PilotAware
British Forum => Technical Support => Topic started by: rogellis on December 25, 2017, 12:48:29 pm
-
.
The second PAW unit I built is also producing some (pulsating) static on the speaker. At a guess, it is about 50 hz.
The static is not from RF, because there is no noise if amp/speaker runs from a different power source, so it must be from the power supply to the amp/speaker and PAW.
But the noise cannot be the fault of the dc converter, because it is coincident with the PAW unit initialising itself (it is silent prior to initialisation). So this appears to be an interference back down the power line from the PAW unit itself.
(I am not in electronics, but I presume a unit can send pulses back down its own power line. The PAW and the amp/speaker both draw their power from a common source.)
R
.
-
.
The second PAW unit I built is also producing some (pulsating) static on the speaker. At a guess, it is about 50 hz.
The static is not from RF, because there is no noise if amp/speaker runs from a different power source, so it must be from the power supply to the amp/speaker and PAW.
But the noise cannot be the fault of the dc converter, because it is coincident with the PAW unit initialising itself (it is silent prior to initialisation). So this appears to be an interference back down the power line from the PAW unit itself.
(I am not in electronics, but I presume a unit can send pulses back down its own power line. The PAW and the amp/speaker both draw their power from a common source.)
R
.
The PAW starts off drawing only a small amount of power, then as it boots up the current draw increases, so this could be conduced emissions from the power supply as the load increases. Many supplies are "quiet" without load then noisy when loaded, alas many is just noisy all the time too. Can you plug something else into the supply and see is the noise returns?
-
Can you plug something else into the supply and see is the noise returns?
I rigged up the dc converter to a mobile phone and iPad, both on charge, plus the amp/speaker - and it was perfectly quiet.
I think the interference is definitely from the PAW, as the noise is perfectly in tune with the flashing LEDs on startup. It then settles down to some static, plus a 5 hz square-wave pulsation. There was a nifty noise generator online, and I was able to replicate the pulsating sound.
http://onlinetonegenerator.com
As a temporary fix, I was able to mute the speaker a bit, and increase the PAW volume to 10 and that muted most of the static. But it is still annoying.
The other PAW gives the same sounds, but much much softer. If I get time I might swap the amps over, but they have different connectors, so it is not straightforward.
R
-
5hz? Your not a whale are you :)
-
Rog,
It should be possible to block interference electronically by filtering. Jeremy Curtis is the ‘expert’ here as that’s his ‘business’ field, but it would probably be worth trying a ground loop isolator in the lead between the PAW and the speaker amp, or use a 3-pole TRS plug which will short the unused ‘Video’ channel at the RPi audio output to ground to see if that helps.
Regards
Peter
-
A lot also depends on the Amp design etc. can you post a link to the amp?
Charging a battery is generally a pretty steady state power delivery, the likes of a PAW has a very 'spiky' power consumption (I think I posted a graph on the forum a while ago showing this), this can also cause problems with supplies to especially if designed as chargers rather than power supplies. When you rigged up the DC converter did you measure how much current was being drawn whilst powering the phones? Should be at least 1A for an effective test.
-
A lot also depends on the Amp design etc. can you post a link to the amp?
When you rigged up the DC converter did you measure how much current was being drawn whilst powering the phones?
Sorry, Im a pilot, not an electrician. The recommnded Anker dc converter also made the same noise with this unit.
This is the amp I used. It looks nice inside, with a mutitude of ic chips, but it is not designed for this application.
Components like this could do with being integrated into the PAW unit. I would happily pay more, not to have this hassle.
https://www.amazon.co.uk/N4U-Online®-Microsoft-Capsule-Rechargable-Black/dp/B00WHPF20E
R
R
-
A lot also depends on the Amp design etc. can you post a link to the amp?
When you rigged up the DC converter did you measure how much current was being drawn whilst powering the phones?
Sorry, Im a pilot, not an electrician. The recommnded Anker dc converter also made the same noise with this unit.
This is the amp I used. It looks nice inside, with a mutitude of ic chips, but it is not designed for this application.
Components like this could do with being integrated into the PAW unit. I would happily pay more, not to have this hassle.
https://www.amazon.co.uk/N4U-Online®-Microsoft-Capsule-Rechargable-Black/dp/B00WHPF20E
So you would be looking for a speaker for use in the cockpit rather than wiring into an intercom or headset?
I've pondered making an "aviation grade" audio mixer that can take the likes of a PAW signal (iPad, or anything else for that matter) with an output for an intercom, aux into a headset or potentially a bluetooth link too. The design I hashed out would also have priority on the inputs, so lower priority inputs would duck for higher priority etc. Naturally all filtered and with some power options to boot.
I may have another look at at some point when I get some free time.
-
I have "pondered" along similar lines on and off for a few years now. A unity gain summing amplifier readily gives the ability to plug any number of inputs into such a box and allows them to be fed to the intercom. The difficult bit is adding the ability to assign priorities amongst them, including giving the comms radio the highest priority, presumably? Seems a good task for a microcontroller I suspect.
Bill
-
So you would be looking for a speaker for use in the cockpit rather than wiring into an intercom or headset?
This is for use in gliders, and we do not use headsets. If PAW wishes to enter the glider market, it is a problem that needs addressing.
