Testing frequency response of a transformer

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

letterbeacon

Well-known member
Joined
Jul 12, 2009
Messages
642
Location
London, UK
I have a few 600:600 input transformers and I'd like to check their frequency response.

Is it as simple as connecting my interface with the terminals on the primary and secondary of the transformer and running something like REW through them?
 
I would suggest
Hook up your transformers with the intended circuit and load.
Have a look on scope while injecting the circuit with a signal generator.
What does a squarewave at 5khz look like?
Is it free from overshooting and ringing? Then do a sweep.
See if the response is flat and at what point you have -3db.
Also if the output is balanced, check symmetry between the ends of the secondary.

If something looks strange, try to treat it with resistive loads and/or capacative loads like zobel nets.
And then measure freq response once again.

Hope that helps.
 
5v333 said:
Hook up your transformers with the intended circuit and load.

But what if the transformer is used to balance the output of a signal gen?  In that case we don't know the intended circuit and load.
 
For a basic test just using REW in a loop is fine.  In addition it's also a good idea to hook up a few different loads to see how it will behave in the real world.  For example 10k, 600 ohm,  and also some capacitance to simulate a long cable run.
 
warpie said:
But what if the transformer is used to balance the output of a signal gen?  In that case we don't know the intended circuit and load.

well you just have to find out what the generators output Z is then and what load the transformer will be hooked up with!
 
the bandwidth between an ext source and the transfomer coupled input could be around 100khz-1mhz.

its been some time i used REW now but isnt it restricted to 44.1 and 48khz FR?
you wil only see what happends up to nykvist (22.05khz and 24khz)
 
warpie said:
But what if the transformer is used to balance the output of a signal gen?  In that case we don't know the intended circuit and load.

The OP was asking about testing input transformers in which case you presumably do know the load.

If you want to test an output transformer then you need to know the driving source as a minimum.

In both cases you also need some idea of the likely maximum level the transformer will be expected to handle and then test it at that level.

As a rule, you need to make some choice as to driving and load impedances as well as operating level but using REW to make measurements is very instructive. You can easily see the harmonic dostortion structure as you approach saturation.

Cheers

Ian
 
REW works away fine at 96khz for me , as long as your interface can do 96k  REW can too.
 
A set of variable series and shunt resistances to modify input and load Z are useful with unknown transformers, or to determine how far away from nominal is acceptable.  Test with the highest bandwidth available. 

You will probably find not much significant response deviation with 600:600 types, with level handling capabilities and phase response at the edges being more responsible for sonic differences. 
 
warpie said:
But what if the transformer is used to balance the output of a signal gen?  In that case we don't know the intended circuit and load.
Most xfmrs have a response that varies significantly dependant on the source and load impedances; that's why some manufacturers recommend a specific load impedance or provide for a switchable internal termination.
 

Latest posts

Back
Top