My only comment regarding the DIN barrels is that when you unplug them, if they are still powered and the tip contacts any part of the case, its dead short: and things let out smoke.
Phoenix connectors are protected in the sense that there is no conductor exposed. At all.
I accidentally touched the case with the tip of the barrel while re connecting a kiwi and luckily it only took out the small CM transformer at the input.
I had said previously that KiwiSDR 2 would fit in the existing aluminum box. But of course this is not true because there is now a third SMA connector (ext clock and self-test). And also the horizontal positioning of the other two (antenna/GPS) was widened to make room for it. The length of the PCB is the same but the end plate with the SMA connectors would have to change.
This isn't too much of a problem for purchasers of the bundled product because the plan is to include the enclosure, unlike before.
Also now my mechanical guy is suggesting I use those through-hole, right angle SMA connectors on the PCB as opposed to the current edge-mount ones we use now. This is for increased PCB connection strength and simplicity attaching to the enclosure. But that moves the centerline of the connectors "up" about 6.5mm. So that is a further incompatibility.
I don't think this is a huge issue, but I wanted to ask and get some feedback.
As a former manufacturing engineer, I would agree with through-hole SMA connectors (or anything else that connects to external cabling) for strength if you can do it and costs work for you. One less point of failure and we all know how much a piece of coax can torque a small connector.
Very much looking forward to v2 here. I own several v1 devices and would definitely be in the market for several of the new version as well. Also, happy to do any beta testing as I can put a v1 and v2 on the same antenna via a multicoupler for A/B testing or anything else needed.
I believe it would be ideal to use the through-hole connectors for a couple of reasons:
The same reason mentioned by CommSigma.
And, I imagine it's more than just me but, few people would be disassembling their Kiwi 1's (is it ok to call them that?) to replace just the radio board with a Kiwi 2. But rather I would be purchasing a whole new kit. So if the enclosure needs to change entirely, or if the end plate needs to be different, I say go for it. Especially if it helps improve robustness.
Another idea my mechanical guy had: Simplify assembly and keep the insulation of all 3 SMAs from the enclosure (as was done with the prior enclosure) by using sleeved plastic washers. They would slide onto the SMA barrels just before the PCB was installed into the enclosure. The depth of the sleeve would be flush with the outside of the enclosure.
The antenna SMA needs to be insulated in this design anyway, so do the same for the other two for consistency. No mounting hardware and associated assembly step required. Rely on the TH of the connector and fit of the SMA-washer-enclosure combination for robustness.
I split this sub discussion into a new thread called "Kiwi V/UHF noise" because the number of posts is going to make it more difficult for people searching for information about KiwiSDR 2.
I know it's been quiet recently. But you can't believe the amount of work going on behind the scenes. It's incredible. I had forgotten just how hard this process is. Even for a board that is mostly unchanged.
The first 5 prototypes are due back from MacroFab Houston in 2-3 weeks. Depends on shipping time to NZ which is always a wildcard. MacroFab is slow and not inexpensive. But man, their process is top tier.
JLCPCB is another story. I have been fighting them for over a month trying to get all the parts in stock to build 5 more prototypes. They don't even want to see your Gerbers and quote you a price until all the parts are in and ready to go. I can't really blame them for that. With the volume of orders they process it's the only strategy that makes sense. The important Kiwi parts have to be special ordered from Mouser/DigiKey as LCSC doesn't stock them. So lots of delays. In a way this is okay because the PCB has had recent changes. So the JLCPCB prototypes should be the final design including the thru-hole SMA connectors, SPI interface ferrite beads etc.
Enclosure prototypes are being fabricated now. The CAD models look great.
Looks like we will do manufacturing in NZ. And handle the initial order fulfillment ourselves. I want to have more control of the process this time.
There aren't going to be any new physical connectors added. It's much too late for that and it violates the prime directive of getting back into production as quickly as possibly with the least amount of risk. But it's an interesting idea and there are some other possibilities.