(I suppose the aircraft headset has a different 5v source, so does not have this PAW interference.)
I presume the main problem is the pulsed transmission from the PAW aerial, which I understand goes out at 5hz - the same as the interference clicks. In the old days of motor mechanics, we used to put a suppressor (a capacitor) across the ignition coil, to prevent similar feedback. Would that help? A capacitor across the 5v output from the dc converter? If so, what type and size of capacitor?
I will also try taking the power from the Raspberry output, to see if that smooths out the 5v current a bit. But I have to connect it to the car, as that is my only 12v battery source.
Cheers,
R
-
The PAW transmission goes out every 1.8 seconds or so.
-
50hz sounds suspiciously like mains induced hum ?
Thx
Lee
-
5 not 50hz
-
Hi Ian
I read this ....
.
The second PAW unit I built is also producing some (pulsating) static on the speaker. At a guess, it is about 50 hz.
-
Sorry Lee, was not reading that far back.
50hz to me says 'earth loop'
-
I think the interference is definitely from the PAW, as the noise is perfectly in tune with the flashing LEDs on startup. It then settles down to some static, plus a 5 hz square-wave pulsation.
I think that Roger is saying there is also intermittent 'click / buzz' at 5 x per sec (5Hz) as well as the 50hz noise??
Deker
-
50hz to me says 'earth loop'
Sorry, the mention of 50 hz was just a guess.
I found this noise generator online and was able to precisely duplicate the buzz, and it is actually a square waveform 5 hz.
http://onlinetonegenerator.com/
This might suggest the PAW transmitter, which I understand works at about 5 hz. There is also a background static hiss.
Someone said a PAW pulse every 1.8 seconds, but I was told that PAW transmits 6 times a second.
As a reminder, when the amp is working on battery, there is no noise at all.
And when I switch to working from the Raspberry USB 5v output, the noise is halved.
And the other PAW unit has the same noise, but much quieter.
I tried using a different dc converter, but the noise remained.
Will a suppressor (capacitor) on th 5v input work? - Like we used to place on car ignition-coil systems?
It was a simple solution to a buzzing radio.
If so, what size capacitor should I look for?
Thanks,
R
-
A clicking sound rather than a buzz?
-
As a reminder, when the amp is working on battery, there is no noise at all.
is that with a running PAW connected?
-
A clicking sound rather than a buzz?
A 5 hz click, just like that online tone generator on square wave mode.
Also a background hiss, but this is not as prominent and annoying as the click.
That is ONLY with the PAW connected and running.
With either non amp/speaker connection (battery supply) or no PAW running, there is perfect silence.
On startup, the speaker gives various bleeps and clicks, perfectly in tune with the PAW LED flashes.
When up and running, it gives the 5 hz clicks and background buzz.
R
-
If you leave the GPS and WiFi dongles off do you still have the click? Could easily be conducted emission from one of the other devices plugged into the Pi
-
If you leave the GPS and WiFi dongles off do you still have the click?
Could easily be conducted emission from one of the other devices plugged into the Pi
Ok, got around to some more testing at last.
The primary noise is coming from the ADSB aerial, or rather from the dongle. If I disconnect the ADSB aerial, the noise remains, but if I disconnect the ADSB dongle the vast majority of the noise instantly goes away.
So I am presuming that different ADSB would solve the problem. The other PAW unit makes a similar noise, but much quieter, as if there is some variability in these units.
As an aside, the ADSB connector really need a locking mechanism for aircraft usage, as the aerial is forever falling out of the dongle.
(On its own, the PAW unit has a soft 0.5 hz heartbeat, which is oddly comforting....)
Thanks,
R
-
How does one buy a new ADSB aerial?
I cannot see one on PAW hardware, Amazon, or Ebay.
Is it the Noor Nano-2?
https://www.amazon.co.uk/dp/B018PUYPCA
Cheers,
R
-
I would have said it was the ADSB dongle that is generating the noise, the aerial is just broadcasting it. Best to replace both dongle and aerial, which will come with the dongle anyway.
There seems to be a lot of variability in these units, evidenced by some running hotter than others.
-
The ADS-B dongles are these:
https://www.ebay.co.uk/itm/Mini-USB-DVB-T-RTL-SDR-Realtek-RTL2832U-R820T-Stick-Receiver-Dongle-MCX-Input-PK-/222637370515 (https://www.ebay.co.uk/itm/Mini-USB-DVB-T-RTL-SDR-Realtek-RTL2832U-R820T-Stick-Receiver-Dongle-MCX-Input-PK-/222637370515)
They come with an aerial. If you just want the aerial:
https://www.ebay.co.uk/itm/High-Gain-Signal-5dBi-Antenna-Aerial-Digital-Freeview-For-DVB-T-TV-HDTV-Black/112435558439 (https://www.ebay.co.uk/itm/High-Gain-Signal-5dBi-Antenna-Aerial-Digital-Freeview-For-DVB-T-TV-HDTV-Black/112435558439)
The antenna plug should be a fairly snug fit. They don't take too kindly to frequent plugging and unplugging though. If it's was loose when it arrived from PilotAwareHardware, get Dave to swap it.
