windings. The only winding I could retrieve was the primary.

Is it possible to recalculate the secondary windings taking in consideration:
- cylindrical core area 80mm2
- operates at 50Khz
- ferrite Bsat ~0.3T
- primary (pin 4-5) has 30 turns. Is feed with +120VDC oscillates at 240v pk-pk
- feedback (pin 3-6) is feed with +2.3VDC oscillates at 4.5v pk-pk
- secondary A - (pin 9-10) oscillates at 2v pk-pk
- secondary B pin 1 ground
- secondary B pin 2 -120VDC rectified input
- secondary B pin 7 -220VDC rectified input
- secondary B pin 8 -2.7KV RMS supply to CRT

Regarding the current I can still retrieve the winding gauges.

Screen Shot 2019-09-24 at 12.02.10 PM.png


Joined Jun 22, 2012
That looks like a Line Output Transformer, in a TV or monitor. These are HF ferrite types tuned to their frequency, you could look at the outputs on the diagram these give out 120v, 220v, and 2.7Kv for focus, and measure the resistance or inductance of the primary winding to give you a base line start, the primary is pulsed at 120V..

so that gives winding ratios of 1/1,,, 1.8/1,,, 22.5/1.


Joined Aug 21, 2017
The 0.8 cm2 at 50 kHz will result in 50/0.8/(50 Hz/50 kHz)=0.6 turns for Volt at 1 Tesla. You have oly 0.3 thus factually results a 1.8 turns to Volt. However, seems Youll use this for meandric signal not for sinusoidal, thus apply the factor 1.11 thus most probable is 2 turns per Volt.

Method for accurate measuring - produce the meander signal with proper frequency, for example by means of IR2153 and pair of mosfets like IRF730 or better. Then switch a trafo at half-bridge circuit and regulate the network Voltage by means of LATR (any roll-adjustable autotransformer). Measure the Voltage as well the current. The voltage at which the losses are 5...10% of I(max) for given cross section of wire may be set as nominal feed voltage. Or 1% for super-high efficiency devices. If any voltage gives the losses larger than this - there may be only two causes - or diminish the frequency, or somewhere in bobbin is hidden the shorted turn.