Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
My PA0FRI active loop receiving antenna
|
On Thu, Jan 30, 2025 at 01:05 PM, <biastee@...> wrote:
Hi Leong I looked at? figure 7 "Applying an unbalanced source to a balanced input" of your above referenced standards document. (i did not find a page 64)
There is clearly a requirement to put the source impedance of the unbalanced source als R1. The common mode source impedance of a piece of loop wire with reference to earth/ground is the same as with a short monopole (rod or Whip). Z is high, at low frequencies nearly a pure reactance (-jX is a capacitor of a few picorads), not 50 Ohms. ?
As the source impedance the you need to take an antenna simulation network for a very short rod, as described in Figure 5.? It ist sufficient to replace R1 by a 50 Ohm source and a capacitor of 5 ....15 picofarads in series, which simulates the Hi-Z capacitive Impedance.
?
regards, Fred |
|
Hi Fred,
?
The signal source in the document refers to a test equipment, not the loop aerial (which resembles a short rod where common mode signal is concerned). The figure caption states that the source? impedance is assumed to be 600 ohm - this is a common value for antique LF signal generator. The purpose of the shown network is too provide (resistive) matching between the signal generator & DUT input. |
Hi Fred,
Digression: The Fig. 5D network is very familiar to me as I once toiled in a factory producing car radios. That network's series cap C4 is specified as 15 pF to represent the capacitive reactance of a 1m monopole (electrically very short at medium wave BC) mounted on a car boot, while the shunt cap C5 (60 pF) represents the capacitance of the coax line connecting aerial to the car radio. The resistors R1 & R2 are there to provide a reasonable match to the sig-gen. |
|
On Fri, Jan 31, 2025 at 06:08 AM, <biastee@...> wrote:
The signal source in the document refers to a test equipment, not the loop aerial (which resembles a short rod where common mode signal is concerned). The figure caption states that the source? impedance is assumed to be 600 ohm - this is a common value for antique LF signal generator. The purpose of the shown network is too provide (resistive) matching between the signal generator & DUT input.That's my point, Leong ?
you measured the CMRR of the amplifier standalone, not the CMRR of the active-loop antenna.
?
reagrds
Fred
? |
|
Ah OK thanks, my bad.
?
Regards,
?
Martin
?
On Fri, Jan 31, 2025 at 02:05 AM, <biastee@...> wrote:
Have you got a weblink, or can you upload a scan of the picture ? |
|
I just made some quick measurements on some ready built amplifiers I found in my scrap pile.
?
First, a gain comparison using a loop simulator circuit between a LZ1AQ and two G8CQX (same as PA0FRI) using different transistors, with high and low values of hfe.
?
?
Second, a comparison of CMRR between the LZ1AQ and the higher gain G8CQX. The amplifier inputs were shorted, and the signal fed to the short via a series 15pF capacitor with a 50 ohm load on the generator output. I moved the exact point of connection of the capacitor by a few mm at most, to demonstrate how easy it is to obtain a change in CMRR. The variation is very noticeable, even with a slight alteration in layout, and in a fairly well controlled workshop environment
?
?
Although the G8CQX, offers much better CMRR at the lower frequencies, the differences at HF are much less apparent. Probably due to strays and circuit balance becoming more difficult with increases in frequency. The suggested reduction in CMRR, due to the increase in input impedance of the LZ1AQ is apparent, and would probably be worse with an actual loop.
?
However, I suspect that in real life with a loop connected, the overall differences between the two types of circuit would actually be minimal, as the opportunity for loop imbalance, due to interaction with nearby objects, would be significant.
?
My previous comment that a practical value of CMRR of around 20 to 30 dB is probably the limit, still seems to be a good estimate.
?
Regards,
?
Martin |
|
Hey!