Also, the new attenuator will not withstand any sort of transmitter-level signals. If you're using a separate rx antenna then it might be possible as others have mentioned in the past. But a common antenna will always require an external T/R switch.
Now a software audio mute function could be added that is triggered from a Beagle GPIO pin (P8/P9 connector). You'd have to run a wire and add a buffer to avoid placing the Beagle and/or FPGA in danger. Or use a USB serial port input, Ethernet packet, etc.
It would be really nice for those of us who use the KiwiSDR along with an amateur radio station without having to deal with an external switch.
While an internal mute would be handy, there's some logic level driven relay boards available on the web that one could used for this purpose . The good news is that they're remarkably cheap. It's not an audio mute, but at least it should be capable of protecting the input of the kiwi.
Sure, I hear you. And this can be more properly addressed in the future when we can have a larger PCB to support changes like this. I have a long list of such changes and I'll make sure this is included.
The original Kiwi enclosure utilized brass standoffs for mounting.
I had found that I needed to replace these with smaller nylon standoffs while integrating the BBAI, which, also utilized two risers on each side of the BBAI board to make the cooling clearance needed for the heat sink and fans. This was also necessary to eliminate a common mode path
With the new design and inclusion of the T1-1 isolation transformer, it may be necessary to include the use of nylon spacers to avoid creating a CM path around the much needed isolation when it is installed in a metal case.
Similar to before we will offer two purchase options:
A complete system: Kiwi board, BB Green (software pre-installed), Enclosure, GPS antenna w/ 5m cable, SMA-M/SMA-M 200mm self-test cable.
Kiwi board alone: Tested, ready for you to add to your own BBB, BBG, BBAI or BBAI-64. You must install the Kiwi distro image onto an sd card and flash your BB.
The BBAI has cooling problems and I don't recommend it now that the BBAI-64 is available. The BBAI-64 has some advantages. Due to the faster ARM processor the waterfall will run faster when all 4 channels are being used: 12 fps vs 9 fps. And the AI-64 supports the rx14wf0 mode (14 rx-only channels) whereas the BBB/BBG does not.
The general impression I got was that it was too soon to be spending time on it. Especially now. Better to wait for the dust to settle (and for other people to be doing some debugging) before trying to use it.
I see, I just checked it and looks like there's some issues needed to be ironed out. Looks like I might settle with BBAI-64 or a BBB/BBG for KiwiSDR 2 use. Thanks!!
I am definitely ready for two ready-to-go units whenever and however. Good thoughts going into the -2 model. I know you know that "feature" creep will have to stop at some point and production start. I'm ready. Can I just put in a cautionary word to suggestors, most of whom know this already, but a software revision will also be necessary. . .
Sure. And in the last few releases I've been adding the changes necessary to support Kiwi-2 as I test the prototypes. So a single distro will support both. It makes my life easier.
Any insight on the internet bandwidth requirements for the KiwiSDR2? Will it be similar/the same as the 1, around 320 kbps max with 4 channels?
I've got about 500-600 kbps upload speed reliably at my remote QTH, so the Kiwi fits the bill better than several other options...just want to confirm the 2 will be similar.
Comments
If only Phoenix Connector made them, it would be great. :-)
This will be awesome!
I loved the v1, hate myself that I fried it and will DEFINTATELY be purchasing the v2.
My only comment regarding the DIN barrels is that when you unplug them, if they are still powered and the tip contacts any part of the case, its dead short: and things let out smoke.
Phoenix connectors are protected in the sense that there is no conductor exposed. At all.
I accidentally touched the case with the tip of the barrel while re connecting a kiwi and luckily it only took out the small CM transformer at the input.
Kickstarter is the way to go for new projects
I had said previously that KiwiSDR 2 would fit in the existing aluminum box. But of course this is not true because there is now a third SMA connector (ext clock and self-test). And also the horizontal positioning of the other two (antenna/GPS) was widened to make room for it. The length of the PCB is the same but the end plate with the SMA connectors would have to change.
