80m Tx PA

321 views
Skip to first unread message

NZ0I

unread,
Feb 16, 2017, 3:00:21 PM2/16/17
to Receiver Development Platform
I confess: I do not understand the design of the power FET 80m power amplifier circuit design shown below. I'll explain what I think I understand about its operation, and maybe someone can clarify any misunderstanding on my part.


Q701 (IRF610) is a power switching FET, designed to operate in either of two states: on or off. In either of those two states, it should have very high efficiency - ~1.5 ohms on resistance, and Its off state it should pass only a few 10s of microamps of current. It is being switched off and on at about 3500000 times per second. Each time it turns on zero volts is applied to the drain, the field begins to build up around L701 as current starts to flow through it from the 12V supply. When Q701 turns off, the current through L701 drops precipitously, and the field built up around L701 collapses, inducing a large voltage at the drain of Q701. Some of the energy from the field collapse gets coupled through C703 and into the antenna system. The energy that doesn't get coupled into the antenna is instead absorbed by L701 and Q701, and some is radiated locally by the wires carrying the current. The on/off cycling continues at a 3.5MHz rate, generating the signal radiated through the antenna, along with heat dissipated in L701, Q701, and the wiring. Am I missing anything in that explanation?

What I don't understand is why a capacitor is being used to couple the energy into the antenna system. This seems like it might be an inefficient way to couple the energy into the antenna, because it works sort of like a voltage divider. Much of the energy in the system is held in the field created by the current drawn by Q701, so using voltage coupling doesn't seem like the best option from an efficiency standpoint. Indeed the numbers for efficiency shown here http://www.qsl.net/on7yd/atx80.htm#Function suggest that you may not approach 70% efficiency until you raise the voltage across Q701 to 30V or higher. (I'm not sure how those measurements were taken, so I can't vouch for their accuracy.)

As an experiment I created a SPICE model for the power amplifier shown above, and for a similar inductively-coupled design. The results are shown below. The transformer would probably need to be hand wound on a toroidal form, and each side would be tuned with a trimmer capacitor. The inductive approach shows the peak signal at the output (across 50 ohms) being 10x what you get with a the 100nF capacitor coupling: that would be a 20dB signal improvement. Also note the improved signal waveform at the input to the LPF (VM2).

I bring this up because, if we can make the 80m transmitter's signal just as strong with less current, then we can draw less power and greatly increase the life of the batteries. If these models are indicative then I suspect we can do significantly better than the 60% efficiency cited in the ON7YD link above.

BTW - the LPF design in these models appears to work very well, with a cutoff just above 4 MHz, and a deep notch very close to the 3rd harmonic - the notch being created by L802 and C803.





Patrick Robert Sears

unread,
Feb 16, 2017, 5:17:18 PM2/16/17
to receiver-devel...@googlegroups.com


Hi Charles,

I'll have to look at your comments and try to think through how the IRF610, L701, and C703 interact.  It took me so long to get any signal from the si5351 to the output of the IRF610 that I just put together the rest of Rik Strobbe's design without thinking much.  I had even designed my own 80m filter (using LT Spice and some stuff I read) and put it on at the end of Jerry's 80m microfox.  That's where I first started using the Si5351 (as replacement for the crystal on the microfox).  But I figured that RS knew more about it.  Some of his writing mentions being careful to not change the material on the toroids unless you know what you're doing and I didn't understand that part.  So I just used his design for that also.  So you see, I didn't investigate this part.  But it would be great to get the same power out without drawing as much current.  And besides, it really would be nice to know how it works (I did have an idea but your comments show that my understanding was simplistic).

For now, I'm still studying the changes to get to Rev B.19.  I've been having fun reading the datasheet for the TLE2426.  I only came across the concept of virtual ground for the first time this last week.  I was studying opamps in anticipation of using the OPA355.  It looks like it's only being used in the "VHF Driver/Modulator" part of the board.  It still have much studying to do there.

Which reminds me, I have a small proto-board on which I build a section of Rik Strobbe's 2m transmitter.  It's the one found here:  http://www.qsl.net/on7yd/atx2.htm  And I set it up so the filter is connected to the rest of the circuit by jumpers across a couple headers.  So we have it already built in case we want to test it and can easily be isolated from the rest of the board.  One thing I'm not sure about is how my layout or the particular components I chose might inappropriate for VHF; never having done anything above 7  MHz.  Anyway, I'll bring it to the o-meet on Sunday if you'll be there.