Thank you Steve, Leong and other folks. Steve - Your skills and confidence are super inspiring - thank you for jumping into the problem (i.e. gain drooping with original LZ1AQ design with 2N2222s) and solving it so quickly while retaining the usage of two PZT2222A transistors! The usage of two PZT2222A transistors reduces the risk of PCB design failures to a fair extent. This is my first attempt at building the LZ1AQ amplifier. I am looking for ideas on simplifying the circuit to its essentials and thereby reducing the PCB size. Currently the PCB will fit inside a 100x100mm IP65 box just fine. Also, can this 2-Layer PCB design really work with BFU590 transistors on it? Folks are telling me 'horror stories' of how these GHz fT transistor devices like to oscillate. Cheers, Dhiru |
|
On Fri, Jan 31, 2025 at 12:47 PM, Dhiru Kholia wrote:
Also, can this 2-Layer PCB design really work with BFU590 transistors Hi Dhiru,
please do not believe in horror stories from folks who have no real practical experience.? I've built several LZ1AQ type? circuits with BFU590G and BFU590Q on a two layer board without oscillation problems. Sure, the pcb-layout must follow basic RF-circuit design and grounding rules.? A PCB size? of 100 mm x 100mm seems to be a lot of space, i doubt, you need it. Best is to keep the pcb-connections as short as possible and use one layer as a pure ground layer. I recommend to use Everett Sharps implementation as an orientation, it is a proven design. Instead of the high fT BFU590 transistors you can use the discontinued, but more robust 2SC5551A. It's hard to get genuine ones, but the chinese counterfeits normally do a decent job.
?
?
regards, Fred
?
?
?
? |
|
The 1uH inductors, in series with the base connections of the output pair, help too.
?
Design the PCB as if it was for operation at VHF / UHF, with plenty of plated through holes, and you shouldn't have any issues.
?
Regards,
?
Martin
?
On Fri, Jan 31, 2025 at 01:21 PM, Fred M wrote:
I recommend to use Everett Sharps implementation |
|
There is also the 2SC3357 as mentioned by Leong with an fT of 6.5 GHz. They are available from Digikey and $0.27 in quantity 1. They re in SOT-89 which is nice for PCB layout since many other options are available in the same package. I will try them in the LZ1AQ amp in the near future.
--
=================================================================== Mike M |
|
On Thu, Jan 30, 2025 at 10:46 PM, Fred M wrote:
Yes, pre-amp only.? BTW, I thought the whole group knows Martin & I have been long trying to eliminate the loop aerial while characterizing the pre-amp, by using a dummy aerial.
?
For the life of me, I can't understand why would anyone want to include the loop in the characterization when comparing different pre-amps.
?
?
Hi Dhiru, I don't know if in Fred's mind that I fit the category of "folks who have no real practical experience". IMHO, the risk is real and GHz transistors deserve their notoriety. Most people don't notice because the instability can occur way above the limits of their spectrum analyzer, e.g. Ka band and in temperature extremes. Oftentimes, I discovered it only after having connected a power sensor (the sensor can respond to frequencies above its specified maximum, though un-calibrated).?
?
?
?
Yes, I experienced the same variation too. A small difference in the length of the wires connecting the test source to the balanced input terminals resulted in a different reading. In the case of M0AYF and PA0FRI, I minimized the variability by widely opening the red crocodile clip so that it could contact both terminals (photo below)
/g/loopantennas/photo/300229/3881034
?
?
? |
|
On Fri, Jan 31, 2025 at 02:58 PM, <biastee@...> wrote:
BTW, I thought the whole group knows Martin & I have been long trying to eliminate the loop aerial while characterizing the pre-amp, by using a dummy aerial.
?
That was - in my understanding - to assess the h-field differential mode performance of the loop amplifier.
Now we talk about the common mode rejection, which needs another kind of dummy aerial, which resembles the source impedance for e-field common mode signals with reference to earth/ground. I believe the CMRR measurement in a 50 Ohm environment is not the same as the CMRR from a Hi-Z reactive source impedance.
And there is still the question to answer, how good the CMRR of an active loop must be. If a CMRR between 20 to 30 dB is sufficient, the high CMRR value of the PA0FRI differential Amplifier is not advantageous for the practical receive performance. ?
I fully agree. But there is no reason to become scared by horror stories about the use of BF590 transistors. I firmly believe, if Dhiru follows some basic RF design rules, his LZ1AQ amp will work stable.
|
|
Very high ft is a problem with these '590 BJTs.? But we all know that.? I'll offer a tried-and-true "most-of-the-time" solution:? Instal a 1/4 or 1/8-watt - better yet, an SMD - 22 to 68 ohm resistor in series with the collector/drain.? If you use a leaded resistor, trim the leads to "zero" length and install it as close to the ground layer as possible.? I've used this trick for decades and it tames "most" of the high ft BJTs and FETs. In addition, you should have a spectrum analyzer or other means of verifying there is no oscillation at or below the rated ft.? Without that, you may end up with a "parametric amplifier" which can fool anyone regarding its exaggerated "performance".???? Dave - W?LEV On Fri, Jan 31, 2025 at 12:47?PM Dhiru Kholia via <dhiru.kholia=[email protected]> wrote: Hey! --
Dave - W?LEV |
|
I've just made some more tests.
