Posts: 5
Threads: 2
Joined: Mar 2018
Reputation:
0
Hi All,
Got my Moitessier hat from the post office this morning, and all setup and tested.
Initial results have been very promising - working really nicely so far, got good GPS lock without external antenna, and quite a bit of AIS data with a fairly small antenna.
One thing I've found is that the reception / general range I seem to get with the AIS seems to change quite significantly depending on the power source.
That is to say - when I use mains with a standard pi transformer, I get significantly better range and reception than when running off 12v->USB power.
I was trying to think about possible reasons for this:
- Could it be that there is more fluctuation in voltage on the 12v circuit?
- Could it be related to grounding?
Or...any other thoughts?
It was quite pronounced and noticeable - i'll see if i have some screenshots with openCPN on my phone for comparison.
I've attached a quick picture of the setup.
Posts: 39
Threads: 1
Joined: Jun 2018
Reputation:
7
2018-08-04, 10:14 PM
(This post was last modified: 2018-08-04, 10:18 PM by pinguino.)
In general, the use of the AIS receiver is highly energy intensive. In addition, the new Rpi have more power and require a more intense energy source.
Often, products of Chinese origin are not of sufficient quality to meet this demand, I have seen for myself.
Your fault may be in the 12 volt to 5 volt converter -most likely - or in the cable that sends the current to the Rpi.
Most Chinese adapters ship at most 2.1 amps and that's not enough. You can use a second 12vlt to 5 volt adapter connected to the GPIO pins, it worked for me, or purchase one of the qc 3.0 adapters that usually delivers more power.
There are also pad-type adapters that deliver 3 up to 5 amps or even 10 depending on price.
Why are sailors more daring than other men?
Because: No man is so often afraid
like the sailors.
"Of the questions of the seafaring world", in
The Tree of Science, Ramon Llull
Posts: 1,250
Threads: 103
Joined: Mar 2016
Reputation:
74
Also some 12v/5v adaptors can be very noisy on RF. Possibly not helping.
Sent from my SM-T813 using Tapatalk
Posts: 45
Threads: 0
Joined: Jun 2018
Reputation:
5
Hi!
PaddyB has already mentioned a possible reason for your issue.
Most likely the performance degradation is caused by noise on the power supply, this might be either due to a noisy input source or the serial combination of the external switching regulator and the onboard switching regulator of the Pi, used to convert 5V to 3.3V. If using two switching regulators in series, they might oscillate and regulation is not reliable anymore. If the power source is unstable, sensitivity gets really worse.
Keep in mind to use low noise power supplies. The best solution would be a linear voltage regulator, no switching converter. However, in general, most converters are switching regulators due to its higher efficiency. It is absolutely crucial to use only high-quality converters!!
I will add a list of recommended regulators soon. If there are suggestions from the community, I will test any suggested converter in the lab to have reproducible results.
Regardless of this post, I have already done tests in the past, comparing the reception performance using two different step-down converters. The difference is quite amazing!!! I will publish a video of the comparison in the next few days.
Regards,
Thomas
Posts: 3,048
Threads: 63
Joined: Mar 2016
Reputation:
293
Interesting post, making it sticky...
Posts: 45
Threads: 0
Joined: Jun 2018
Reputation:
5
See a comparison of two different 12V to 5V converters. One uses a TPS 40057 and the other one a LM2596S switching regulator.
https://www.rooco.eu/2018/08/07/improve-...wer-supply
The results with the TPS regulator are really bad. You should not conclude from this, that you should avoid TPS based converters. The problem might not be caused solely by the switching regulator chip, but also by the converter design itself (capacitor values, capacitor types, inductor quality, PCB design etc.)
In general, do not use the cheapest converter you can get. Rule of thumb: The cheaper the converter, the higher the noise. Low-cost designs do not really care about noise, ringing, oscillation, etc. These converters might work satisfactorily to load your smartphone battery but are not recommended to power RF device, like the Moitessier HAT.
Posts: 43
Threads: 9
Joined: May 2016
Reputation:
1
2018-08-07, 05:21 PM
(This post was last modified: 2018-08-07, 05:22 PM by NahanniV.)
Maybe the next version of this hat could include a wide input, efficient, low noise power supply ?
It would be handy and increase reliability and repeat-ability.
Posts: 45
Threads: 0
Joined: Jun 2018
Reputation:
5
(2018-08-07, 05:21 PM)NahanniV Wrote: Maybe the next version of this hat could include a wide input, efficient, low noise power supply ?
It would be handy and increase reliability and repeat-ability.
The HAT is powered by the Raspberry Pi, so it does not include its own power supply, nor will it include one in the future. However, the voltage input on the HAT is filtered, so there is already a noise reduction onboard, but this is of course limited.
Choose a suitable converter and everything will work fine.
Regards,
Thomas
Posts: 176
Threads: 44
Joined: Mar 2016
Reputation:
0
Try a external Antenna with good grounding, that is as far as possible from the power regulator.
May be you can make a test with a masttop antenna. I guess the difference is gone.
Posts: 2,365
Threads: 21
Joined: Jun 2016
Reputation:
75
I could be wrong, but consider that "mains" power actually acts as a better antenna improving reception.
Not sure where the 12v -> 5v is coming from, but even if the 12v is from mains, it might have filtered this differently.
Try using a really good vhf antenna.
|