Cheers,

Patrick

--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/41f4b846-7ec1-4a36-aa32-86cac9cef7fa%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

NZ0I

unread,
Feb 16, 2017, 8:53:52 PM2/16/17
to Receiver Development Platform
I've taken a look at possibilities for simply going from an Si5351 clkout (.6V - 2.7V) to a the power FET's gate voltage (0V - 20V) in two steps or less. A trick I have used in the past, and it actually made its way into a piece of commercial equipment, is an opto-isolator. In an opto-isolator the only thing that flows from one side to the other is photons, so provided you can drive the LED on one side, you can control any voltage within the rating of the device - up to kV values for some parts. The downside to optoisolators is their current drain, cost, and sometimes their lifetime. Since we probably wouldn't want the Si5351 to directly drive an LED, there would probably be one voltage level shift required to properly drive the LED. That approach remains a possibility, but while looking for opto-isolators I came across this part: http://www.digikey.com/products/en?keywords=336-2386-ND

The Si8710 is essentially a capacitance-coupled replacement for opto-isolators. It requires less drive current (~5 mA) at a voltage slightly below the maximum output from the Si5351. It has better reliability than LED opto-isolators, and can switch at speeds up to 15 MHz. In short, it might be possible to go from Si5351 to driving the IRF610 in a single step using this part, making the 80m transmitter very simple.

I would still like to give the processor the capability to adjust the CW-only 3.5 MHz output power, but I'm thinking of doing that by allowing the processor to adjust the DS (drain/source) voltage applied to the power FET using a variable switching supply. This could make for a very simple, very efficient, variable-power 80m transmitter.

NZ0I

unread,
Feb 16, 2017, 10:11:13 PM2/16/17
to Receiver Development Platform
I did some more jiggering with the PA design (see below). The output RF transformer is now a 1:1 transformer, with only the primary tuned to resonance. L701 is now serving simply as an RF choke, and its value isn't critical, so long as its reactance is relatively high at the transmit frequency. Some degradation of the signal going into the LPF can be observed on the Oscope, but the output at the antenna has gone up another 20dB. The transmitter is now putting out 20V p-p, or about 4 Watts into a 50-ohm antenna... if I've done my math right. That is for a 12V power supply. We would want to lower the voltage to 9V or less if we see these actual results in hardware. I would guess the efficiency is better than 80% with this design, I'll have to see if I can take some current measurements (simulated) to determine the input power being consumed.

This makes me think that perhaps the original design works as well as it does (on an actual board) only because you are using a real L701 component that doesn't function as a perfect inductor, or perhaps there is some coupling going on between L701 and the LPF inductors. It is quite strange that it models so poorly. But maybe my model is wrong. Still, I would expect better than 60% efficiency from this Class C PA, so maybe the real-life results also aren't all that great.



On Thursday, February 16, 2017 at 3:00:21 PM UTC-5, NZ0I wrote:

NZ0I

unread,
Feb 17, 2017, 10:32:15 AM2/17/17
to Receiver Development Platform
So here is what I am proposing for the 80m transmitter design, subject to modeling and testing - I'm sure there will be changes. The oscillator, isolator, and PA should work just fine from just above DC to >10MHz. You'd just need to switch to different transformers and LPF configurations for the various bands you want to operate on. It also supports FSK with a simple port pin toggle by the uC.

There is currently no way provided to key the transmitter other than to turn on/off the oscillator clock output via I2C. Or you could turn off one of the selectable clock outputs, and switch to the disabled output to turn off the transmitter. 

The variable power supply +5V to +12V that powers the PA FET should have some current limiting applied, to avoid having components and traces fry if the power FET ever gets turned on with no signal applied (e.g., failure of the isolator). There are various ways to accomplish the current limiting.

The design below strays from the "tried and true" capacitor-coupled output technique. Personally, since I can't understand why the ON7YD design has poor efficiency, models poorly, or functions well at all, I'd rather go with a design that at least I understand. A better explanation of the ON7YD design could persuade me to change back to it.




