1:1 Coupling Transformer 20Hz to 10KHz

Why do you need a transformer? Can you float the load?

10W is a lot, could you pick off some and use a lower power audio transformer?

George H.

I don't understand the 1:1 thing either. Couldn't one just use two audio transformers connected "back to back" 1:whatever -> whatever:1? If isolation isn't required then I don't think the response of the transformers has to be particularly linear; one could just take feedback from the "secondary" to the driver stage to linearize it.

But if isolation isn't needed, then I don't understand the 1:1 requirement...

Audio transformers have a purpose. Coupling 1:1 at higher powers in the 8R load range isn't one of them. It just makes more sense to drive directly, or to couple at a lower-power signal level, before power is involved. Plenty of 1:1 transformers for 600 ohm line.

If this isn't audio, then perhaps you might reconsider bandwidth and flatness. The LF end will be the hardest spec.

RL

RL

as always it would help if you told us what you're trying to do. Otherwise one can only really make widely differing suggestions at random.

Another semi-random suggestion: an audio transformer kit.

NT

There's a better trick than that. Use a 1:1:1 transformer - drive the first winding, compare what you see on the second winding with the signal you ar e trying to send, and use local feedback to control the current through the first winding to get the voltage output you want from the third - isolated - output winding.

It's nowhere near as good if you are pulling a lot of current from the outp ut winding, but it does allow you to lose the magnetising current from your error budget.

In the ideal case, only the first winding needs to be wound with heavy wire to carry the magnetising current, and the other two windings can be much l ess bulky.

Me either. Could be a DC offset problem, but there are much cheaper ways of dealing with that.

--
Bill Sloman, Sydney

I think some of the better spring reverb tank drivers in guitar amplifiers use a setup sort of like that. They have a current sensing resistor in series with one of the transformer primaries to ground, and do transresistance(?) feedback to the driver stage.

maybe try a 50VA toroidial power transformer with a split 10V secondary. insulate the primary and leave it unconnected.

--
  \_(?)_

** That will work very well.

Give you 25Hz to 50kHz response at least, with exact 1:1 turns o.

..... Phil

** Not correct.

What IS commonly done is to connect the drive transducer coil inside the feedback loop of an op-amp - creating current drive and hence a much flatter response.

Reverb systems only have transformers when a valve is being used to drive the transducer - for impedance matching. Usually 15kohms to 8ohms.

.... Phil

Thanks. That is the best suggestion so far.

David King

Ah I gotcha. I was a little confused by some of the solid state reverb amp schematics I saw - the reverb driver is not driving a transformer, but that coil is internal to the tank and they're wrapping current sensing feedback around the input transducer.

--


----Android NewsGroup Reader---- 
http://usenet.sinaapp.com/

1:1 isn't odd, it's just a unlikely requirement above some 100s of milliwatts. Sometimes it's just the quickest way to deal with transmission ground loops, emi and hum.

RL

Interesting, thanks. I would never have guessed that an AC transformer would work up to 50 kHz.

George H.

ould

It can work fine. It can be that you need to feed it with a lot of high fre quency current to sustain all the current circulating in the shorted turn i nside the transformer core, but as long as you can do that, the flux throug h the core will follow the voltage applied to the input winding and generat e the same voltage on the output winding.

The core and the input winding can get warm.

You can end up having to provide even more primary current to drive the int er-winding capacitances of the input and output windings, but a sufficientl y low impedance driver can cope.

We had fun with this on the scan drive for an electron-beam microfabricator once. My boss - a very clever man - hadn't been fully exposed to the trans former equations, and was rather happy with their implications when we work ed it all out.

--
Bill Sloman, Sydney

bill

** What " shorted turn" ?

Toroidal types work as well or better than E-cores as wide band audio transformers, without the need for layered windings.

.... Phil

There's always a conducting loop inside a transformer core. Using a laminated core helps, using a ferrite core helps more, and using a nickel-zinc high resistance ferrite is more or less essential if you want to go up to MHz ferquencies.

Toroidal cores are frequently wound on a tape core, in which the high permeability core material is formed as a thin ribbon before being wound as circular core. Search on "metallic glass core materials"

for the extremely expensive stuff, which may be peddled to audiophools.

--
Bill Sloman, Sydney

bill

** "Toroidal types work as well or better than E-cores as wide band audio transformers, without the need for layered windings." ** The context is supply frequency toroidal transformers - wound from grain oriented, silicon steel strip. E-core transformers for audio frequencies ( valve output etc) typically use the same material.

I am at a loss to know just what your point is.

Pun intended.

.... Phil

o.

transformers, without the need for layered windings."

in oriented, silicon steel strip. E-core transformers for audio frequencies ( valve output etc) typically use the same material.

The point is that there is a frequency dependent loss in any transformer co re.

Supply frequency transformers are designed to make this low enough not to m atter much at 50/60Hz. The loss will be a lot higher at higher frequencies. This won't stop you getting high frequency output from them, but you will have to drive the input coil progressively harder to get the same output vo ltage swing at progressively higher frequencies.

You haven't given us any kind of explanation why wound toroidal cores would work better at high frequencies than E-I laminations. You don't need to, b ut some kind of explanation would be interesting. It could be that it's pra ctical to use a thinner steel tape in a wound toroidal than it is in stacke d E-I laminations, but that's pure speculation.

--
Bill Sloman, Sydney

Hi Phil, your answer to use a toroid is fantastic (I just wanted to say that).