[TangerineSDR] [HamSCI] RE: Notes from PSWS / TangerineSDR call of 07-26-2021

Jonathan emuman100 at gmail.com
Wed Jul 28 15:16:22 EDT 2021


Phil,

This design (Integrated VLF antenna/preamp in PVC pipe with Raspberry Pi box) was used for an MIT study and the author of vlfrx-tools built a bunch of these antennas and boxes for those conducting the study. This was in 2018. Since this might have fallen under your umbrella, did you hear of it or know who the PI might be? The DC-DC converter has an operating frequency of 100-600 kHz, so it doesn’t really affect anything below that. Paul really liked the performance of his design. He even wrote a guide on how to set up the Radpberry Pi as a VLF SDR using his vlfrx-tools software. 

Jules,

I really appreciate the supplied schematic. Should I use the Schottky diode where you drew the regular diode, in parallel like you said? I could also probably solder the MLCC across the leads of the electrolytic and install it. Also, that TVS is bidirectional. Should a unidirectional TVS be used so it clips any negative spikes closer to 0V? 

David, 

Tom mentioned the same. I’ll be working on trying to fit an electrolytic and a ceramic capacitor. 

Khan,

This is good news. Email me if you have issues. 

Jonathan
KC3EEY



> On Jul 28, 2021, at 2:27 PM, Julius Madey <hillfox at fairpoint.net> wrote:
> 
> Jonathan,
> I still think you need to get rid of the series polarity protection diode and suggest the following schematic (with 22uf 50 volt Aluminum electrolytic)
> 
> 
> <eaeficdplljflikc.png>
> 
> The PTC poly fuse will open either on overload, sustained voltage above TVS rating or reverse polarity.  This will also snub transients on the line the series diode may have permitted.  There     might be some virtue in adding an MLCC cap of ~ 0.47uf in parallel with the electrolytic for lower overall ESR.
> 
> Regards,
> Jules-K2KGJ
> 
> 
> 
> 
> 
> 
>> On 7/28/2021 12:41 PM, Jonathan wrote:
>> Hi Tom,
>> 
>> That makes a lot of sense. I’ll try fit a 0.1uF and a 22-47uF electrolytic on the feedline loop. 
>> 
>> Thanks again! 
>> 
>> Jonathan
>> KC3EEY
>> 
>> On Jul 28, 2021, at 10:55 AM, Tom McDermott <tom.n5eg at gmail.com> wrote:
>> 
>>> Hi Jonathan - a concern is that the DC-DC converter input stage would become unstable and oscillate
>>> with the cable inductance.  If that happens then there could be high frequency AC voltages present at the
>>> input of the converter.  Would those over-voltage the input?  Possibly.   Would it radiate RFI from the cable?  Likely.
>>> 
>>> Good design dictates sufficiently low source impedance to the DC-DC converter.  Almost all linear regulators
>>> require it.  An electrolytic and ceramic bypass pair right at the DC-DC input would be considered good design by many.
>>> 
>>> -- Tom, N5EG
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>>> On Wed, Jul 28, 2021 at 7:30 AM Jonathan <emuman100 at gmail.com> wrote:
>>>> John, 
>>>> 
>>>> Thanks for the suggestion! I think I’ll start with the overvoltage age spike of 50V for 100ms or less based on the datasheet spec of the DC-DC converter. I was thinking of some sort of fusable circuit protection now that I’ll be using a TVS diode to mitigate stress from the follow current. 
>>>> 
>>>> 
>>>> 
>>>> Tom,
>>>> 
>>>> The datasheet wasn’t clear on the need for input filtering. On the output of both the IP2415S, I have 0.1uF ceramic capacitors. In the datasheet, it only mentions a spec of “Input Reflected Rated Current” with a series inductor and shunt capacitor. I don’t really understand exactly what this is, but it lists “20mAp-p through a 12uH inductor and 47uF                     capacitor”. Nothing else is mentioned about input protection or filtering.
>>>> 
>>>> In both situations that the DC-DC failed, the cable length was only 4’ of cat 5. This circuit that was proved and tested by Paul, including the receiver and schematic I attached, did not have such an issue, but the only difference is that the voltage was lower than 30V. That’s why I was thinking it was a turn-on/off spike. I think that in a lot of typical applications, the input voltage isn’t that high, so spikes don’t normally present a problem. The DC-DC that feeds the receiver DC-DC has never failed and it was always fed with 18V or less. 
>>>> 
>>>> 
>>>> 
>>>> Jules,