On Thursday, February 16, 2017 at 3:00:21 PM UTC-5, NZ0I wrote:

NZ0I

unread,
Feb 18, 2017, 8:43:45 AM2/18/17
to Receiver Development Platform
Thinking about this a little more, I suppose the ON7YD 80m tx PA design isn't so different from the 2m transmitter design we are considering, that uses the BLT50. Provided you have a 50-ohm antenna (or dummy load) connected at the antenna output, it should present as a 50-ohm impedance at the output of the final transistor, or FET. This will be a much lower impedance than the RF choke through which the DC collector/drain bias is supplied. So one would expect most of the energy to be directed into the antenna. I am probably getting myself wrapped around the axle by worrying about the inductive effects of the RF choke... it should probably be ignored. I think many 1000's of 80m transmitters with this design have been built and used successfully over the years, and it is hard to argue with success.

Still, I don't understand why it doesn't model better. And probably that's due to a deficiency in my model. It also seems like the PA should be more efficient than ON7YD's numbers, but again that's just based on assumptions of power FETs being used in places like switching power supplies.

I'll play around with the model some more, and do some more googling. Perhaps it would be worth supporting both PA approaches in the first experimental board, just so they can be compared in practice.


On Thursday, February 16, 2017 at 3:00:21 PM UTC-5, NZ0I wrote:

Gerald Boyd

unread,
Feb 18, 2017, 6:31:56 PM2/18/17
to NZ0I, Receiver Development Platform
I think c703 main purpose is to function as a DC block to keep DC of the antenna connector.


Sent from my iPad
--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.

NZ0I

unread,
Feb 18, 2017, 8:59:09 PM2/18/17
to Receiver Development Platform
I agree with the DC-blocking function of C701. I should have been more coherent about my confusion... but when I'm confused it is hard to be coherent.

In addition to blocking the DC from reaching the antenna, C701 passes the RF into the LPF and antenna system. I'm fine on those points. But I am aware of two options for coupling the RF into the antenna system: through a DC-block capacitor, or through a transformer. In this case, we have a low-voltage system (12V) but potentially high current (amps) thanks to the low on-resistance of the power FET. And, at 3.5 MHz, it shouldn't be too challenging to build a small transformer using a ferrite toroid. So it seems that both design options are possibilities here.

If we consider L701 to be functioning purely as a choke, passing DC but blocking RF, then L701 doesn't contribute any kind of inductive "kick" to boost the voltage in the system. So what we have on the left side of C701 is a 12V square wave: the voltage goes up and down between 0 and +12V at a 3.5MHz rate. Let's assume that the square wave crosses C701 unattenuated, and we therefore have a signal of the same amplitude and frequency on the right side of C701. That square wave passes through the LPF and becomes a nice sinusoid, and let's assume that it becomes a sine wave with the same amplitude as the square wave had: 6V (i.e., it is 12V peak-to-peak). Let's assume a perfectly resistive antenna impedance of 50 ohms. The power being delivered to the antenna is then  E^2/R, where E is the voltage RMS amplitude of the signal. That gives us  (6V * .707)^2 / 50 = 0.36W. So the maximum power that can be delivered to the antenna is 360 mW for a 12V system. I'm not an RF engineer, so I am probably missing something, and my argument could be totally wrong. But that is the best explanation I can provide about why the DC-blocking capacitor approach doesn't make sense to me. Also, when I model the final amplifier with the DC-blocking capacitor, I get output power that is disappointingly low.

360 mW, or even 200 mW, would probably work just fine for most practice hunts. So I'm not concerned so much about the power being low. I'm especially unconcerned about the low power prediction since I acknowledge that my analysis is probably wrong, and these transmitters are probably capable of doing much better than my numbers suggest. But still it seems to me that it should be possible to do better using a transformer, which would couple the energy from the relatively-high current into the antenna system.

To explore that possibility, I created a second SPICE model, and this time I substituted a transformer for the DC-blocking capacitor. (Actually, I left the capacitor in, but added a transformer as well. The primary of the transformer connects between the FET's drain and L701. A 10uF capacitor was added at the top of the primary to decouple the DC. The secondary of the transformer went between ground and the input to the DC-blocking capacitor. Otherwise, no changes were made.

The SPICE model with the transformer shows a 40V peak-to-peak sine wave being generated at the antenna connection point. Using the same formula for calculating power, that works out to 4 Watts being delivered to the antenna from a 12V system. Using the simulation to measure the DC current being delivered to the FET, I found that the amplifier was consuming 5 Watts of power. That works out to 80% efficiency. That is in line with how I would expect the FET to operate in, say, a switching power supply. So I find it credible.

