Poll: Remove electronic attenuator in future Kiwi-2 builds

jksjks
edited January 22 in General Chat

Update: Thanks everyone for voting and for the comments. Final result about 70% for removing the attenuator. But, we have decided to keep it for secondary reasons (e.g. the risk of making PCB changes mid-production, the distributors having to explain to customers why a feature suddenly disappeared without warning, the work required to make the necessary software changes, etc).

In a related development, Peter and Martin have been working on an SDR Protection Circuit that is now available in our store: kiwisdr.nz/products/sdr-protection-circuit See the product description for details and test results.

----

We're thinking of removing the electronic attenuator starting with the next Kiwi-2 build. And would like to get your opinion.

The attenuator has been nothing but a liability that has caused customers and us grief. And we don't think that many people actually use it. People keep destroying the attenuator by transmitting while the Kiwi is connected to a receive antenna. And once destroyed you either get a permanent high level of attenuation or no reception at all.

You could get away with this on a Kiwi-1 to some degree without causing damage. But the attenuator on the Kiwi-2 simply can't handle that kind of abuse. Worse, we don't have a cost effective repair program for these situations due to the two-way shipping and SMD repair costs. We have another solution, but both us and customers lose money in the process.

So please vote in the poll below and comment. Thank you.

Comments

  • Hello, I think the attenuator is a great idea, and I use it all the time. My antenna has a high output level. I attenuate at least 5 dB. But sometimes the KiwiSDWr input saturates (Tremolat KiwiSDR #2). I attenuate then by 10 dB. On days with very low propagation, however, I attenuate only by 3-4 dB.

    Martin had made a very good suggestion. That of controlling the attenuator automatically according to an average overload of the input converter. With a fairly large attack time and release time. The success of this automation will be in a good design of the envelope which will control the attenuator.

    Now, nothing serious. If this attenuator disappears in future manufacturing, we can always install a fixed or variable attenuator upstream of the Kiwi's HF input.

    I completely understand John's decision. If this attenuation IC causes problems, it should be removed or replaced with a more robust one, or protected at its input by large diodes. I installed two large protection diodes (30A peak) on my KiwiSDR.

    Best regards, Philippe, AutreRadioAutreCulture, Trémolat, France

  • edited January 19

    Could a physical jumper be implemented in future models to allow for the user to open the case and bypass the attenuator in the event it's unneeded or gets damaged? Assuming a small jumper(s) could be had that won't act like an antenna for internal noise.

    Or alternatively, how about an external optional electronic attenuator that can still be controlled by the Kiwi UI? That would also open up the option of an attenuator for those running KiwiSDR 1s. Something that could easily be attached to some GPIO and power pins via a short shielded cable.

    The attenuator is very useful given there are so many Kiwis in areas that are susceptible to overload, and conditions that change that can affect the amount of attenuation needed day by day or hour by hour. Such as AM stations changing between day/night power, nearby SW stations coming on and going off, or when using an antenna switch.

  • I successfully use an external attenuator for KiwiSDR1. The attenuator is assembled on the PE4302 board and ESP8266. Control is via USB connection to KiwiSDR. The attenuator works in a simple automatic mode. At night, a signal attenuation of 15 dB is used. In the daytime, the signal attenuation is set to 5 dB. An algorithm for controlling the attenuator in an intelligent mode with feedback by reading the ADC_overvoltage values ​​was also developed. Attenuation is selected automatically several times per minute so that ADC overloads are as small as possible.

    https://forum.kiwisdr.com/index.php?p=/discussion/2996/external-attenuator-for-kiwisdr-1

    Nate_RnitroengineTremolatjolo22
  • Personally, I think the internal attenuator is more trouble than it is worth.

    An external attenuator controlled by the KiWi, as a separate accessory, may be a better option for those that need to handle day / night propagation changes.

    However, in most cases, an amplitude / frequency slope equaliser and / or Broadcast band notch filters can fix most problems, without having to dynamically change the overall gain distribution once it has been initially set. Having a fixed system gain, makes it easier to observe changes in propagation and noise levels.

    Regards,

    Martin

    studentkra
  • I know a lot of Kiwi owners are most interested in HF and SWLing, but the public Kiwis are also very useful to AM/MW DXers for targeting or identifying signals that may be local or within range to those Kiwis. Unfortunately, there's many Kiwis in desirable locations that have been fitted with steep BCB filters, rendering their use for MW DX limited to useless.

    An ideal solution would be an add-on external attenuation module that works with the attenuator slider control in the Kiwi control panel. Sold as an accessory alongside the Kiwi. With a disclaimer on the attenuator accessory module that it is susceptible to damage if exposed to strong signals.

