@PhaseLockedLoop @THEkitchenSINK
I have two of these on the way (for the use-case of compensating for runs of coax line).
2 Likes
They have interesting issues
If you get one that runs hot it’s cuz of a defect. Also they need a clean 5V to run the PA. So you might as well power off of USB and charge the batteries with an adafruit ina3221 or 226 and solder the i2c lines from that to the rak for battery monitoring
1 Like
Yes sir, will be some redesign of the current “two radios one cup” build for sure. Happy to get this stuff fleshed out before I make the high ground build.
Deff getting obvious I like the building part more than the operation part.
Addendum to the pi print server, I needed to disable wifi power management else the pi zero kept falling off of the network.
Its getting to the point where my solar and batteries are overkill enough (and not far away from me for service) that I’m researching esp32 based mesh builds via home assistant ESPhome. The latter would make changes and updates a breeze and node logs easier.
And thinking of a T1000E as a daily carry
Firmware bugs can cause the t-echo LED pin to stay hot even after shutdown.
Going to put in a brute force manual switch recessed in the back using the 3d printed case thickness to act as the recess making accidental actuation less likely.
1 Like
Finding builds where 5v is simply run via usb-c and 3.7v battery plugged into wisblock and thats it. Wisblock apparently has step up internally for 3.7v battery operation at 1w tx load. Better yet when powering over USB the wisblock can charge the batteries at a higher amp limit than via the solar input connector.
If this is the case it simplifies my parts list greatly. My solar panels already output a regulated 5v to USB.
Don’t do that
What happens when you provide enough current via the USB provider from the battery and the wizblock tries to charge the battery?
You got a loop
Unless your solely providing solar based 5V and hope the onboard step up does a good job (it gets hot)
Looks like I won’t have to change a thing on my current builds power setup wise, just swap the board.
All my builds (x3) so far are 5v regulated output solar panels to rak solar input and 3.7v batteries connected to the battery connector. That combo should suffice for 1w operation.
Im also going to install a set of these for each case
OM f’ing G. The unicorn is coming.
Zigbee, zwave, thread, PoE
Because as much as id like to be 100% zigbee or 100% thread, there just isn’t enough device coverage (equivalents throughout the protocols). Basically this
2 Likes
Initially I didn’t want to use both the airframes cavity filter along with the wisblock 1w’s SAW filter but teh googles says this is good.
Using a Surface Acoustic Wave (SAW) filter inline with a cavity filter combines high-Q narrowband selectivity with steep-skirt out-of-band rejection, ideal for removing strong, nearby interference. Place the cavity filter first to handle high power and provide initial filtering, followed by the SAW filter for sharp rejection, while ensuring impedance matching.
What do you think? How much am I loosing by using a couple feet coax + cavity filter with a 4631 setup?
Reference: Cavity filter is 906.875 +/- 2mhz (Tastic in my region 906.875)
Wisblick 1w I’m assuming the SAW filter is for the full 863-928 MHz range, Meshcore in my region uses 927.875
With that info I was going to buy additional 1w’s so that my two mesh one cup builds (two of them now- meshtastic and meshcore in one box) are both 1w’s but I really think the meshtastic issues I’ve seen are at the firmware level and/or region’s user implementation/lack of firmware updates and/or the hellscape of the network for my region (saturated AF, some asshats having meshtastic nodes pump out stats data every second- why? Yes you can ignore them in the app but they are still stinking up the air waves, I think of them like packet collisions).
Right now my design philosophy is a PACE plan of meshcore is Primary with the 1w wisblock, no addition of the airframes filter as the meshcore in my region runs a freq outside of the airframes tune anyhow. Then meshtastic is my Alternate + just a simple chat feed like JS8call for HAMs to “monitor the situation” if one comes up- get about 10-20% of whatever meshtastic is coming over the air in my region. So meshtastic gets my now old spare 4631’s that are freed up from the 1w’s that arrived + use of the airframes filters I have, but erks me I’m using the high gain antennas with coax runs with such low power.