-
I would have said it was the ADSB dongle that is generating the noise, the aerial is just broadcasting it. Best to replace both dongle and aerial, which will come with the dongle anyway.
There seems to be a lot of variability in these units, evidenced by some running hotter than others.
I thought the ADS-B aerial was receive only?
Why does it seem to be working much harder than the PAW aerial, which is transmitting?
A portable FM radio does not work very hard, and does not use so much power....
Cheers,
R
-
These dongles are Software Defined Radios, and are pretty power hungry and inefficient, however they get the job done and are as cheap as anything as they are churned out in huge numbers. Unlike an "normal" radio they do a lot of signal processing to extract the signal of interest and the design of these things is based on cost rather than efficiency or adherence to emissions standards (regardless of what the label says). You could buy a box of 500 from a Chinese supplier and get 10 different versions in the pack, cost is king for these things.
The "ideal" way to receive ADSB (and all such similar signals) is to have an RF receiver tuned for those frequencies and then stick a small(ish) FPGA behind it to extract the data from the signals received. But you won't get one of them for $5 delivered from china. It's a reasonable amount of work to make one of them, but the performance would be orders of magnitude better than an SDR dongle. If you did this, you would also take quite a load off the Pi too as it spends lots of time running software to decode the ADSB and similar signals from the stream of data from the SDR dongle, which opens all sorts of possibilities for Lee's PAW code to be ported to a more elegant bespoke bit of hardware - but the end result would be more expensive and there would be no BYO option available.
-
I thought the ADS-B aerial was receive only?
Cheers,
R
Rog,
A great explanation from Jeremy C as to the limitations of commercial SDR dongles.
WRT the 'normally receive' ADSB antenna 'broadcasting' the noise, its simply that if the SDR dongle is generating 'Radio Frequency Interference' (RFI) internally, the attached (normally receive) antenna, or its cable (or any other piece of connected wire come to that) can then act as an antenna and radiate that interference locally, where it is probably being picked up by the additional wiring and amplified through your amplifier / speaker system.
In the case of the PAW 'Bridge', which is a much more expensive piece of kit, the transceiver is screened inside a grounded metal case with the input and output filtered to prevent the transmission of any off frequency signals. To do this with standard SDR Dongles would significantly increase production cost, weight, size and complexity and has never been found necessary for 'normal' PAW use.
Hopefully swapping your dongle and taking care with component selection and placement will solve your interference problems, however as space seems to be less of an issue than the interference, you could try one of the 'better quality' SDR dongles available, such as the NooElec Temperature Controlled (TCXO) dongles which we use in the PAW OGN-R Rebroadcast stations (primarily because they are less subject to frequency drift - which is not an issue for normal PAW 1090 MHz use, but critical when looking for very low power glider signals on 868 MHz). See...
https://www.amazon.co.uk/NooElec-NESDR-SMArt-Enclosure-R820T2-Based/dp/B01HA642SW/ref=sr_1_2?ie=UTF8&qid=1515074676&sr=8-2&keywords=nooelec
The NooElec SDR can be fitted as a direct swap replacement for the normal PAW SDR and has a metal case and standard SMA antenna connection, so perhaps worth a try.
Best Regards
Peter
-
Thank you both for the answers x it is beginning to make sense.
How about this even smaller NooElectric dongle, would this work too?
There is a Nano-2
https://www.amazon.co.uk/NooElec-Compatible-Foreflight-FlightAware-Applications/dp/B01K5K3858
And a Nano-3
https://www.amazon.co.uk/NooElec-NESDR-Nano-Bundle-Stratux/dp/B076GWF6FF
It is tuned to the required frequencies, and appears to have a screw-in fixing. The cable dropping out of the other version is a continual problem.
Would this NooElectric have a lower energy consumption than the standard one? We are working on batteries, so any reduction in power consumption would be appreciated.
Thanks,
R
(PS - just noticed the optional heat sink, so perhaps it is not so efficient....)
.
-
There won't be much, if any, power usage difference between them - it's a "feature" of using this type of dongle. Double whammy really, using an SDR means you need take quite a chunk of the Pi CPU horsepower so both end up needing more current and become voltage sensitive too, the RF noise reduces sensitivity of the system as a whole. It's a cost/benefit thing and give a bit of insight into where there is such a price gap between a PAW and other solutions available.
-
Rog,
I just discovered the N3 today myself, so can't comment until I get a chance to test one. IIRC the Nano 2 doesn't have a particularly great reputation in the SDR world, though I can't remember the details at present. I do know the one I suggested works, hence why I suggested it. It also uses SMA for the antenna, which would get round your disconnect worries.
Regards
Peter
-
.
It looks like a nice unit - designed and built in the US. And for aviation, we should not really be using the cheapest, especially if there is a chance of overheating on the inferior versions. Hopefully the electronics are of a similar standard......
I have asked NooElec if they can produce a dongle and remote aerial as a package, as the aerials on their package are not suitable. They will get back to me.
R
-
I've just ordered one of the single ones - https://www.amazon.co.uk/gp/product/B073JZ8CC2/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1
I'll let you know how it turns out ....
-
The Nano 3 arrived today..
Initial impressions are good - the build quality is very good, and the supplied heatsink / SMA-MCX adapter are nice touches ...
