For one of our current development projects, it was necessary to find a transformer capable of coupling signals into a circuit over a wide bandwidth. The requirements for this transformer were to provide linear transmission from the lowest possible frequency (approximately 10 Hz) up to a frequency of 10 MHz. We initially ordered the commercially available SSR10VS/HS-10135 choke. To test whether it would also function below the frequency of 1 kHz specified in the data sheet, we needed to measure it.
Bode diagrams via USB
For the simple, digital measurement of these Bode plots, an Analog Discovery 2 from Digilent was used. This is a digital USB multifunction device that, in addition to its two oscilloscope channels, includes two signal generators, a programmable laboratory power supply from 0.5V to 5V, a logic analyzer, and several other features.
Setting up the Discovery 2 turned out to be surprisingly easy once you've got the relevant software and documentation. It's largely self-explanatory, making it "plug and play." After plugging in the Discovery and installing the associated WaveForms software, the measurement process becomes almost intuitive, saving you the trouble of connecting a real oscilloscope to your PC via a network.
In order to be able to use oscilloscope probes or BNC cables in general on the Discovery 2, we used an associated expansion board on which the two oscilloscope channels and the two signal generators are implemented as a single BNC socket.
Measurement results of the DIY transformer
When measuring the commercial choke, we were able to confirm the frequency response from the datasheet, but at the same time, we discovered that below the frequency shown in the datasheet, it didn't have the linear relationship we were looking for. Researching various electronics blogs, we came up with the idea of winding the required 1:1 transformer ourselves from twisted-pair cables. Since we still had old network cables and coil cores lying around from other projects, the next logical step was to replicate the approach we had found. After a bit of winding work and a toroidal transformer with a total of 40 windings, we measured this new homemade component again with the Discovery 2. This completely fulfilled our expectations:
To increase the stability of our new transformer, we also gave it a plastic housing with BNC connectors:
Conclusion
During the development phase of a product, when the physical parameters are not yet entirely certain, coils and transformers are particularly suitable for self-winding. This offers the opportunity to replace or modify windings or toroidal cores, or generally adapt them to the final required parameters. This example demonstrates once again that a DIY solution is by no means inferior to an industrially manufactured component. Furthermore, the existing transformer is even more cost-effective than a store-bought equivalent.
If you have any questions about our solution or need support with a similar project, please contact us!