This cheap bias-T claims to operate from 10 MHz to 6 GHz. But does it actually deliver?
A bias-T allows DC power to be supplied through an RF signal path, making it particularly useful for powering remote LNAs and active antennas. I tested one of these inexpensive bias-T modules using the Rigol DNA814 VNA, measuring its S21 transmission characteristics. While the upper frequency response was actually okay, the low-frequency performance leaves a lot to be desired.
The original module exhibited a 3 dB cutoff at approximately 164 MHz and a 1 dB cutoff at 328 MHz. Hardly impressive for something supposedly operating down to 10 MHz! Opening the module revealed a 24 nH RF choke. Interestingly, calculating the theoretical insertion loss of a 24 nH shunt inductor in a 50 Ω system produced a curve that almost perfectly matched the measured response. The inductor was clearly limiting the low-frequency performance.
To improve its performance, I replaced the original inductor with a 75 nH WE-KIHC high-current ceramic inductor from Würth Elektronik ( @wuerthelektronik ), part number 744917175.
The results speak for themselves: The original bias-T with its 24 nH inductor achieved a 3 dB bandwidth of 164 MHz to 5.37 GHz and a 1 dB bandwidth of 328 MHz to 3 GHz. After replacing the inductor, the 3 dB bandwidth improved to 55 MHz – 5.8 GHz, while the 1 dB bandwidth expanded to 112 MHz – 3 GHz. That’s a remarkable improvement in low-frequency performance without sacrificing the upper bandwidth!
For my upcoming RF project, I need less than 1 dB insertion loss above 118 MHz. With this simple modification, the inexpensive bias-T now meets that requirement. A great example of how understanding a circuit’s RF behavior can help identify and correct its limitations with a single component change. Pretty cool, right?
#electronics #engineering #technology
source
Dude, where'd you get the extra $10K for that RF network analyzer? 🤔
Please explain RF it's my dream learning it but don't know we're to learn , please make series were you teach it 👏
So, they almost built it perfectly. 😊😊😊
That is a great solution. I have 3 of these that I bought in case I need them for testing future high frequency preamplifier projects. I will make the same modification to mine. Many thanks, they will be much more useful now.
Adding a 24nH inductor is a practical attempt to compensate for the terrible insertion loss. Verifying the broadband impedance match across the RF path is essential after this modification.
But is the circuit still stable (assuming it was before the modification) ?
That layout is kinda sus…. those T traces and stuff, all in all very sus
What's your project…?
can you also further explain what you actually gain. and why this change was important.
Well its not a .5 amp bias t anymore
I also bought a 5-pack, mainly for the enclosures. I heard the variation is quite big, so my experiences might not apply to you, but here they are anyway.
The build quality is just terrible. The SMA connectors are mounted using countersink screws, instead of buttonhead ones. The PCB mounting has the same problem, except that they reused the el cheapo bias-t pcbs, so neither the connectors nor the mounting holes line up with the casing. The PCB is actually thicker than optimal for the case, so the SMA center conductor is bent. Very low build quality there.
When I removed the PCB to fix it up, I found another problem: there is a partial pocket under it! The PCB is not flat against the bottom of the box, there are holes. Also, the box itself is anodized, it's not really conductive. I would guess these two mainly f-up the S11 and S22 curves, altought they are not that big of a deal as that small inductor is. If you plan to re-use those boxes, disassemble them first, and replace the screws with proper ones!
I previously used four pairs of these bias-tees for a cross-dipole antenna array. The failure rate was quite high—three out of the four pairs were defective, specifically suffering from poor, unreliable solder joints ("cold solder" joints) and overall low durability. This may have been due to the poor quality of the components supplied; consequently, I designed my own dual bias-tee circuit for the project, and everything has been running very smoothly since. The project concerns radio astronomical interferometry. 🥰🥰🥰