It's physically smaller than the 'stock' SDR so doesn't obscure the USB port it sits next to.
PAW has identified it straight away (shows as a NooElec NESDR Nano 3 in the home screen), and connection to the whip antenna brings in ADSB signals as you'd expect.
I'll try out various configs later tonight to see how it behaves with the USB extension and the tuned 1090Mhz NooElec antenna (though I have now 'tuned' my whip antenna, with noticeably better results...)
I'll leave it running for a while and see how hot it gets....
Scheduled to be flying tomorrow so I'll take it up with me and report back ...
-
Hi Sean,
My Nano 3 arrived today too..
I agree with your initial impressions, it certainly looks good quality and after plugging it in, mine was also immediately recognised as 'NooElec NESDR Nano 3' on the PAW Home Screen.
I have been running it at home with the standard PAW 1090MHz antenna and another SMA option and both performed as expected, so it certainly works fine.
It has been running for 12 hours now (without using the supplied heat sink) and although I haven't had any issues, my one concern is that it gets VERY hot - and the heat seems to be 'held' by the aluminium case and transferred through to the Pi USB block, which probably isn't good for the other components. It's now so hot in fact that I can only touch it for a few seconds without it actually burning my fingers!!
I will continue testing tomorrow with the heat sink attached.
Regards
Peter
-
HI Peter
I agree on the heat side of things, though more testing today with the Nano 3 in the short USB extension I'll be using removes the concern for the heat transfer to the Pi - So far there have been no issues using the Nano 3 with the USB extension and NooElec 1090Mhz antenna ...
Weather stopped any off airfield flying today so I didn't bother hooking the gear up in the cockpit for testing ....
-
It has been running for 12 hours now (without using the supplied heat sink) and although I haven't had any issues, my one concern is that it gets VERY hot - and the heat seems to be 'held' by the aluminium case and transferred through to the Pi USB block,
Thanks for that, Peter. I have asked the NooElec engineers if they would like to reply with any advice. Unfortunately, my Nano-3s have not arrived as yet. I was trying to organise a package deal with antenna.
Roger
-
The heat from the Nano 3 is sufficient for me to mount it outside on the V1 'All in one box' project ... and I've added the heatsink.
With the heatsink it's not so hot that you cant touch it anymore, but the extra space it needs means it wont fit inside !! (it's tight in there....)
I think the V2 one will need a slightly larger enclosure ... apologies for the quality of the pic ..
More to follow when it's finished .... (nearly there)
-
I think the V2 one will need a slightly larger enclosure ...
I made a similar enclosure, as the PAW would not last on the back-shelf of a glider cockpit.
No Nano to try yet, but will probably place it on a lead inside the enclosure.
R
-
.
I had a reply from the NooElec engineer, who said:
It is a very good thing that the device is hot, as it means the heat is not concentrated into a single smaller area where it is likely to cause problems down the line. Hopefully that makes some sense!
I presume this means that an insulated plastic casing to the chip might overheat the chip itself, rather than the casing, and ruin the chip. The heat has to go somewhere.
.
Unfortunately it is a byproduct of the main chips themselves, which we cannot change as we do not manufacture them. The best way to reduce power consumption is to use a host device with lower consumption (higher efficiency) and by ensuring the SDRs are not active when they are not being used.
By host device, I preume he means the aerial, so a tuned aerial would be better. I will confirm this with him.
Ralph
-
No, he means the RPi. As the SDR is required all the time, there is no power to be saved there.
-
Rog,
Presumably the engineer means... My bold!
'Its a very good thing that the device is hot, as it means that the heat is [being transferred to the metal casing (and equalised there *), rather than being retained inside the dongle and] concentrated into a smaller area where it is likely to cause problems down the line [namely causing failure of internal SDR components].
I don't necessarily disagree with him as a general principle - especially taking account of his second comment (which Ian has already covered), but in our case it means that if we want to use this particular SDR Dongle with PilotAware, we need to ensure that the heat is not simply transferred to the Raspberry Pi - which is what happens in practice - from where it can potentially cause damage to components in the RPi or in the other attached dongles.
If you decide to use this device, I would certainly suggest fitting it on a USB extender, adding an extra heat sink and either mounting it outside your case, or modifying the case to incorporate a cooling fan (which you might be advised to do anyway as the case seems to have no obvious provision for cooling the RPi).
* Note: I suspect that the use of a metal casing to retain heat is deliberate to maintain a constant temperature inside the dongle, so that the crystal oscillator doesn't drift off frequency, - which is the whole point of a Temperature Controlled Crystal Oscillator (TCXO). If I am correct, we don't want to cool the dongle excessively either, as that could negate the effectiveness of the dongle.
Regards
Peter
-
But in our case it means that if we want to use this particular SDR Dongle with PilotAware, we need to ensure that the heat is not simply transferred to the Raspberry Pi - which is what happens in practice - from where it can potentially cause damage to components in the RPi or in the other attached dongles.
If you decide to use this device, I would certainly suggest fitting it on a USB extender, adding an extra heat sink and either mounting it outside your case.
But is the Chinese dongle simply exporting its heat through the USB connector, instead of through its casing? Logic would dictate that both dongles consume about the same power, and thus give out the same amount of waste heat, which has to escape somewhere. And if it is not exiting through the casing, it must be going through the USB.