This author: http://www.qsl.net/on7yd/atx80.htm#Function claims to get 2.9 Watts of power, at 61% efficiency, from his DC-blocking capacitor design using a 12V power source. I have no reason to doubt his results, since I acknowledge the limitations of my knowledge to explain the functioning of the DC-blocking approach. But ignoring the power output, the efficiency seems less than one might expect from the FET being used. 

I suspect that L701 is not functioning purely as a choke, and that indeed it is serving to boost the voltage being delivered to the antenna. But I don't know why SPICE isn't modeling that behavior. Regardless, there might indeed be an efficiency advantage to the transformer-based design.

So I'm still confused, and seeking to better understand and model the DC-blocking capacitor PA design. But I am leaning toward including the option to build either a DC-blocking capacitor, or a transformer-based PA, on the experimental dual-band transmitter PCB design.


On Thursday, February 16, 2017 at 3:00:21 PM UTC-5, NZ0I wrote:

NZ0I

unread,
Feb 18, 2017, 10:37:18 PM2/18/17
to Receiver Development Platform
As Gilda Radner used to say, "never mind". I found the problem in my SPICE model. I had failed to set the signal driving the gate of the FET sufficiently high. Once that problem was fixed, the DC-blocking capacitor design began to work as advertised. It does indeed deliver power levels on the same scale as the transformer approach. The voltage at the drain of the FET is indeed higher than +12V, it is +43V p-p, and is affected by the inductance and resistance of L701. So there is a voltage step-up effect being created by the FET driving a changing current through L701. Physics triumphs again.

The modeled efficiency of the PA is about the same as for the transformer-coupled approach. The efficiency depends a lot on the quality of L701. So use a high-quality 5uH choke to improve efficiency - one with a low DC resistance and high current rating.

Based on this, I think I will abandon the transformer approach altogether. There is no real advantage to it. I'll also stop bothering you guys with my rants and confusing arguments. Thank you for your patience.



On Thursday, February 16, 2017 at 3:00:21 PM UTC-5, NZ0I wrote:

Gerald Boyd

unread,
Feb 19, 2017, 4:44:54 PM2/19/17
to NZ0I, Receiver Development Platform
Glad to see the model is now working for the original FET non transformer version. 
It's always good to look into things just in case there is a better way.
Jerry

Sent from my iPad
--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.

Patrick R. Sears

unread,
Feb 20, 2017, 8:21:40 AM2/20/17
to receiver-devel...@googlegroups.com


Hi all,

Just spent the entire weekend building and debugging, except for Sunday afternoon at the o-meet.  Unforeseen problems with the course timing system meant that I never got to the 2m work.  But I think I solved all the timing system problems.

About the final stages on the transmitter.  I don't understand how all the components are chosen by ON7YD but before abandoning the transformer, take a look at his ARDF 2m transmitter page:  [ http://www.qsl.net/on7yd/atx2.htm ].  Even the final stages looks quite a bit different from his 80m transmitter and he does include a transformer.  Those final stages are the ones I mentioned earlier I already built.  I brought it to the o-meet, but had left it in the car when I saw you Charles, and it slipped my mind.  I may even have some files where I took pictures from the oscilloscope when I was testing it.  When I get home, I'll see if I can find those and a schematic of exactly which part of the Rik Strobbe's circuit I built; if it would be useful.

Cheers,

Patrick

Charles Scharlau

unread,
Feb 20, 2017, 9:50:16 AM2/20/17
to Patrick R. Sears, Receiver Development Platform
Hi Patrick,

I'd be interested in learning more about the system you are working on, should you decide to share. 

Regarding the ARDF 2m transmitter, the transformer in the design shown here  http://www.qsl.net/on7yd/atx2.htm is functioning as a common mode choke, keeping RF restricted to the LPF, and preventing it from reaching the modulating circuitry. It passes the audio-modulated DC through the windings to the final transistor. So, in effect, it is not so different from the design we are considering in the latest dual-band transmitter design, the main differences being the use of a common mode choke (instead of a straight choke inductor), and he is modulating the final amplifier stage, whereas we are applying modulation in the driver stage, where it should be more efficient.