?
The first thing I need to clarify is that the CMRR I measured in this graph
?
?
doesn't use the same signal reference level as the gain plots in this graph, which has an additional 30dB of attenuation in the feed from the signal source.
?
?
The losses through the loop simulator circuit and the 15pF / 50 ohm whip simulator, are not the same, either. So it's difficult to equate the two in terms of absolute CMRR at any given frequency. However, it's clear that the G8CQX in comparison to the LZ1AQ has better CMRR at the lower frequencies.
?
As an experiment, I tried adding a Coilcraft 1:1 transformer onto the LZ1AQ input, to see if this would improve the CMRR, but it actually made it worse. Placing a common mode choke on the RF cable feeding the amplifier, did improve the CMRR on the lower frequencies by about 10 to 15dB, but it didn't make much of a difference at the higher frequencies.
?
To be honest, I'd thought that adding a 1:1 transformer on the amplifier input would have made a fairly large change, by improving both the isolation and balance, but no matter how much I messed around with the configuration, I couldn't make it any better than the same amplifier, without a transformer input.
?
Regards,
?
Martin
?
On Fri, Jan 31, 2025 at 12:38 PM, Martin - Southwest UK wrote:
I just made some quick measurements on some ready built amplifiers I found in my scrap pile. |
|
With regards to how much CMR is required, it should be remembered that a loop connected to an amplifier even with zero CMR will still show some nulls. The nulls may not be deep or exactly through the axis of the loop however. Any CMR that the amplifier offers will be added to the loops side rejection.? |
|
Hi Tom,
?
This was behind my thinking about CMRR, and how much can actually be achieved in practice.
?
My observation is that a maximum null depth of 20 to 30 dB is about the limit for a stable null, which is the basis for my CMRR guess estimate.
?
Regards,
?
Martin
?
?
On Fri, Jan 31, 2025 at 05:03 PM, Tom - VE3PSZ wrote:
With regards to how much CMR is required, it should be remembered that a loop connected to an amplifier even with zero CMR will still show some nulls. |
|
It should be remembered that the CSA Aperiodic loop design (in the files section), deliberately introduces an E-Field component, in order to unbalance the amplifier, and change the polar diagram from a standard "figure of 8" to a much more directional cardioid shape.
?
Regards,
?
Martin
?
On Fri, Jan 31, 2025 at 05:03 PM, Tom - VE3PSZ wrote:
The nulls may not be deep or exactly through the axis of the loop however. |
|
Yes and the CSA Aperiodic loop does work. I tried it myself. Nice cardioid pattern. But a word of caution just before everybody starts cutting into their loop antennas. It may not be obvious from the literature, but adding the top resistor compromises LW sensitivity and completely destroys the VLF frequency response. MW is affected but still usable, and HF is relatively untouched. |
|
Agreed, which is why I have shorted out the top resistors in mine, and replaced the amplifiers with Wellbrook clones or LZ1AQ's.
?
The CSA's are only really useful when deployed as designed, and used in an array.
?
Regards,
?
Martin
?
On Fri, Jan 31, 2025 at 06:44 PM, Tom - VE3PSZ wrote:
Yes and the CSA Aperiodic loop does work. I tried it myself. Nice cardioid pattern. But a word of caution just before everybody starts cutting into their loop antennas. It may not be obvious from the literature, but adding the top resistor compromises LW sensitivity and completely destroys the VLF frequency response. MW is affected but still usable, and HF is relatively untouched. |
|
Thanks for all the tips, info and help!
I have access to TinySA Ultra but I believe monitoring my power supply for unexpected current consumption might work better to detect oscillations. I was reading the following page earlier today: ... I am attaching the PCB files to this email to aid the review process. The complete set of files is open at the following link: Cheers, Dhiru |
to navigate to use esc to dismiss