    Another external Kiwi accessory to look into would be to offer a range of tunable notch filters that could allow one or more strong problem signals to be attenuated without affecting the entire spectrum or resorting to filtering out an entire band.

  • edited January 20

    What input signal level can damage the PE4312? I'm ready to check it out. But I don’t have the necessary equipment (do I need a signal generator?) A similar attenuator PE4302 costs 0.8 USD (PE4312 1.4 USD) on Aliexpress including delivery.

  • Personally I vote for removal of internal attenuators in future designs, they tend to be weak/vulnerable in the end a pain, especially in high RF or active atmospheric environments.

    My preference leans toward hardier external attenuators which are relatively inexpensive, plentiful and easily replaceable.

  • Thunderstorm I missed months ago did not destroyed attenuator U401 but did L401 and U405 (LTC6401-20) which got shorted inputs. While replacing U405 I destroyed L405 and L406 (wire replaced). Attenuating still worked but 0dB was about S3 below (comparing to KiwiV1 which survived). After shorting L401, connecting U401 In with Out there is no waterfall or audio difference to KiwiV1 sharing now the same antenna. Should I put in place proper L401, L405, L406 parts? How can I know their specs? Should I remove attenuator?

    This is an embed external element. It can be deleted using the delete key or the backspace key. To view the full element, press the preview button below.
    Image
    thunderstorm-repair.jpg
    1.44 MB


    HB9TMC
  • jksjks
    edited January 26

    Hi Stan.

    Good job replacing the U405 ADC preamp. I'm not sure I could do that myself.

    In your description you say U401 for the attenuator (that you jumpered over). But on the Kiwi-2 schematic it is actually U408. U401 is the ADC.

    You might be seeing the -3 dB because L405, L406 and C411 form a 40 MHz low pass filter between the ADC preamp and the ADC. So if you replace the inductors with jumpers now C411 (100 pF) sits alone across the differential output of the ADC preamp which is not so good.

    Same with replacing L401 of the 30 MHz input LPF. I don't know what the response of that filter becomes with that inductor removed.

    I'm not sure if there's a problem with jumpering over the attenuator while it's still in circuit. I wouldn't think so.

    So you should really replace all those inductors. The part numbers of the inductors are as follows. You can substitute parts as long as they are wirewound and have similar Q factors. But the Murata parts should be easy to find.

    L401 330 nH Murata LQW18ANR33G00D

    L405 L406 150 nH Murata LQW18ANR15G00D


  • edited January 27

    I carefully studied the photo and it seems to me that the copper paths of the input and output of the attenuator are cut. That is, attenuator is completely disconnected from the circuit.

  • Yes, I think you're right. Should be okay then.

  • jks, thanks for inductors specification, you are right, my mistake, it is U408. After repair, attenuator set to 0dB still reduced signal to S3 below Kiwi1 S-meter. After crossing attenuator In-Out both kiwi S-meters almost equaled but attenuator slider and buttons on RF page still worked in some range.

    studentkra, you are right too, completely forgot it, could not then risk another damage to surronding elements so decided to cut In Out paths.

  • edited January 27

    I tried a short-circuited inductor L401of filter in the simulator. The characteristics have not changed much, only starting from 24 MHz there is a slight attenuation of the signal. About 3dB at 30 MHz.

    Also, the short-circuited inductors L405 and L406 at the output of the amplifier do not have much effect.

  • studentkra, thanks for interest. L401 simulation mirrors reality. Now both kiwis share the same antenna and we can compare. Months ago, after weeks of fixing attempts, nothing bothered me when finally got working radio which went to other place. Attenuator Poll reminded me that I took a photo after repair and decided to ask experts here.

    This is an embed external element. It can be deleted using the delete key or the backspace key. To view the full element, press the preview button below.
    Image
    kiwi-2above-1below.jpg
    399.85 KB


  • The 40 MHz LPF between the ADC preamp and ADC is important because it limits the amount of wide band noise going into the ADC for low frequency applications. See the Analog Devices data sheets for a discussion of this situation.

  • Yes, sure. But we already have an LPF in front of the preamp with a very steep decay level. And the filter in the differential output line of the preamplifier is just an addition. The decay level is very smooth since it is only first order. But in some situations it is necessary, as stated in the Analog Devices documents.

  • You will have additional aliased noise contribution from the LNA, which is another reason for the 40MHz LPF.

    Regards,

    Martin

Sign In or Register to comment.