There are some advantages to modulating the final, instead of the driver. It is simple and straightforward. It is also inherently inefficient, having a theoretical maximum efficiency of 50%. But, as I've mentioned before, it can be argued that throwing away a Watt or two is not all that big a deal. I also see that no one else seems too concerned about a clean sine wave modulated signal, though the design shown in the link you provided does appear to be applying a simple RC filter to the modulation, probably resulting in a more sinusoidal modulation. The 2nd order active filter in my initial design is probably overkill, and a simpler (by 1 resistor and 1 capacitor) 1st order filter would be plenty good enough.

With ARDF, the objective of most designs is to make them simple and functional. Period. And I agree with that philosophy if the goal is to post schematics on the web that others can emulate with a high likelihood of success. But my vision for the ARDF equipment designed using the Receiver Development Platform is a bit different. I'm shooting for something that is something more than the bare-bones fixed-frequency designs like ON7YD and others have posted. I'm shooting for something more flexible, modern, and usable. At a minimum the frequencies should be programmable, and the power output adjustable. The simplest way that I'm aware of to accomplish that is to use a processor to program a frequency synthesizer. But supporting frequency entry, and power setting, means a user interface of some sort. And that necessarily increases cost and complexity, and leads to "feature creep" when one discovers all the additional things that can be done with just a tweak to the software.

Back when Nadia and I were putting on ARDF practices and events, I became convinced that the whole sport is just too burdensome for a small group of organizers (e.g., 2) to sustain year after year. There is a lot of equipment, duplicated across 6 transmitters and two bands, and more recently even more transmitters uniquely designed for Fox-O and Sprint events. The equipment is too fragile, too complex for the user, and there is just too much of it. The result is that no one can consistently put on a clean event - there is almost always one transmitter (or more) that fails to function to standards, and/or there are long delays while the organizers go out and swap/adjust to get things working. Competitors have to carry spare receivers, or purchase/borrow new ones at the last minute when theirs fails. This has been true at the World Championships of the sport, where those problems crop up every two years like clockwork. If the sport is to grow, I think that has to change. 

So, in addition to making the new receivers and transmitters more flexible, modern and usable, they should also be rugged and very simple for the organizers and competitors. Everything should be dual band (whenever possible) to cut the transmitter and receiver count in half. Anyone who can open an umbrella or tune an AM radio should be able to deploy a transmitter, or operate a receiver. Making the equipment simple and foolproof is the goal, and I think today's technology can accomplish that at a reasonable cost.

But "simple for the user" is not the same as simple for the builder. In fact, it is the opposite. Making receivers that are pretuned to the right frequencies and that work reliably; transmitters that turn themselves on/off and which can be safely collected by untrained volunteers; integrating the functionality into single devices that can perform all the needed functions, and be programmed simply, reliably and with little risk of error, makes all the equipment more complex internally. It is like the modern cell phone: simple enough to use its basic functions and almost as simple to use advanced voice recognition and killer apps, with rarely a failure. But look inside at the complexity involved! However, no one needs to understand the inner workings of their cell phone, nor do they need to build the phone themselves. As with cell phones, ARDF equipment ultimately needs to be available to everyone in built-and-tested modules, or commercial devices. IMO.

I've strayed well off topic. Getting off my soapbox now. But if I sometimes seem to be ignoring the tried-and-true in favor of new-and-unproved, it is because that is in fact what I am doing. I think that's the only way forward.


To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.
--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/f4329eab-ac84-7ef4-e940-b3f86c6f3bec%40gmail.com.

NZ0I

unread,
Feb 20, 2017, 10:46:07 AM2/20/17
to Receiver Development Platform
My attention was brought to this unusual design: http://213.114.137.49/audio/push-pull-reg_0.htm

Using two voltage regulators as RF power amplifiers is certainly an outside-the-box approach, and an example of what you can get away with at low frequencies. I'm not proposing that we use this design, however, but thought you might find it interesting.


On Thursday, February 16, 2017 at 3:00:21 PM UTC-5, NZ0I wrote:

Patrick R Sears

unread,
Feb 25, 2017, 10:24:07 AM2/25/17
to receiver-devel...@googlegroups.com


Hi Charles,


Kelly and I are in Augusta GA visiting Isaac for the weekend.  There won't be much more work done on this end till next week.

