Oscillating transformer

[Ishanhuss]

Hi there,

I am hoping that someone could tell me what it is; oscillating transformer.

I have a transformer, centre tapped on primary side (3 wires like a voltage divider). Input is 5vdc regulated and secondary was about 700vdc (2 wires) when coupled to circuits.
When not coupled to circuits, input of 5vdc register 36vdc on secondary.

My question is
- how was the voltage raise to that level? There is 2 diodes tied to the secondary side.
- in what way can I make identical transformer? I don't have any specifications other than
measuring voltages from a good working ones.
- if it oscillate at 1000cps, does it mean 500 cycle/sec (500hz)

Thanks in advance.

 

[MikeMl]

Transformers do not oscillate by themselves. Sounds like you are describing a Push-Pull Converter as shown in Fig 12 of this PDF.

If it oscillates at 1000cps, it means 1000 cycles per sec or 1000Hz.

If it steps 5V up to 700V, then the turns ratio between primary and secondary must be ~1:100.

To determine the actual turns ratio, connect the primary to 1000Hz sine wave from an audio signal generator that has an output impedance of 50Ω or lower. With no load on the secondary, measure the rms voltage at the primary and secondary. The turns ratio is Vsec/Vpri

 

[dr pepper]

You can make a really simple blocking oscillator with just a resistor and a transistor as well as a trans, is this what you meant?

 

[Ishanhuss]

Hi Mike, dr pepper,

Thanks for the respond. I have attached a drawing of the said circuits. I'm not sure if it's a push-pull converter as Mike suggested. Oh yes MIke, thanks for the attached reference you include. It helps me to understand more.
Since transformer are base on mutual induction of alternating current, there is a capacitor tied to the input side of 5Vdc reg supply.

dr pepper, I'm trying to make a similar oscillating transformer (that's what the manual refer it to) base on existing working ones.



Thanks again guys. Appreciate it very much.

 

[MikeMl]

Not a push-pull, but single-ended. C8 is for bypass, energy storage. R14 is for voltage drop and filtering. Q4 makes a blocking oscillator with out-of-phase feedback from lower half of the transformer primary. C7 affects frequency, but is there mostly for spike suppression. R15 affects amplitude. CR2-3 C5-6 form a voltage doubler (reduces necessary turns-ratio in T1). R13 Z1-5 form a shunt regulator.

This supply delivers high-voltage, but only a tiny current.

ps: found this phrase in an old patent: Miniature transformer TRIAD TY202X Primary-Secondary turns ratio 1:200 Primary Winding Feedback ratio 50:1.

This appears to be an obsolete transformer no longer made by Triad.

 

[dr pepper]

A classic blocking oscillator, or as now popularly called 'joule thief'

If you can get the same trans then you'll be able to make a copy, even if you cant its still workable, if you can still get the data for the original trans you ought to be able to get something similar, blocking oscillators are usually forgiving on parameters.

 

[Ishanhuss]

HI Mike, dr pepper,

Thanks again guys. I follow up to the patent literature Mike and found the origin of the text. The trans was meant for flickering neon light for it's high voltage output. Now for my attempt to make a similar trans, the transformer maker needs spec such as VA, input and output voltages. With info of Primary secondary turn ratio 1:200, primary winding feedback of 50:1, input voltage of 5Vdc, would that be sufficient spec? I measure a working trans an inject a 5V on the primary side, without connecting any other component, what I get was 36V on the secondary (open circuits). This info, will it be sufficient? What do you think guys?

 

[dr pepper]

Possibly, allthough I'm not sure I get the 50:1 on the feedback ratio, are you saying the driven side of the primary has 50 times the turns as the feedback side, that sounds a bit high, but if thats how the original is.

5v to 36v might well be ok if you measured the trans with a sig gen or the like, being effectively a switcher turns ratios as per classic mains trannys dont apply in the way you think, in the application there will be more on the 5v side due to flyback.

The winders will need to know the max voltage on the secondary and the difference between that and the primary insulation wise, and they might need to know the max operating temp for the insulation, operating frequency is also important and maybe too duty cycle, ie how longs its going to run for at a time.

 

[MikeMl]

I didn't get the 50:1 feedback statement in the patent, either? I have built blocking oscillators like this one where the primary was just a simple center-tapped winding. In other words, there are an equal number of turns on either side of the center tap.

 

[Ishanhuss]

Hi Mike, dr pepper,
Thanks for the last posting. I didn't manage to cook up the transformer and it cost us about S$16k to replace new units. It was like a spoil mouse and we're changing a new pc.
Triad gave me the spec from their archive but I guess it's a bit too late cause the boss decided to make a purchase for a new unit.
Btw, I'm just curious, I was told that the usual 50/60hz laminated core transformer was not able to handle a 1khz frequency. A ferrite core or that sort was utilized instead. Is it true? Thanks again guys.

 

[dr pepper]

A laminated core might be able to handle 1kc as its not that high, valve amp o/p transformers can go to 20kc.

Switching transformers on the other hand tend to operate around 20 - 40kc upwards so ferrite is popular for them.