>>>> 
>>>> You recommend the Schottky diode in parallel with the TVS, then a series PTC, so during a spike, the TVS will conduct and the PTC would fuse, and during a reverse polarity condition, the Schottky diode will conduct with limited current due to the PTC fusing? I can see the benefit that the series PTC in normal a normal circuit condition would drop much less voltage that the Schottky diode. 
>>>> 
>>>> 
>>>> 
>>>> All in all, I agree with adding some capacitance to filter voltage spikes and reduce or eliminate them. I’ll need to be sure not to exceed the DC-DC max load capacitance, which is 47uF in this case. 
>>>> 
>>>> But, to me, the most plausible cause for the failures are turn on/off voltage spikes, even with a short length of cable. Adding electrolytics is a bit difficult due to the tight space constraint as seen in my previous email, so I’ll try out the TVS and a PTC. Other than voltage spikes, does anything else come                     to mind? 
>>>> 
>>>> 
>>>> 
>>>> Thanks guys. 
>>>> 
>>>> Jonathan
>>>> 
>>>> KC3EEY
>>>> 
>>>> 
>>>>> On Tue, Jul 27, 2021 at 9:34 AM Jonathan <emuman100 at gmail.com> wrote:
>>>>> Hi Everyone,
>>>>> 
>>>>> Before the meeting ended last night, I wanted to discuss the issues I’ve experienced with my VLF active antenna. As I mentioned, the DC-DC converter (XP Power IP2415S) failed after I applied power. I turned the volume up because I wasn’t hearing sferics, and shortly after, I heard the squeal of the DC-DC converter. When I pulled out the smaller foam core partially, I heard a squeal coming from the DC-DC converter itself. 
>>>>> <image3.jpeg>
>>>>> 
>>>>> I had a failure of the DC-DC converter when I                             first built the VLF preamp. It was being powered by the Raspberry Pi box (same power circuit as my text box). I had a power failure and when the power was restored, the                             DC-DC converter failed.  In both situations, it was being supplied with 30VDC, with the limit is 36V. The reason for using 30V is to overcome the voltage drop from long cable lengths.
>>>>> 
>>>>> After both failures, there was a low impedance across the input of the DC-DC converter causing the supply to be pulled down. Because the input and output are isolated, I figured the fault had to have been caused by a voltage spike on the input due to power being turned on and off. The datasheet mentions that it can withstand 50V spikes at 100ms, but I can’t think of anything else that could cause the DC-DC to fail, especially that 30V is used to power it and any voltage spike would ride on top of that 30V. I contacted XP Power about the issue. 
>>>>> 
>>>>> Here is the schematic of the power path. The                             blocks are the IP2415S DC-DC converter.
>>>>> <image1.jpeg>
>>>>> I’m powering the DC-DC converter with 30V from another DC-DC converter in the test box and Raspberry Pi box (Raspberry Pi box not shown, but same circuit). Keep in mind, the length is still relatively short in my testing, so cable inductance isn’t the primary reason for the spike. Plus, at long lengths, the voltage would be much lower due to voltage drop across the length, and I think this was why Paul might not have seen this issue when he originally tried and built this design. 
>>>>> 
>>>>> To mitigate this problem with the supply voltage at 30V, I decided to use a TVS diode across the input of the DC-DC converter. Here is the schematic: (my apologies for the TVS diode being backwards)
>>>>> <image2.jpeg>
>>>>> I chose a unidirectional TVS so there is also protection if the voltage spikes have any negative components. I picked a TVS with a working voltage of 33V, a Littel Fuse P4KE39A. According to the datasheet, it looks like this would be adequate. TVS diodes have a PN junction that is more rugged, with a larger surface area for greater current density. One thing I overlooked with this type of circuit protection is that once the TVS conducts, current from the power supply will be shunted too, known as the follow current, and in some situations, this may cause issues requiring a fuse. In this case, since the spikes are probably short duration, and since the DC-DC converter in the text/Pi box has short circuit protection, this should not be an issue. 
>>>>> 