About sharing the course timing system:  Definitely.  I'll get stuff out soon.

Regarding the ARDF 2m or dual-band transmitter or other ARDF equipment:

- Thanks for the info on ON7YD's design.  I looked at it some after reading your description.  I'll have to study it a bit more.

- About:  "... simple and functional. ... philosophy if the goal ... that others can emulate with a high likelihood of success."

  I agree 100%.  That's what hobbyists like me need.

- About:  "...the frequencies should be programmable, and the power output adjustable"

  I had the same concern.  That was my main reason I went to the Si5351 and adding the variable voltage regulator in the 80m transmitter.

- About:  "... my vision for the ARDF equipment designed using the Receiver Development Platform ...  transmitters more flexible, modern and usable, they should also be rugged and very simple for the organizers and competitors ..."

  I agree 100%.  It's definitely not something I could accomplish but I'm very excited for it.

- About:  "... "simple for the user" is not the same as simple for the builder. In fact, it is the opposite. ..."

  I agree 100%.  I have two personalities:  The hobbyist that likes to build stuff and needs the simple stuff and the ARDF enthusiast that wants to make ARDF events happen.  The enthusiast personality only really cares for this point your making.  I want it rugged and I want it simple to use on the outside.  That's it.

- About open source, which you did not mention here.  It's great for people like me.  But I'm not sure it makes sense for the rugged simple-to-use system.  Don't get me wrong.  The hobbyist in me would love everything to be open source.  It allows me to learn so much and gives me a community to which I can contribute and feel good about what I'm doing.  But the ARDF enthusiast in me just wants to buy it and use it.

I wanted to hit a bunch of the points you made specifically because I felt so strongly aligned with your comments.  I'll stand on the soap box with you any day to push these points : )  And I'm super excited that Jerry and you are taking on this project.

Cheers,
Patrick
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/CAK-GiwaHCAawDjxkHRXzJCevdkfzcRB0%2B9UtLC-jkqCYHMFLKQ%40mail.gmail.com.

Charles Scharlau

unread,
Feb 25, 2017, 1:38:31 PM2/25/17
to Patrick R Sears, Receiver Development Platform
Just a few comments regarding open source software:

I'm no expert on open source, and have not contributed much to it personally. But I have from time to time taken advantage of GitHub projects that others have generously shared. From what I know about open source, there is no downside from the perspective of the user who downloads and uses it. The only downside, for those who will be utilizing the shared code in derivative products, is understanding the impact of the licensing agreement. The legalities can get tricky for companies wanting to incorporate open source into proprietary for-profit products. But experimenters, makers, hackers, and pretty much everyone else, open source just means you can access, analyze, control, and keep the software that goes into your projects. Oh, and you can also contribute bug fixes and improvements if you so desire, but there is no obligation. What's not to like?

For most projects incorporating one or more processors, the software is as essential as the hardware for making a working product. You really want to control both hardware and software. Even if you never make any changes to the code, it is good to have full access to the code that makes your project function. Just for peace of mind, if nothing else. 

What I hope for from the dual-band ARDF receiver/transmitter project is the following:

1. Open hardware design (KiCad), that everyone can analyze, modify, or simply re-use to create PCBs for their own use, or use by others.
2. Open PCB sourcing (OSH Park), so everyone can order bare boards to populate by hand, or to have assembled by automation.
3. Open software design (GitHub), that everyone can analyze, modify, or simply use to generate binary files to program their hardware.
4. Technical details, documented on the web, that everyone can access to quickly come up to speed and efficiently accomplish their goals.
5. Designs that provide competition-grade performance, ruggedness and durability, low product lifetime cost, and unparalleled simplicity of use.
 
If all those goals are accomplished, then I think others will want to manufacture them in minute quantities - onesies, twosies. The designs will almost certainly be too complex for the average tinkerer to successfully build, but should be within the abilities of skilled kit builders and experienced experimenters. If the transmitters and receivers are good enough, and hold some appeal for the general ham community, then small production runs (dozens?) of factory-assembled modules might make economic sense. I think that a crowdfunded project would help answer the question of whether the designs are worthy of building by automation. Once a prototype product has demonstrated what the transmitters and receivers can do, and at what cost, the next steps will be to examine the cost of production, and gauge the interest from the ARDF/ham community.