If that is so then the NooElec is probably better, as it is dissipating the energy away from the R-Pi. I do not have my NooElecs yet, so I cannot test this. The NooElec engineer said that their dongle consumes 280 ma of power on average, or about 1.5 w. Which is not very much, so I am surprised this results in such a high temperature.
Yes, by 'host device' the engineer did mean the R-Pi. I did ask if there were any tweaks that could be made to reduce dongle energy consumption, but it appears not. Even if less data is extracted from the dongle (like excluding weaker signals from contacts further away), it seems like the dongle will keep on doing its own thing.
Rog
-
The "low power" dongle has a switched mode power supply rather than linear, and apparently takes 100mA less than the standard dongle.
But as for the Nano 3 and heat dissipation, what happens with the standard dongle is that the chip (the R820T IIRC) gets hotter. The standard dongle does get hot and does heat up the Pi to some extent, but not by that much generally, as the heat transfer through the PCB isn't that great. What the Nano 3 does is dissipate the heat away from the R820T chip, but that means everything else gets hotter.
This doesn't show the Nano 3, but it does show the heat signature from some of the other dongles.
https://www.rtl-sdr.com/tag/low-power/ (https://www.rtl-sdr.com/tag/low-power/)
-
The Nano 3 does run considerably cooler (you can actually touch it) with the supplied heatsink installed ...
I'm running mine on a short extension due to the space available in my case - that said, it also keeps it away from the Pi's USB ports ...
-
But as for the Nano 3 and heat dissipation, what happens with the standard dongle is that the chip (the R820T IIRC) gets hotter. The standard dongle does get hot and does heat up the Pi to some extent, but not by that much generally, as the heat transfer through the PCB isn't that great. What the Nano 3 does is dissipate the heat away from the R820T chip, but that means everything else gets hotter.
You are talking about the Stratux low power SDR, which might be a good idea but it appears to be out of stock.
https://www.rtl-sdr.com/product/stratux-rtl-sdr-low-power-rtl-sdr-limited-stock-only
But this is not the Sodial SDR that normally comes with the PAW, which has the same power consumption as the NooElec.
https://www.amazon.co.uk/USB-DVB-T-Stick-RTL2832U-Receiver-Black/dp/B01LYGU4Z2/ref=sr_1_4?ie=UTF8&qid=1516194326
Not sure what you mean by: 'What the Nano 3 does is dissipate the heat away from the R820T chip, but that means everything else gets hotter'.
The heat only has two pathways out of the SDR dongle - either through the dongle casing or through the USB connector. Since the Sodial has a plastic casing, which cannot dissipate the heat, the heat goes through the USB and into the R-Pi. If you take the Sodial SDR out of the PAW unit and touch the USB, you can burn your fingers.
Rog.
-
Going back to the noise problem, this is an audio of the speaker sounds. In sequence they are.....
Normal PAW sound, with buzz and click
Silence with a battery source for the speaker.
Back to normal PAW sound, with buzz and click
Deletion of buzz, when I take out the SDR dongle.
Deletion of click, when I take out the wifi dongle.
https://instaud.io/1F2N
So the buzz is from the SDR dongle and the click is from the wifi dongle.
Rog
-
Hi Rog,
Just noticed your most recent post trying to bring the subject back on line, but as I have spent ages testing and typing this, I decided to post it anyway :(
[Paul]
You are talking about the Stratux low power SDR, which might be a good idea but it appears to be out of stock.
https://www.rtl-sdr.com/product/stratux-rtl-sdr-low-power-rtl-sdr-limited-stock-only
Out of stock at the moment and not available direct from Stratux, though reportedly available through Amazon.com (US site). Be careful though - I ordered a 'pair' from the same seller's (not Stratux) Amazon UK site but got the 'old' version - which draw pretty much the same current as the Sodial ones, not the 'new' Stratux Low Power ones.
But this is not the Sodial SDR that normally comes with the PAW, which has the same power consumption as the NooElec.
https://www.amazon.co.uk/USB-DVB-T-Stick-RTL2832U-Receiver-Black/dp/B01LYGU4Z2/ref=sr_1_4?ie=UTF8&qid=1516194326
Not sure what you mean by: 'What the Nano 3 does is dissipate the heat away from the R820T chip, but that means everything else gets hotter'.
The heat only has two pathways out of the SDR dongle - either through the dongle casing or through the USB connector. Since the Sodial has a plastic casing, which cannot dissipate the heat, the heat goes through the USB and into the R-Pi. If you take the Sodial SDR out of the PAW unit and touch the USB, you can burn your fingers.
Rog.
What Paul is (presumably) referring to is that TCXO dongles (Nano 3 etc) are designed to remove the heat from the 'chips' to the metal casing, which stops the chips getting too hot while also maintaining a constant temperature inside to control the temperature of the crystal oscillator (which is part of the circuit used to set the frequency). The metal casing should then allow excess heat to dissipate gradually from the casing, but the danger is that much of this excess heat can simply be transferred to the metal USB 'manifold'. In the case of the Nano 3, NooElec have obviously decided to supply an external heatsink because the dongle is so much smaller than 'normal' TCXO dongles such as their own NESDRSmart or RTL-SDR.com's RTL-SDR V3 and so won't be as efficient as dissipating excess heat directly from the casing.
