|
Components and Circuits For discussions about component types, alternatives and availability, circuit configurations and modifications etc. Discussions here should be of a general nature and not about specific sets. |
|
Thread Tools |
30th May 2017, 9:55 am | #21 | |
Octode
Join Date: May 2011
Location: Lancashire, UK.
Posts: 1,349
|
Re: RF signal path in this circuit...
Quote:
Draw what you want and we can try it but I'm not sure it will prove much tbh. |
|
30th May 2017, 10:10 am | #22 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: RF signal path in this circuit...
The base is grounded right up to the frequencies where the 100nF decoupler goes inductive and its Z starts to rise, then things get very dodgy. The Miller-positioned capacitor on the other side is a lot more fun.
By the way, the Miller effect is still effective on the right hand side, but where that decoupler is good the voltage gain ought to be rather low. Anyway, this circuit is enough of a mess that it would be easier to start again with a new design rather than try to fix this one. It looks like it's already had more fixing than any circuit should. David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
30th May 2017, 10:47 am | #23 | ||
Dekatron
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
|
Re: RF signal path in this circuit...
Quote:
Quote:
Thank you for your analysis and comments, I'm getting the complete picture now. I recently started to analyse all RF circuit designs that are not from a known trusted source and come up with anomalies like this quite often, but have only partial insight into the whole shebang. I enjoy the process but I still rely on a lot of external expertise . DC operating conditions are challenging enough, let alone RF design and best practice! Another anomaly was the choke value shown ...wild parasitic oscillations at 3Mhz occured. 220 uH works and is what I am using , although of course changing the operating conditions of that output transistor in an unquantified way!
__________________
Al Last edited by Al (astral highway); 30th May 2017 at 11:14 am. |
||
30th May 2017, 11:05 am | #24 |
Dekatron
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
|
Re: RF signal path in this circuit...
That is the AC voltage measured at the collector of the output transistor unloaded; basically swings twice the power rail voltage.
__________________
Al |
30th May 2017, 11:24 am | #25 | |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: RF signal path in this circuit...
Quote:
Essentially what you have is a oscillator followed by a variable gain amplifier. The first stage of the amp may have its input capacitance modulated due to Miller effect, which will cause the FM you reported in another thread. |
|
30th May 2017, 11:40 am | #26 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: RF signal path in this circuit...
I'm not that clued up on RF engineering but couldn't the parallel resonant circuit (loop plus capacitor) be connected between the collector and the +ve supply?
Lawrence. |
30th May 2017, 2:28 pm | #27 |
Dekatron
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
|
Re: RF signal path in this circuit...
I can't see that working . This circuit can either (more commonly ) have a tank circuit in the collector of the output transistor or an RF choke, as shown. The load is then in either case the antenna. In this case it it a resonant loop but normally ( since this is a pantry transmitter) it is a short (3m) long antenna with matching inductor/ maybe a pi network or some simple way of removing harmonics ...
__________________
Al |
30th May 2017, 2:31 pm | #28 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: RF signal path in this circuit...
Your loop antenna circuit is a tank circuit...?
Lawrence. |
30th May 2017, 3:18 pm | #29 |
Rest in Peace
Join Date: Sep 2008
Location: Solihull, West Midlands, UK.
Posts: 4,872
|
Re: RF signal path in this circuit...
It could be in the collector circuit, but you still need to ensure that the AC and DC go to the right places. If the loop is connected to the 12V supply then you would need to be careful never to short it to the case. Generally best if antennas are at DC 0V.
|
30th May 2017, 3:39 pm | #30 |
Octode
Join Date: May 2011
Location: Lancashire, UK.
Posts: 1,349
|
Re: RF signal path in this circuit...
You can get around -/+20 volts but then the curious frequency multiplication occurs (looks to be X3).
|
30th May 2017, 3:43 pm | #31 |
Guest
Posts: n/a
|
Re: RF signal path in this circuit...
You could couple the loop to the transmitter with a small feeder loop like http://www.cvarc.org/resources/Tech_...op_antenna.pdf
|
30th May 2017, 6:14 pm | #32 | |
Dekatron
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
|
Re: RF signal path in this circuit...