>>>>> So the questions I have are: do you also think it could be voltage spikes that caused the failures, and if so, was my choice of TVS diode adequate. Also, one important question, if I connect the anode of the TVS diode to the receiver side ground, will I lose isolation? I know that during spike current conduction, I will. I’m sure I can still get adequate protection with the TVS diode anode connected to the negative of the isolated power loop too, but I just wanted to be sure. 
>>>>> 
>>>>> Lastly, since the magnetometer board will undergo another revision, I feel it may be worth it to add some sort or adjustable regulator or use of an external power source (not connected to the 5V pin of the Pi header) because even with paralleling wires or pairs, some installations will require very long lengths (especially since interference-free locations are often far away from the shack) and the voltage may drop to where the LDO cant regulate. Using a higher voltage at the Pi end (like I do with the VLF preamp) may be required for some installations. I think this should be considered if it already wasn’t addressed. 
>>>>> 
>>>>> Jonathan
>>>>> KC3EEY
>>>>> 
>>>>> 
>>>>> 
>>>>> On Jul 27, 2021, at 7:21 AM, Dr. Nathaniel A. Frissell Ph.D. <nathaniel.frissell at scranton.edu> wrote:
>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> Thank you, Tom.
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> The recording will be available later today at 
>>>>>> 
>>>>>> https://youtu.be/AammohuygMw and hamsci.org/telecons.
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 73 de Nathaniel W2NAF
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> From: TangerineSDR <tangerinesdr-bounces at lists.tapr.org>
>>>>>> 
>>>>>> On Behalf Of Tom McDermott via TangerineSDR
>>>>>> 
>>>>>> 
>>>>>> Sent: Monday, July 26, 2021 10:02 PM
>>>>>> 
>>>>>> 
>>>>>> To: TAPR TangerineSDR Modular Software Defined Radio <tangerinesdr at lists.tapr.org>
>>>>>> 
>>>>>> 
>>>>>> Cc: Tom McDermott <tom.n5eg at gmail.com>
>>>>>> 
>>>>>> 
>>>>>> Subject: [TangerineSDR] Notes from PSWS / TangerineSDR call of 07-26-2021
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> Notes from PSWS / TangerineSDR call of 07-26-2021
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 1. Bill is using chart.js for magnetometer charting. He is setting up a database using Django web and database framework for Python.
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 2. Scotty is looking at the Intel (Altera) Arria 10 GX FPGA 10GX270 for the version 2 Data Engine (supporting 10GE). These FPGAs appear to be more available than the MAX10 FPGAs. The intention is to develop DE Ver 1 and DE Ver 2 in parallel
>>>>>> 
>>>>>> while awaiting FPGA component availability. The 10 GX development boards are pretty expensive. 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> -- Tom, N5EG
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>>  
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> -- 
>>>>>> 
>>>>>> 
>>>>>> Please follow the HamSCI Community Participation Guidelines at http://hamsci.org/hamsci-community-participation-guidelines.
>>>>>> 
>>>>>> 
>>>>>> --- 
>>>>>> 
>>>>>> 
>>>>>> You received this message because you are                           subscribed to the Google Groups "HamSCI"                           group.
>>>>>> 
>>>>>> 
>>>>>> To unsubscribe from this group and stop receiving emails from it, send an email to hamsci+unsubscribe at googlegroups.com.
>>>>>> 
>>>>>> 
>>>>>> To view this discussion on the web visit https://groups.google.com/d/msgid/hamsci/SA0PR03MB5547CBF733EFEB4D829E659FF2E99%40SA0PR03MB5547.namprd03.prod.outlook.com.
>>>>>> 
>>>>>> 
>> -- 
>> Please follow the HamSCI Community Participation Guidelines at http://hamsci.org/hamsci-community-participation-guidelines.
>> --- 
>> You received this message because you are subscribed to the Google Groups "HamSCI" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to hamsci+unsubscribe at googlegroups.com.
>> To view this discussion on the web visit https://groups.google.com/d/msgid/hamsci/59FAEE9C-11F5-4CDF-B960-2E583E42E89C%40gmail.com.
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tapr.org/pipermail/tangerinesdr_lists.tapr.org/attachments/20210728/03ea51e5/attachment-0001.html>


More information about the TangerineSDR mailing list