The amount of heat generated with the standard PAW SDR dongles does vary quite a lot from dongle to dongle (these aren't high quality dongles), though I have never had any problems with any of mine. A few users, however, have had issues with SDR dongles failing (presumably due to heat damage), and in a few cases we have also had WiFi dongles melting and think that excess heat from the SDR may be a contributory factor in these cases.
If that is so then the NooElec is probably better, as it is dissipating the energy away from the R-Pi. I do not have my NooElecs yet, so I cannot test this. The NooElec engineer said that their dongle consumes 280 ma of power on average, or about 1.5 w. Which is not very much, so I am surprised this results in such a high temperature.
Yes, by 'host device' the engineer did mean the R-Pi. I did ask if there were any tweaks that could be made to reduce dongle energy consumption, but it appears not. Even if less data is extracted from the dongle (like excluding weaker signals from contacts further away), it seems like the dongle will keep on doing its own thing.
Rog
I ran a couple of fairly long term tests yesterday with my Nano 3, and have repeated them again this morning. With the dongle connected to the RPi through a USB Power Meter, yesterday's tests showed an initial current draw of 390mA, rising to a peak of 580mA, then settling to between 390 and 400mA. This morning's test started off at 220mA, which rose to 440mA when the PAW's 'Bridge' came on line (not sure why). The current then 'settled' at between 360 and 430mA which is significantly higher than the figure quoted by your NooElec engineer. This would equate to nearer 2 Watts, which is a fair amount of heat in a small dongle.
I'm not criticising, by the way as I really like the idea of small TCXO dongles, but the advantages have to balance out without creating additional problems elsewhere. I will continue testing.... :)
Regards
Peter
-
Going back to the noise problem, this is an audio of the speaker sounds. In sequence they are.....
Normal PAW sound, with buzz and click
Silence with a battery source for the speaker.
Back to normal PAW sound, with buzz and click
Deletion of buzz, when I take out the SDR dongle.
Deletion of click, when I take out the wifi dongle.
https://instaud.io/1F2N
So the buzz is from the SDR dongle and the click is from the wifi dongle.
Rog
Hi again Rog, very interesting audio post. You can clearly hear the interference and the effects of removing the SDR and WiFi dongles. I have to say that I am never aware of this level of interference in my plane, but then again I am flying in an open cockpit, so the same noises could well be present, just not audible.
I've just re-read the entire thread. A while back I suggested trying a 'Ground Loop Isolator' between the PAW and your audio interface. Did you ever get round to trying this? It's not really the optimal solution, but I and others have had considerable success using this method to reduce audio interference where the source was difficult to eliminate otherwise.
These are the sort of things I would try, I have used both to good effect previously....
https://www.amazon.co.uk/gp/product/B01IETQQQK/ref=oh_aui_search_detailpage?ie=UTF8&psc=1
https://www.amazon.co.uk/gp/product/B01LYYRT7M/ref=oh_aui_search_detailpage?ie=UTF8&psc=1
Worth a try.... :)
Regards
Peter
-
I have to say that I am never aware of this level of interference in my plane, but then again I am flying in an open cockpit, so the same noises could well be present, just not audible.
A while back I suggested trying a 'Ground Loop Isolator' between the PAW and your audio interface. Did you ever get round to trying this?
You would not notice any noise in a light aircraft, as your headset will have a difference electrical source to the PAW. I only have this problem because the PAW and the speaker run from the same dc converter (the speaker is attached to the PAW unit).
I did not know what a GL isolator was, but I see it is advertised to cure the very problem I appear to have. And it is not very big, so I think I will get one.
Cheers,
Rog
-
The metal casing should then allow excess heat to dissipate gradually from the casing, but the danger is that much of this excess heat can simply be transferred to the metal USB 'manifold'. R may be a contributory factor in these cases.
Still not sure about this argument.
The Nano will take the generated heat to the metal casing. This will dissipate much of the heat directly to the atmosphere, via radiation and gas conduction. So perhaps 60% is dissipated through the casing, and only 40% goes through the USB.
With the Sodial, the heat cannot be transferred to the insulating plastic casing. Since heat cannot be destroyed, it only has one option, which is to make the i ternal chips much hotter, and to eventually travel down the USB. So perhaps 90% of the heat will go through the USB in this case.
So the Nano has to be the better option (if these dongles are drawing the same current). BTW, the different current draw will be because of the number of contacts it is tracking. The more contacts, the greater the processing power required to decipher it all.
Thanks for the current-draw figures. I might go back and ask the engineer where he got his figures from. All I know is that I ran the complete PAW unit for 10 hours from from 12v, 5 a/hr battery. Admittedly there was not too much traffic to track, but that probably only works out at 4w average draw.
Cheers,
Rog
-
A while back I suggested trying a 'Ground Loop Isolator' between the PAW and your audio interface. Did you ever get round to trying this?
Th GL isolator was delivered this morning (why...?) via Amazon, and the problem is solved. This gadget really is a PFM box, because it was the noisiest chanell ever, as you heard, and suddenly it is quiet. All I need to do now, is find room in the box for yet another component.
