Magnetic components ( i.e. inductors and transformers ) are important elements used in a wide range of electric and electronic applications such as power supplies, DC-DC converters, filters and RF circuits. For these reasons they should be included as a major ingredient in the course of study for any electrical engineering student.
There are a couple of considerations that should be taken into account when selecting the transformers used in student lab activities based around the Analog Discovery hardware. First there are certain limitations in the hardware. The range of frequencies that both the function generators and scope inputs support is from DC to around 10 MHz. This range includes audio frequencies, most of the common DC-DC converter operating frequencies and although these days when engineers think about RF frequencies they are thinking in the 100’s of MHz if not GHz, recall that the standard AM radio band spans from about 500 KHz to nearly 2 MHz. So it would be advisable to select transformers with useable bandwidths that overlap the most with this range of frequencies.
Also the function generators can deliver a maximum unloaded voltage swing of +/- 5V and up to around 35 mA of current. The internal fixed +/-5V power supplies can supply a similar 35 to 40 mA.
A second consideration is the package configuration of the transformer and how it can be connected to the lab circuit. Using solder-less breadboards for building the circuits is a convenient and common technique which allows the students to build multiple configurations while reusing the same components. These solder-less breadboards require leaded or so called through hole components. Surface mount technology (SMT) components will need to be mounted to some sort of a small PCB that adapts them to a set of through hole pins that can plug into the breadboard.
The lower limit on the usable frequency range of a transformer is generally set by the impedance level of the circuit in question and the inductance of the transformer windings. If we assume the common 50Ω standard as our starting point we can calculate the lower frequency bound based on the published winding inductance from the manufacturer’s datasheets. In some cases the datasheets will specify the usable frequency range. In general it is common practice to pick the reactive component, in this case the inductance, at the lowest frequency of interest, to be at least 4 times the resistive component, in this case the 50Ω source resistance.
Coilcraft offers a couple of possible candidate product families.
The Hexa-Path product family contain six 1:1 isolated windings that can be connected in series or parallel. By connecting the windings in series or parallel, the Hexa-Path components can be configured as inductors, coupled inductors ( common mode chokes ) and transformers. This versatility allows them to be used in a variety of applications, mainly power supplies and DC-DC converters.
Unfortunately for us the Hexa-Path products are only available in SMT configurations. Small through-hole to SMT adapter PC boards will be needed to use these with our solder-less breadboards. The Schematic and board CAD files to adapt the HPH1-XXXX package style for through hole use with breadboards are attached to this blog. The surface mount leads on the HPH2-XXXX package are on standard 0.1” spacing and 0.6” wide so it can actually be soldered to a standard double wide machine pin DIP socket like the 24 pin version shown here. The extra pins can of course be cut off.
24 Pin machine pin 600 mil wide DIP package
The Hexa-Path products come in a number of varieties. To provide the maximum frequency range overlap we will be selecting from the HPH1 and HPH2 high inductance grouping. The winding configuration for the HPH1 and HPH2 Hexa-Path transformers is shown in figure 1. The flexibility of 6 independent windings provides the ability to build a large variety of example Lab circuits.
Figure 1 Hexa-Path HPH1-XXXX winding configuration
Measured1 3dB BW
1 : 1
0.002 – 3.00 MHz
1 : 1
0.016 – 2.00 MHz
1 : 1
0.003 – 5.00 MHz
 The measured bandwidth was done with 50Ω source and ( secondary ) load resistances configured as 1 : 1 ratio for the maximum inductance with 3 windings in series for both the primary and the secondary. See figure 2. Scope Channel 1 was used to measure the amplitude across the primary and Channel 2 was used to measure the amplitude across the secondary.
Figure 2 Bandwidth test circuit
Plot 1 Hexa-Path frequency response from 1 KHz to 5 MHz
Plot 1 shows the amplitude vs. frequency response seen at the primary and secondary windings for the three models tested. The vertical axis is in dBV. The flat portion of the response will be at -6 dB where the resistive portion of the transformed RL is equal to RS. As we can see from the curves in plot 1, the higher inductance HPH1-1400L has the lowest low frequency roll off of the three examples with the HPH2-1600L not very far away. The HPH1-0190L with its much lower inductance, 27uH vs. 202 uH, has a low frequency roll off around 8 times higher. It is also useful to note that at higher frequencies above 1 MHz the response of the primary and secondary amplitudes starts to diverge. So we can conclude that the widest usable bandwidth is from 10 KHz to 1 MHz for the HPH1-1400L and HPH2-1600L.
Both Coilcraft and MiniCircuits also offer a family of wideband RF transformers in 6 pin DIP through-hole packages. MiniCircuits calls their DIP package X65, figure 3. The W38 flat pack surface mount package option appears to be like the X65 package but with the leads not bent down, figure 4. These packages work perfectly with standard solder-less breadboards.
Figure 3 Mini Circuits X65 style package
Figure 4 W38 flat pack surface mount package style
These DIP transformers come in the three configurations shown in figure 5. MiniCircuits also has a three winding configuration shown in figure 6. Configuration 1 is a simple two winding transformer, configuration 2 has a center tap on the secondary winding and configuration 3 has a center tap on both primary and secondary.