This isn't too much of a problem for purchasers of the bundled product because the plan is to include the enclosure, unlike before.
Also now my mechanical guy is suggesting I use those through-hole, right angle SMA connectors on the PCB as opposed to the current edge-mount ones we use now. This is for increased PCB connection strength and simplicity attaching to the enclosure. But that moves the centerline of the connectors "up" about 6.5mm. So that is a further incompatibility.
I don't think this is a huge issue, but I wanted to ask and get some feedback.
As a former manufacturing engineer, I would agree with through-hole SMA connectors (or anything else that connects to external cabling) for strength if you can do it and costs work for you. One less point of failure and we all know how much a piece of coax can torque a small connector.
Very much looking forward to v2 here. I own several v1 devices and would definitely be in the market for several of the new version as well. Also, happy to do any beta testing as I can put a v1 and v2 on the same antenna via a multicoupler for A/B testing or anything else needed.
Rob
I believe it would be ideal to use the through-hole connectors for a couple of reasons:
The same reason mentioned by CommSigma.
And, I imagine it's more than just me but, few people would be disassembling their Kiwi 1's (is it ok to call them that?) to replace just the radio board with a Kiwi 2. But rather I would be purchasing a whole new kit. So if the enclosure needs to change entirely, or if the end plate needs to be different, I say go for it. Especially if it helps improve robustness.
Okay, so I think this might work.
Another idea my mechanical guy had: Simplify assembly and keep the insulation of all 3 SMAs from the enclosure (as was done with the prior enclosure) by using sleeved plastic washers. They would slide onto the SMA barrels just before the PCB was installed into the enclosure. The depth of the sleeve would be flush with the outside of the enclosure.
The antenna SMA needs to be insulated in this design anyway, so do the same for the other two for consistency. No mounting hardware and associated assembly step required. Rely on the TH of the connector and fit of the SMA-washer-enclosure combination for robustness.
I split this sub discussion into a new thread called "Kiwi V/UHF noise" because the number of posts is going to make it more difficult for people searching for information about KiwiSDR 2.
If there were a kickstarter for a Kiwi 2, I would sign up for at least one, probably two.
It would be nice to both add a little extra monitoring capacity, and to have a spare one for doing development (and lower priority monitoring?).
I know it's been quiet recently. But you can't believe the amount of work going on behind the scenes. It's incredible. I had forgotten just how hard this process is. Even for a board that is mostly unchanged.
The first 5 prototypes are due back from MacroFab Houston in 2-3 weeks. Depends on shipping time to NZ which is always a wildcard. MacroFab is slow and not inexpensive. But man, their process is top tier.
JLCPCB is another story. I have been fighting them for over a month trying to get all the parts in stock to build 5 more prototypes. They don't even want to see your Gerbers and quote you a price until all the parts are in and ready to go. I can't really blame them for that. With the volume of orders they process it's the only strategy that makes sense. The important Kiwi parts have to be special ordered from Mouser/DigiKey as LCSC doesn't stock them. So lots of delays. In a way this is okay because the PCB has had recent changes. So the JLCPCB prototypes should be the final design including the thru-hole SMA connectors, SPI interface ferrite beads etc.
Enclosure prototypes are being fabricated now. The CAD models look great.
Looks like we will do manufacturing in NZ. And handle the initial order fulfillment ourselves. I want to have more control of the process this time.
More information soon..
Thank you for the info!
With the addition of the attenuator is there any possibility that an input could be added for muting via an external PTT signal?
It would be really nice for those of us who use the KiwiSDR along with an amateur radio station without having to deal with an external switch.
Well, it depends on what you mean.
There aren't going to be any new physical connectors added. It's much too late for that and it violates the prime directive of getting back into production as quickly as possibly with the least amount of risk. But it's an interesting idea and there are some other possibilities.