I will still check out the Nano-3, when it finally arrives...
Thanks,
R
-
Hi Rog,
Thanks for the update and glad the Ground Loop Isolator has sorted your audio problem.
The Nano 3 actually seems to work fine, and with the heat sink fitted - although it still runs hot - 'HOT' is even printed on the casing - it's not so hot you can't touch it. It's just with my (fire service) background I get a bit paranoid about putting hot things inside a plane. (Engine is OK in my case - it's outside !)
Best Regards
Peter
-
The Nano 3 actually seems to work fine, and with the heat sink fitted - although it still runs hot - 'HOT' is even printed on the casing - it's not so hot you can't touch it.
Peter,
The Nano-3 arrived today. And it makes the same feedback buzz noise as the Sodial dongle, so no change there.
Regards the heat, it is much as I expected. If you run the Nano for an hour, the casing is very hot just as you say. But if you pull it out, the metal of the USB is almost cold. ... Conversely, the Sodial casing is only warm after an hour, but if you pull it out the USB is burning hot.
So it does appear that the Nano-3 is much better for the Ras-Pi motherboard, because it is dissipating the heat to the environment, rather than through the USB to the Ra-Pi.
So I think I am now ready for my second build.
DC converter,
Inline fuse,
Ground Loop Isolator,
Amplifier,
Speaker,
Nano-3 with radiator,
GPS on top of the Ra-Pi,
Wires galore.....
Its going to be a jam-packed box.....
Cheers,
Rog
-
Rog / Sean,
Thought you would like to know I have just arranged to return my Nano 3 to Amazon.
It crossed my mind yesterday, that during my recent tests, when I was concentrating on measuring the current drawn by the dongle, I hadn't noticed any traffic 'on screen'. At the time I subconsciously put it down to the heavy snow falling (Glasgow Airport was closed during part of the test period) and/or perhaps just one of Edinburgh's 'less-busy' periods, so thought nothing more about it at the time.
Thinking about it again later, I began to wonder, so checked my screen grabs from my initial Nano 3 tests on 8th January, which showed the Nano 3 'Connected' and recognised by PAW, with 1090 Traffic visible on the PAW Traffic Page and on Sky Demon, so it was working fine then. Unfortunately I hadn't bothered taking screen grabs while running the load tests, so I re-ran a test yesterday morning and the result was - Nano 3 Connected but no ADSB Messages and a Red ADSB 'light' throughout the test.
I repeated the test with 2 x different RPis, interchanging the Nano 3 with a standard 'Sodial Type' dongle, but using the same 'standard' antenna for both dongles that I had used for the previous tests. I was careful to only exchange the dongles when there was CAT ADSB traffic visible on screen inbound to Edinburgh (at least 3 aircraft with others following and some visible at altitude from my window). The traffic was visible on screen with the standard dongle, but disappeared when I replaced the standard dongle with the Nano 3 (yes, I waited for the dongle to become re-established each time) and reappeared again as soon as went back to the standard dongle. The results were exactly the same with both RPis, so it seems that the Nano 3 has now 'died', possibly resulting from (or in) the extreme heat and unexpectedly high current draw (compared to that quoted by NooElec) ?? :-\
Undeterred, I will order another one (or perhaps two) and continue testing. Hopefully, I just had a 'Friday duffer'.
Regards
Peter
BTW - If no-one has any objections, I'm thinking of asking the Mods to split this thread and move 'Answers 30 - 45. 47, 50 and 54' to a separate thread titled NooElec NESDR Nano 3. It will make both threads much easier to follow by anyone reading them in future.
-
Rog / Sean,
BTW - If no-one has any objections, I'm thinking of asking the Mods to split this thread and move 'Answers 30 - 45. 47, 50 and 54' to a separate thread titled NooElec NESDR Nano 3. It will make both threads much easier to follow by anyone reading them in future.
Good idea.
I will reply separately, in that case.
R
-
Peter,
Regards the noise from the speaker, I thought you might like to hear the results.
This recording has two 'PA' calls in the normal configuration, then some shuffling as I fit the Ground Loop, and two 'PA' calls with the Ground Loop fitted.
https://instaud.io/1HIo
R
-
Peter,
Regards the hot Nano-3 - did you manage to measure the current draw for the Chinese dongle?
(Compared to the 400 mA for the Nano...)
I must admit my Nano does get unbearably hot - although the USB is still stone-cold, which is a bonus.
So at least the Ras-Pi is not overheating...
I will do a long term test again, to judge the total time on a 7ah lead-acid battery (perhaps 5ah available).
The Chinese dongle managed 10 hours, before the system died, indicating a total draw of about 5amp.
(I don't have an ammeter....)
.
I noticed this blog report, that said the NooElec Nano-2 was very RF noisy (presumably meaning wasteful too), compared to other Nano dongles. And it ran 40ºf hotter too.
But this posting did not mention that the cooler dongles may be simply dissipating their extra heat into the Raspberry-Pi. The similar power consumptions of the two nano dongles would indicate that they are generating similar heat, so the temperature comparison may not be valid.
Again they are giving a 270 mA draw for the Nano-2, rather than your 400 mA draw for the Nano-3.
https://www.reddit.com/r/stratux/comments/614xn9/new_low_power_v2_sdrs_and_comparisons/#bottom-comments
Below is a graph of the RF noise from four different SDR dongles.
The Stratux low power V2 looks like a good option, if it goes back into production.
It appears to be not available at present.
Rog
(https://www.rtl-sdr.com/wp-content/uploads/2017/04/low_power_plot_stratux.png)
.
-
This is the Stratux low power V2.
But it is sold naked, and vulnerable. Not so bad if everything is inside an enclosure.
https://www.amazon.co.uk/Stratux-1090ES-UAT-Radio/dp/B01N9NNS2H
R
-
That link is NOT the Stratux low power V2. Ask me how I know 🤔
The correct version has a black PCB and no IR Detector, I have one.
Here are some tests I did, the top pair is the ones linked above.
The attached image shows the dongles collection I have, the power consumption in Watts and the temperatures in three locations with an IR spot thermometer. The three locations are 1. the chip and USB shield junction, 2. The R820T chip 3. MCX connector. Room temp was 70 degrees F and each SDR was run for 30 minutes before readings taken
The Stratus Low Power V2 has a white insulator in the USB plug and the PCB is reversed compared with the others. PCB is also black or dark blue.
I don't have nor intend to test the ability to hear weak signals at our chosen frequencies.
http://whmc.org.uk/RTL-SDR_Dongles.jpg
-
That link is NOT the Stratux low power V2. Ask me how I know 🤔
The correct version has a black PCB and no IR Detector, I have one.
I got the link from the Stratux page. The 5th advert down. It says:
Stratux RTL-SDR dongles are designed for use in ADS-B reception on small aircraft, but they can also be used by anyone requiring lower power usage. For example on solar powered Raspberry Pi units. Note that these units have no enclosure, and do not have a TCXO. When installing in Zadig these show up as "Low Power V2"
But I must admit it does not look like yours.
https://www.rtl-sdr.com/buy-rtl-sdr-dvb-t-dongles
R
-
The Stratus Low Power V2 has a white insulator in the USB plug and the PCB is reversed compared with the others. PCB is also black or dark blue.
Regards the Stratux V2. The engineer from NooElec says of the Stratux V2....:
Power consumption (of the Stratux LV2) should be about 60-80mA less, but it comes at the cost of all kinds of stray EMI being emanated from the device, as it utilizes switching power supplies instead of analog supplies to reduce power consumption. In other words, the primary ICs have the same power consumption as other SDRs. We are not comfortable offering such a product which is sometimes used in sensitive environments like cockpits, which is why we did not implement such a solution in our devices.
But this statement is the opposite of this graph, which shows the analogue powered NooElec Nano-2 being more noisy than the digital powered Stratux V2.
So who is the more correct here....?
Rog
(https://www.rtl-sdr.com/wp-content/uploads/2017/04/low_power_plot_stratux.png)
-
Hi Rog,
I have been keeping an eye on the rtl-sdr.com site for a while now, waiting for the Stratux dongles coming back in stock (I have now ordered from the China link). In the meantime the Amazon links generally led to the Amazon.com US site, which showed the older style dongles. I got caught out by this myself and after comparing the Amazon US and UK sites, ordered a Stratux ‘matched pair’ from the Amazon UK site, but when they arrived they were the ‘old type’ so I sent them back. Interestingly, the Canadian Amazon site shows the Low-Power version, but they won’t deliver to the UK!
Re the NooElec comments on the Stratux Low Power dongles, yes they are probably correct that the switch-mode power regulator Stratux use WILL generate additional RFI over a linear analogue regulator, but remember that NooElec design their dongles for use across the band as SDR ‘Radios’, whereas the Stratux ones are designed specifically for use on 1090MHz and 978MHz and have been designed to minimise RFI at these specific frequencies. To my mind, reducing power draw and excess heat is more relevant to us than a low noise floor across the full dongle frequency range. Ian’s tests in this respect are looking very positive.
Best Regards
Peter
-
I have been losing confidence in the NooElec. Every time I swap over from one to the other, the Stratux has more contacts than the NooElec. Unfortunately, I no longer have two units to compare at the same time.
I have reordered the V2 Stratux dongles direct from the RTL-SDR site. Not too sure about the naked pcb, as there is a lot of wiring in my repackaging that could apply pressure on the pcb and break it.
Cheers,
Rog
-
Rog, I have a spare case I can post you if you are not happy being naked :o
-
I have been losing confidence in the NooElec. Every time I swap over from one to the other, the Stratux has more contacts than the NooElec. Unfortunately, I no longer have two units to compare at the same time.
I have reordered the V2 Stratux dongles direct from the RTL-SDR site. Not too sure about the naked pcb, as there is a lot of wiring in my repackaging that could apply pressure on the pcb and break it.
Cheers,
Rog
I kind of know what you mean - during testing I always seem to get more contacts with the stock ADSB dongle - the NooElec one always seems to show less .... which is a shame as its really well built.
I don't yet have two Pi's to try them side by side, but should have shortly
-
Rog, I have a spare case I can post you if you are not happy being naked :o
Did not know you could get them.
Are they specific to this size dongle? On Amazon?
Cheers,
R
-
No, you cannot get them online. It's one I removed from a dongle.