Figure 5 Winding configurations
Figure 6 Three winding configuration
Again we want to down select from the range of product options for the lowest minimum usable frequency for the largest overlap with the capabilities of the Analog Discovery hardware. For Coilcraft the inductance and frequency range values in the table are from the datasheet. For MiniCircuits the inductance values are estimated from measurements and the frequency range is from the datasheet.
Tested ( recommended ) model numbers:
1 : 1
0.005 – 100 MHz
1 : 1
0.005 – 100 MHz
1 : 1
0.005 – 100 MHz
1 : 1
0.080 - 200 MHz
1 : 1
0.015 - 300 MHz
2.5 : 1
0.010 - 100 MHz
4 : 1
0.020 - 250 MHz
4 : 1
0.050 - 200 MHz
5 : 1
0.300 - 300 MHz
1 : 1 :1
0.010 - 10 MHz
To verify the manufacturers frequency range claims, tests were done using the Analog Discovery Network Analyzer instrument. The measured bandwidth response was again done with 50Ω source and ( secondary ) load resistances when configured for 1 : 1 ratio transformers, figure 2. RL was adjusted for other transformer ratios as needed such that the amplitude measured by CH1 was the same for each impedance ratio.
The various frequency plots below compare the model numbers from the two tables above. To make the comparisons between the different impedance ratios, the amplitude vs. frequency of the signal seen at the primary winding ( CH1 in figure 2 ) is plotted. For models with center tapped windings, the center tap was left floating. Both step up and step down orientations are plotted for those models. Remember that RL is adjusted such that the flat band amplitude is the same for all cases. Plot 2 shows the frequency response for the higher inductance models including all three from Coilcraft and the MiniCircuits model with T1-6T. As we can see the three Coilcraft parts have almost identical response to the 1:1 MiniCircuits model.
Plot 2 Higher inductance 1:1 models
I also had a number of -6 parts with different impedance ratios. These are plotted in plot 3 when configured as both step-up and step-down transformers. As step-down transformers they all responded much like the 1:1 transformer. As step-up transformers the low frequency cut off point is significantly higher for the 2.5:1 and the 4:1 models.
Plot 3 Higher inductance models as step up and step down
In plot 4 response curves for the lower inductance -1 model transformers are plotted. Again these are plotted when configured as both step-up and step-down transformers. As step-down transformers the 4:1 and 5:1 parts responded with lower minimum frequencies than the 1:1 transformer. As step-up transformers the low frequency cut off point is higher for the 4:1 and the 5:1 models much as we saw in the -6 models.
Plot 4 Lower Inductance -1 models
In the final plot 5 three different configurations for the three winding T-626 transformer are plotted. The step down configuration showed a lower minimum frequency much as the conventional models did.
Plot 5 T-626 3 winding configured as 1:1 1:2 and 2:1
In addition to their use as transformers, these devices can be used as inductors. By connecting the multiple windings in either series or parallel various inductance values can be obtained from the single component.
Formulas used to calculate electrical characteristics of multi-winding transformers:
Manufacturer datasheets list certain electrical characteristics for the devices. Probably the most important for our purposes is the winding inductance. For power conversion applications the DC resistance (DCR), the maximum rms current (Irms), and saturation current Isat are also specified.
Connecting windings in series:
For higher inductance, multiple windings (WN) can be connected in series. As the inductance increases, energy storage and Irms remain the same, but DCR increases and Isat decreases.
Inductance = Inductancetable × (WN)2
Note: this Wn2 factor is only valid when the coupling factor between windings is exactly ( or very nearly ) one. A more general formula is LT = L1 + L2 + 2M
DCR = DCRtable × WN
Isat = (Isattable × 6) ÷ WN (connected in series)
Irms = Irmstable
Where Inductancetable , DCRtable , Isattable and Irmstable come from the manufacturer’s datasheet.
Connecting windings in parallel:
To increase current ratings, multiple windings (WN) can be connected in parallel. DCR decreases, current ratings increase, and inductance remains the same.
Inductance = Inductancetable
DCR = 1 ÷ [WN × (1 ÷ DCRtable)]
Isat = (Isattable × 6) ÷ WN ( connected in parallel )
Irms = Irmstable × WN
In conclusion, the 6 winding Hexa-Path transformers provide the most flexibility and a good low frequency cutoff point but are not very good above 1 MHz. They are, however, only available in surface mount packages. The wide band RF transformers can also provide a similar low frequency cut off point for the high inductance models and have high frequency performance well beyond the upper limit of the Discovery hardware. They are available in multiple configurations including the three winding configuration T-626 from MiniCircuits. And importantly they are in through hole DIP packages ready to use with solder-less breadboards.
As a final comparison the three best choices based on minimum frequency are plotted in plot 6. Based almost entirely on the DIP package options but slightly on performance above 1 MHz, I think the T1-6T, WB1010 (or WB2010, WB3010), and T-626 are better choices over the SMT HPH1-1400L.
Plot 6 All three have similar usable bandwidths
As always I welcome comments and suggestions from the user community out there.
The CAD files for the HPH1 footprint SMT to 12 pin DIP PCB are attached here: