autoformers

Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic

   Hi
wondering if anyone knows any good references on using autoformers/autotransformers rather than L-pads to match driver sensitivity. seems potentially useful - particularly as you can boost sensitivity as well.

particularly I am wondering:

though working out the turns ratio is simple enough - what determines the total number of turns?

I've seen references to using the autoformer both to adjust the impedance and to act as an inductance making up part of the crossover - how does this work?

why is this technique so rarely used? is it just cost?

cheers
Phil

Author	Message Topic Search Topic Options Post Reply Create New Topic Printable Version Translate Topic
snowflake Members Profile Send Private Message Find Members Posts Add to Buddy List Old Croc Joined: 29 December 2004 Location: Bristol Status: Offline Points: 3120	Post Options Post Reply Quote snowflake Report Post Thanks(0) Quote Reply Topic: autoformers Posted: 24 May 2016 at 9:43pm
	Hi wondering if anyone knows any good references on using autoformers/autotransformers rather than L-pads to match driver sensitivity. seems potentially useful - particularly as you can boost sensitivity as well. particularly I am wondering: though working out the turns ratio is simple enough - what determines the total number of turns? I've seen references to using the autoformer both to adjust the impedance and to act as an inductance making up part of the crossover - how does this work? why is this technique so rarely used? is it just cost? cheers Phil

_djk_ Members Profile Send Private Message Find Members Posts Add to Buddy List Old Croc Joined: 23 November 2004 Location: United States Status: Offline Points: 6002	Post Options Post Reply Quote _djk_ Report Post Thanks(0) Quote Reply Posted: 25 May 2016 at 3:46am
	" what determines the total number of turns?" The required inductance. " is it just cost?" Yes. Klipsch uses very lage inductance value autoformers, and they parallel them with the smaller (correct) value inductor the crossover requires. JBL frequently used the exact value autoformer the crossover required, sometimes with taps for level adjustments (which also required resistor and/or capacitor changes as well, via a switch bank).
	djk

snowflake Members Profile Send Private Message Find Members Posts Add to Buddy List Old Croc Joined: 29 December 2004 Location: Bristol Status: Offline Points: 3120	Post Options Post Reply Quote snowflake Report Post Thanks(0) Quote Reply Posted: 25 May 2016 at 9:45pm
	using the Klipsch method can you choose whether to put the crossover inductor on either the primary or secondary side depending on whether you are boosting or cutting to keep the coil value low? using the JBL method do you find the inductance by designing the autoformer for the appropriate power, minimum frequency, and primary voltage? and then select a shunt resistor to make the crossover load correct for the inductance and desired crossover frequency? if you want fine adjustment could you not put a few dB L-pad on the driver rather than messing around with multiple taps?

rodneps Members Profile Send Private Message Find Members Posts Add to Buddy List New Member Joined: 10 June 2016 Status: Offline Points: 7	Post Options Post Reply Quote rodneps Report Post Thanks(0) Quote Reply Posted: 10 June 2016 at 10:13pm
	I have used the inductance of the auto as part of the filter . One can start with designing the filter with a simple inductor, and then add turns to reduce the output (assuming the tweeter sensitivity needs to be brought down to the LF driver's level) . The input cap value is reduced by (approximately) the square of the turns ratio. The filter response is also affected by the coupling coefficient between the windings, One great advantage of using an auto transformer is that one gets a higher input impedance if you are using in in the 'step-down' mode Derek

snowflake Members Profile Send Private Message Find Members Posts Add to Buddy List Old Croc Joined: 29 December 2004 Location: Bristol Status: Offline Points: 3120	Post Options Post Reply Quote snowflake Report Post Thanks(0) Quote Reply Posted: 14 June 2016 at 2:13pm
	thanks for the replies - few questions, sorry if they are basic but I'm having trouble finding anything that explains this topic well. Is the idea of the higher input impedance that it increases the 'damping factor' of the system i.e. the ratio of the load impedance to the output impedance of the amp (plus cable resistance)? should the Re of the driver be understood as part of the load impedance or should it be lumped in with the cable resistance and therefore part of the output impedance? when using an L-pad or series resistor to attenuate a tweeter is this in effect lowering the damping factor to below one - the series resistor is greater in value than the load resistance? what is an acceptable ratio of load resistance to output resistance? could you use an off-the-shelf 100V step-down transformer to do most of the attenuation and then some low value resistors to fine tune the level?

rodneps Members Profile Send Private Message Find Members Posts Add to Buddy List New Member Joined: 10 June 2016 Status: Offline Points: 7	Post Options Post Reply Quote rodneps Report Post Thanks(0) Quote Reply Posted: 14 June 2016 at 11:55pm
	!00v line transformers are normally not particularly high quality, I personally would avoid them. I don't think damping factor is very relevant to tweeters where the fo should be at least an octave or so below crossover frequency (unlike bass drivers where the fo in in the passband & part of the LF alignment) but I would feel that the crossover slope should be such as to make the level at tweeter fo at least 20dB down on the main tweeter level. The starting point would be the crossover frequency required, the amount of attenuation needed and the slope needed to match the acoustic slope of the LF driver with it's crossover. Auto transformers will generally give better attenuation at tweeter fo because of their lower output impedance, but I have used both those and resistive attenuation to good effect depending on the particular design. Derek

_djk_ Members Profile Send Private Message Find Members Posts Add to Buddy List Old Croc Joined: 23 November 2004 Location: United States Status: Offline Points: 6002	Post Options Post Reply Quote _djk_ Report Post Thanks(0) Quote Reply Posted: 15 June 2016 at 1:30am
	It does sound different to use an autoformer rather than a straight L-pad. The main problem being cost and the lack of fine adjust-ability, ±1dB is the best I have seen.
	djk

rodneps Members Profile Send Private Message Find Members Posts Add to Buddy List New Member Joined: 10 June 2016 Status: Offline Points: 7	Post Options Post Reply Quote rodneps Report Post Thanks(0) Quote Reply Posted: 15 June 2016 at 5:57am
	It depends on how much attenuation you need, whether an L pad design gives too low an impedance, and the specific crossover circuit you want to use. On recent designs I have used an Auto plus a low value L pad to give fine adjustment. When you say cost, do you mean cost of buying an auto ? Derek

rodneps Members Profile Send Private Message Find Members Posts Add to Buddy List New Member Joined: 10 June 2016 Status: Offline Points: 7	Post Options Post Reply Quote rodneps Report Post Thanks(0) Quote Reply Posted: 15 June 2016 at 6:12am
	It depends on how much attenuation you need, whether an L pad design gives too low an impedance, and the specific crossover circuit you want to use. On recent designs I have used an Auto plus a low value L pad to give fine adjustment. When you say cost, do you mean cost of buying an auto ? Derek

rodneps Members Profile Send Private Message Find Members Posts Add to Buddy List New Member Joined: 10 June 2016 Status: Offline Points: 7	Post Options Post Reply Quote rodneps Report Post Thanks(0) Quote Reply Posted: 15 June 2016 at 7:05am
	I am guessing this is a one off project. Do you have test equipment to verify your design? Do you have a crossover circuit you are looking to adapt with an auto-transformer? It's not too difficult to take as existing inductor and add extra winding to make it an auto-transformer, especially if wound on a bobbin. BTW, sorry for double post.

snowflake Members Profile Send Private Message Find Members Posts Add to Buddy List Old Croc Joined: 29 December 2004 Location: Bristol Status: Offline Points: 3120	Post Options Post Reply Quote snowflake Report Post Thanks(0) Quote Reply Posted: 15 June 2016 at 6:43pm
	looking at re-designing the crossovers for my unity horns. the present crossover has a third order HPF and two resistors in series and parallel to attenuate the tweeter level. it also has a protection bulb in series with the tweeter. if I switched to a second order crossover I could use an autoformer and reduce the amount of heat dissipated by resistors. using the autoformer as the filter inductor means there isn't much extra cost. wondering how the protection bulb will impact on the filter behaviour. in the current design there is a resistor in parallel with the tweeter and bulb which stabilises the impedance seen by the filter. is a similar effect possible when using the autoformer as part of the filter. can some of the advantages associated with an autoformer be achieved by modifying a standard crossover: making the series resistor much higher (~22R) and selecting a parallel resistor to get the correct attenuation will result in a high impedance load; rearranging the components so that the series resistor is on the amp side of the filter but the parallel resistor is still across the tweeter - the tweeter now has the filter inductor parallel across it without passing through a high value resistor. will this help damp back emf?

rodneps Members Profile Send Private Message Find Members Posts Add to Buddy List New Member Joined: 10 June 2016 Status: Offline Points: 7	Post Options Post Reply Quote rodneps Report Post Thanks(0) Quote Reply Posted: 16 June 2016 at 8:17am
	You could still use a 3rd order filter. You would need to reduce the input cap by the square of the auto turns ratio. For example, if your present inductor had 100 turns & you added another 100, you would get 6db attenuation and need an input cap 1/4 of the original. If the bulb is in series with the horn the effect should be very similar as the current through the horn should be the same for a given response, that may not be true if the resistor is in parallel with the horn and after the bulb, I'm not sure from your description if that is the case. Changing to a high input resistor will not have the same effect as the auto, you will split the dissipation but not reduce it much. It will also Chang the response quite a lot as the filter now sees a high source impedance. You can are correct about the better damping using the auto, although with the 3rd order you still have the output cap impedance, but that is a trade off with the higher attenuation slope.