But automated or not, anyone building these designs - or improved designs based on these - will have all the information they need to build, improve, and control the hardware and software that goes into their project. That's my goal anyway.



--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development-platform+unsub...@googlegroups.com.

Gerald Boyd

unread,
Feb 26, 2017, 7:41:50 PM2/26/17
to Charles Scharlau, Patrick R Sears, Receiver Development Platform

I think this is a valid goal.  Based on feedback from the talk this is inline. There was one person that wants to follow our project as he may want to modify the design into a QRP rig.

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.

--
You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.


For more options, visit https://groups.google.com/d/optout.

 

--

You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.

 

--

You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.

 

--

You received this message because you are subscribed to the Google Groups "Receiver Development Platform" group.

To unsubscribe from this group and stop receiving emails from it, send an email to receiver-development...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/receiver-development-platform/CAK-Giwa4-vG3VW7w7yEbrFbNbAnLtxsDqsdgjR43Ced%2BphQP1Q%40mail.gmail.com.

Patrick Robert Sears

unread,
Mar 4, 2017, 11:06:48 AM3/4/17
to Charles Scharlau, receiver-devel...@googlegroups.com


Hi Charles,

No problem sharing with RDP.  At first, I replied separately because I wasn't sure you wanted it sent there since it isn't receiver or transmitter project.  Anyway, I CC'ed the RDP in this email.

Correct.  I'm using the Adafruit #3013 dev-board with the DS3231 on it.  I will be happy to share all I find with them.

For synchronization, I'm just setting them using the programming cable and the arduino IDE's serial monitor.  I first sync the computer to www.time.gov.  I then plug in the programming cable, bring up the serial monitor, enter in the value of a time that will come up in a few seconds, wait a bit, and then send a command via the serial monitor when I see the clock reach that time on the computer.  I'm hoping that I won't have to resync them before the summer.  For the prototypes, that's good enough.  For timing kids, even if they drift by 5 to 10 s over the course of the season, I'm ok with that for this season.  In the future, I'll definitely want to get a better system.

Note about communication options:  The Adafruit dev board uses I2C.  There is also a Sparkfun dev board (https://www.sparkfun.com/products/10160) product-ID BOB-10160 which uses the DS3234.  I'm not certain, but the way I understand it, the DS3234 is basically the SPI version of the DS3231.

This RTC stuff has been a real discovery for me since last December.  It's very cool.  I definitely want to go with the system you're describing with the 80m transmitters I have already built.  I would like to design a small board with an RTC on it to plug into my spare pins header.  Then I'll shift the timing control code out of the ATmega and onto the RTC.

About the IDE I'm using.  The fox control boxes use an Adafruit Pro Trinket 3V 12MHz (ID 2010).  It still uses an ATmega328 but it will not work with the regular Arduino IDE.  I have to download a custom Adafruit version of the IDE to make it work.  Also, it's possible to program it via USB but it's not possible to get the serial monitor to work that way.  But it is possible to program it via an FTDI-to-usb cable and that allows the use of the serial monitor.

About the transfer of info to the computer.  Yup, I just pull the SD card from the master box and just copy the files onto my computer.  From there, I can write some scripts to process the data any way I want, or I can open the files with a text editor or spreadsheet program.  It's just a regular file system and text files on the SD card.  Super simple and easy for anyone to customize for their own needs.

About explaining the system and a diagram.  Yes, definitely.  Some kind of a block diagram and maybe a flow chart.  It's still super early stages and I don't even have the suppliers and parts numbers in the BOM yet.  There is still so much work to do there (and on the 80m transmitter page where I still only have pre-version 1 info while I'm on version 5).

As an aside, I just spent the morning doing voltage and current measurements on the 80m transmitters.  I think I'm going to drop the number of batteries from 16 AA to 10 AA.  I hope to get to the antenna work very soon, but with the antennas I used last season, I'm pretty confident that using a 1 W output (measured into a 50 ohm dummy) will work fine.  If so, I just don't need 16 AA.  I didn't really understand how linear regulators work.  But from the reading I've done this last month and the measurements I've done this last week, it looks like the transmitters draw the same current regardless of the voltage input to the linear regulator (about 50 mA when not transmitting and 290 mA while transmitting continuously); and all that's not used just goes to heat.  I'm using an LM317 so that I can adjust the voltage being used for generating the RF output (different regulators are used for the ATmega and the Si5351).  Anyway, this has been a real learning experience.