Quote:
Only..by using one of the online calculators there, I realise the whole design concept of the magnetic loop shown is shonky. That's the original one, let alone mine. I now have some doubts about the plausibility of claims made for its viability from the author, although he may well just have forgotten to mention a few things along the way. To be fair, he does say somewhere that best results are at the top of the broadcast band, which is quite a way from the 1MHz that I chose. But still... My magnetic loop was always designed to be built around my own idea, a frame of six vintage snooker triangles that I salvaged from the old snooker hall down the road, a place regularly used by Hurricane Higgins in the past. I call this the 'snooker triangle six.' So I resisted the design based on copper plumbing pipe and used 3mm wire, that I already had to hand, instead, wound in a single turn around the conjoined triangles. A few tweaks with some hand wound smaller inductors for tuning, and away we go... I knew there'd be differences in inductance etc but I could design for them and compensate. And so I did. In the end, the results are woeful - and that's the published design! Here's a comparison, using a Java Applet in the article posted by MM. 1) Original, Copper plumbing pipe design, (15mm dia.) Outside diameter 2.02m Efficiency -49dB (0%) Bandwidth - 3.19KHz at 1MHz - REJECT! Q=314 (actually too high, hence the poor bandwidth, and the author's attempt to damp the circuit with a 10K parallel resistor) Circulating current =2.6A with 150mW in 2) 'Snooker triangle six'- 3mm wire. Outside diameter 2m Efficiency 0% Bandwidth 38KHz - perfect! Q=25, dire. Circulating current =1.16A with 150mW in. A note in the article posted by MM advises that small antennas are inherently inefficient (no surprises there) and that for this centre frequency of 1MHz a winding length of 36 to 72 metres is ideal. So that's very far from just 2 metres! Of course we can get resonance using the high parallel caps but that doesn't undo the inefficiency... Conclusion 1) I will investigate the resource provided by MM, thank you! 2) I like the look of my 'Snooker Triangle Six' and a lot of work (for me in my current state) went into making it, so I will keep it as the basis for either a receiving loop antenna instead, or a better transmitting one. ( looks good. I'll post photos another time, when its application is settled.) That will mean a more sensible number of turns, likely with Litz wire to keep inductance sensible. 3) -conclusion and note to self - don't always trust the claims made by authors of circuits without doing the maths. 4) Remove a few spurious components (as discussed above) from the transmitter and reconfigure. The quality of modulated waveform is high and the output is nice and stable. I can decide whether to adapt it for a 3m antenna or (preference) rework the Snooker Triangle Six.
__________________
Al |
|
30th May 2017, 8:22 pm | #33 |
Nonode
Join Date: Jun 2013
Location: Surbiton, SW London, UK.
Posts: 2,801
|
Re: RF signal path in this circuit...
You would be better off using a conventional tuned loop, with a coaxial feed from the collector to a single turn un-tuned loop with a separate 15 turns on the same former tuned by a 500pF variable capacitor.
An alternative method is the leaky feeder system with the output fed to a length of un-terminated coax. |
30th May 2017, 11:13 pm | #34 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: RF signal path in this circuit...
Outside diameter 2m?
So with six snooker triangles making a hexagon, the triangles must be 1 metre on each side. Someone was either playing with very large balls, or an abnormal number of normally-sized ones. If you used that dimension to calculate your inductance, Q etc, your results could be a bit off. 2m circumference perhaps? David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
31st May 2017, 8:08 am | #35 |
Dekatron
Join Date: Apr 2011
Location: Cornwall, UK.
Posts: 13,454
|
Re: RF signal path in this circuit...
2M might be the length of the wire not the diameter of the loop, ie: "circumference"?
Lawrence. Last edited by ms660; 31st May 2017 at 8:21 am. |
31st May 2017, 8:37 am | #36 |
No Longer a Member
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
|
Re: RF signal path in this circuit...
Hello again Astral Highway.
I can offer some advice on loop antennas for AM transmitters, since I have been designing & experimenting with various types for decades. The radiation resistance of the loop antenna is proportional to the square of its cross sectional area and the number off loop turns squared. Also proportional to the fourth power of its operating frequency. This means as the size of the loop is smaller, the number of turns lower and the frequency lower the ability of any loop to radiate well is severely reduced. The military have had some good success with a single loop of modest size (1 meter square) operating around 30MHz, but for the AM radio band a small loop like this is hopeless, that is, if it is only a single turn of wire or a copper pipe. It requires many more turns, unless the loop is physically very large. I have attached a photo of a couple of compact loop antennas (both operate at 1480KHz). The the loop is wound with 60 strand Litz wire and inside PVC conduit, that makes it appear as a solid loop. The input impedance, via the matching network, results in 50 Ohms. This sort of design is very easy to build. In the photo the grey one is fed by a modified Tek sine wave generator that has had an amplitude modulator added to turn it into a variable frequency AM transmitter, the other in the orange housing is a custom built Xtal controlled AM transmitter which contains audio compressors and soft clipping to prevent over-modulation. I would recommend going down the road of a multi turn loop. |
31st May 2017, 8:56 am | #37 |
Dekatron
Join Date: Nov 2006
Location: London, UK.
Posts: 3,496
|
Re: RF signal path in this circuit...
Ahaha! That's a hilarious image , David! Checking just now, the applet does ask for 'length of wire' ; i.e., cirumference, which I knew by heart. I entered 'diameter' in my write up , just distracted !
__________________
Al Last edited by Al (astral highway); 31st May 2017 at 9:02 am. |
31st May 2017, 9:12 am | #38 |
Moderator
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,801
|
Re: RF signal path in this circuit...
Well, Higgins did have a bit of a reputation.
David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done |
31st May 2017, 9:17 am | #39 |
Hexode
Join Date: Jul 2007
Location: Edinburgh, UK.
Posts: 344
|
Re: RF signal path in this circuit...
Hi Argus25
Just out of interest which of the Tektronix TM500 modules are you using? Ross |
31st May 2017, 9:59 am | #40 |
No Longer a Member
Join Date: Oct 2016
Location: Maroochydore, Queensland, Australia.
Posts: 2,679
|
Re: RF signal path in this circuit...
Its the SG503.
|