Also, the new attenuator will not withstand any sort of transmitter-level signals. If you're using a separate rx antenna then it might be possible as others have mentioned in the past. But a common antenna will always require an external T/R switch.
Now a software audio mute function could be added that is triggered from a Beagle GPIO pin (P8/P9 connector). You'd have to run a wire and add a buffer to avoid placing the Beagle and/or FPGA in danger. Or use a USB serial port input, Ethernet packet, etc.
It would be really nice for those of us who use the KiwiSDR along with an amateur radio station without having to deal with an external switch.
While an internal mute would be handy, there's some logic level driven relay boards available on the web that one could used for this purpose . The good news is that they're remarkably cheap. It's not an audio mute, but at least it should be capable of protecting the input of the kiwi.
John K5MO
Sure, I hear you. And this can be more properly addressed in the future when we can have a larger PCB to support changes like this. I have a long list of such changes and I'll make sure this is included.
Hi John, great news about the prospect of an SDR2. I'll certainly grab another one when (assuming) production starts up.
73
Colin
ZL2FL
The original Kiwi enclosure utilized brass standoffs for mounting.
I had found that I needed to replace these with smaller nylon standoffs while integrating the BBAI, which, also utilized two risers on each side of the BBAI board to make the cooling clearance needed for the heat sink and fans. This was also necessary to eliminate a common mode path
With the new design and inclusion of the T1-1 isolation transformer, it may be necessary to include the use of nylon spacers to avoid creating a CM path around the much needed isolation when it is installed in a metal case.
I would like to buy Kiwi sdr v.2.
Do you know approx. price?
73'
To prepare for the next KiwiSDR 2, which BeagleBone do you recommend I buy?
Thanks and 73's de IK8SUT Antonio
@IK8SUT jks wrote earlier, that the plan was to sell the Kiwi2 bundled with a BBB.
But I do have also a spare BBB waiting for a Kiwi board :D
You may not need to buy your own BeagleBone.
Similar to before we will offer two purchase options:
The BBAI has cooling problems and I don't recommend it now that the BBAI-64 is available. The BBAI-64 has some advantages. Due to the faster ARM processor the waterfall will run faster when all 4 channels are being used: 12 fps vs 9 fps. And the AI-64 supports the rx14wf0 mode (14 rx-only channels) whereas the BBB/BBG does not.
Will the BeagleV-Ahead work with the KiwiSDR 2? I'm complimenting on getting the BBAI-64 or that.
Re the BeagleV-Ahead: Nothing has been done about that yet. We had a thread discussing the situation https://forum.kiwisdr.com/index.php?p=/discussion/comment/16419#Comment_16419
The general impression I got was that it was too soon to be spending time on it. Especially now. Better to wait for the dust to settle (and for other people to be doing some debugging) before trying to use it.
I see, I just checked it and looks like there's some issues needed to be ironed out. Looks like I might settle with BBAI-64 or a BBB/BBG for KiwiSDR 2 use. Thanks!!
I am definitely ready for two ready-to-go units whenever and however. Good thoughts going into the -2 model. I know you know that "feature" creep will have to stop at some point and production start. I'm ready. Can I just put in a cautionary word to suggestors, most of whom know this already, but a software revision will also be necessary. . .
THANKS!
Don
But a software revision will also be necessary...
Sure. And in the last few releases I've been adding the changes necessary to support Kiwi-2 as I test the prototypes. So a single distro will support both. It makes my life easier.
I moved the KiwiSDR 1 external attenuator posts to their own thread.
Any insight on the internet bandwidth requirements for the KiwiSDR2? Will it be similar/the same as the 1, around 320 kbps max with 4 channels?
I've got about 500-600 kbps upload speed reliably at my remote QTH, so the Kiwi fits the bill better than several other options...just want to confirm the 2 will be similar.
-Nate
Exactly the same.