OK, got to go.

Cheers,

Patrick





On 03/04/2017 09:24 AM, Charles Scharlau wrote:
Cool! It looks like you've been very busy. That looks like a very cool timing system! I need to read through it more carefully, but I'm guessing that the results can be transferred from the control box to a computer using an SD card? Any other options? A diagram might be effective for describing how the whole system works.

If I am understanding correctly, you are using this part from AdaFruit for the RTC: https://www.adafruit.com/products/3013

If so, then you are using the same DS3231 chip that I've incorporated into the Control Head board. Anything you learn about the performance of that device could be very helpful for me, if you can share it. The plan is for the transmitters (and receivers and control head too) to rely on the accuracy of the RTC to be responsible for all the critical timing. The internal clock of the ATMEGA328, which is not terribly precise, will run the processor and coarse timing functions like ADC reading, tone generation, periodic internal tasks/interrupts, and the like. But the RTC will be responsible for generating precise 1-second and 1-minute interval interrupts and alarms for regulating transmit timing, turn-on time, time-of-day display, and any other precise timing functions.

Although the specs of the DS3231 look promising for those applications, there are still question marks about whether it will be necessary to apply calibration to the RTC, if temperature differences will cause larger RTC drift, and whether there are less expensive RTC devices that might work just as well.

As you learn about clock drift between timing boxes and the master box, and temperature dependence, I would be very interested to hear about your observations. Also, it wasn't clear to me how you are accomplishing clock synchronization between all the devices... that's something else that could have application to the receivers, transmitters and control head.

Is it OK to share information about your system to those on the Receiver Development Platform email list? If so, I'd like to start a thread there.

Thank you for sharing!

73,
Charles

On Sat, Mar 4, 2017 at 5:22 AM, Patrick Robert Sears <patrick...@gmail.com> wrote:


Hi Charles,

I finally started working on Kelly and my website again.  And I got some information about the course timing system on there.  You can find it at http://islandcreativetime.com/y2017/projects/course_timing_system/main.html

Cheers,

Patrick




On 02/20/2017 09:50 AM, Charles Scharlau wrote:

Charles Scharlau

unread,
Mar 4, 2017, 2:47:31 PM3/4/17
to Patrick Robert Sears, Receiver Development Platform
Thanks Patrick, that all sounds great. I look forward to seeing the timing system in April.

Regarding linear regulators, you can just think of them as self-adjusting resistors: they automatically adjust their resistance in order to drop the voltage at their input to their specified regulated output voltage (e.g., 3.3V). So the higher voltage you put into them, the greater voltage that they must drop. All that voltage drop results in heat that is dissipated by the regulator per Ohm's law: I2xR, where R = the resistance across the regulator. Linear regulators are inherently inefficient, and the greater the voltage you apply to their input, the more inefficient they are. If you put 12V into a linear regulator that provides 5V at its output, and passes 100 mA of current, the regulator must drop 7V at 0.1A = 0.7 Watts of power. The circuit receiving regulated 5V is only dissipating 0.5 Watts of power. So the linear regulator is only operating with 42% efficiency.

Switching regulators work on a different principle. They generate voltages at their outputs using power FETs that operate in either of two very efficient states: On or Off. An ideal FET, performing as an ideal on/off switch would be 100% efficient in either of those states. They switch current through a high-Q inductor, which also operates at high efficiency. The switched current is passed through the inductor with varying frequency and/or duty cycle to achieve either a higher or lower voltage from voltage pulses resulting from the inductor's magnetic field that builds up when the FET turns on, and collapses when the FET turns off. Regulated output power can be generated at greater than 90% efficiency with some switching regulator designs. Efficiency tends to increase with greater load currents, so switching supplies are ideal for providing high currents, but become less efficient under light loads. The main drawbacks to switching supplies: they tend to generate a lot of electrical noise (both radiated by the inductor, and conducted on the power supply lines), and switching supplies are more complex and expensive than linear regulators.

Transmitters are good applications for switching supplies because transmitters can tolerate more noise (versus a receiver for instance) and draw appreciable current while transmitting. We should be able to improve the efficiency of the transmitters by turning off the switching supply (and powering down the transmit stages) when the transmitter is off the air. 
Reply all
Reply to author
Forward
0 new messages