In addendum to the Alternate’s use-case, I’d swap Meshtastic back to Primary if making a small ad-hoc network remotely (camping, convention, off-road meetup/trail run etc) as then I feel it’s current implementation is it’s strength.
I agree with Andy over at hackaday:
And with The Comms Channel (TL:DR, same observations above and I’ve had)
I’d love to just build out my own private meshcore on a different freq but I just don’t have the location hookups to get the regions/friends I want.
Or would be cool to find a large enough meshtastic group in my region that has kicked over to a diferent freq/settings and it runs much cleaner (essentially what meshcore did in this region).
More than you’d like. If optimal insertion loss is somewhere in the neighborhood of two to three DB.. then you’ve lost half your power. Just from insertion loss at the optimal point
Like realize one of the things that we had to do in Utah is we realized that we needed to move to 1 W nodes because of this insertion loss.
It’s fine, just don’t expect giant miracles from it
The war between meshcore and meshtastic is pretty much a bunch of bullshit. There’s a lot of people that give meshcore credit but they’re not giving credit where credit is due. They’re just saying it’s better. There’s a few things it does better, but ultimately it will absolutely suffer from the problems that meshtastic suffers from because of the topology and the mesh network setup. It’s the mesh itself, the mesh topology that causes the problems that you are seeing on mesh networks. There’s no way around it. You can do your best to mitigate it. You can make it slightly more efficient and be slightly larger, but you’re eventually going to run up against the same fucking problem. I try to stay out of it because it’s like Linux distro wars. It’s pointless
Utah is pretty okay with both and we have big networks with both and both function relatively the same way. We may actually setup a reticulum network too for the extra paranoid
The TCP aspect might help with the drops I suppose.
Thats why I don’t treat it like a war, but just look at use-case and saturation. They are so similar its not like one is somehow totally immune to the other’s issues, its just the meshtastic air waves where I’m at matches our freeways- jammed packed. Not even up on the new setup that long and meshtastic is already showing about 1300 nodes.
Going forward I need to research and dig into if there is an ‘alt’ meshtastic network, tune to that possibly throwing the airframes filters in the parts box and move on.
Well, sh*t. Guess I’ll be buying some more 1w’ers later.
@PhaseLockedLoop @THEkitchenSINK
I’m such a derp, I finally finally broke the seal off of the nanovna H4 and checked some antennas.
The TE Connectivity ANT-916-CW-HW-SMA I got via Mouser is TOPS! Amazing SWR and broad for the 900-930 MHz you can find yourself playing in.
The little stubby antennas that came included with the 1w wisblock kits have almost a total open/high SWR unless holding it or touching it to a ground plain then its decent. Not sure how this will work on the t-echos I want to put it on so I’m going to research the little SMA rat tail mod.
The ALFA 915s didn’t really need a ground plane but their SWR was much better in the high 800s low 900s, better for my meshtastics at 906 but not good for meshcore 927.
Some rando Amazon (I can’t find the record so maybe mouser?) wifi looking hinged ones longer than the TEs did ok, not great, not terrible.
And the finale-
I had just redeployed my meshtastic + filter + ‘gucci’ Rokland uber dBi antenna with uber quality RFC400 3ft coax that could only RX and but TX meshtastic… so attached the nanovna
AND its maxed out SWR, some kind of open somewhere I guess. Explains that.
Wisblock 4631 looking back at me:
Man, screw all the different antennas and the big HAM radio looking stuff- ANT-916-CW-HW-SMA FTW - not that big, rock star performance.
Soooo back down with the antenna mast, remove some bolts, untape some stuff, lots of SWR readings to come, see whats breaking the link here and stubbornly try to fix what I have being a cheap ass vs. only going 100% direct mount to case antennas avoiding any external runs of coax.
1 Like
You can compensate with the gain of the antenna. A bit
So let’s say that you’re plugging in a filter that is going to take about 2 to 3 DB away. Well, then you can add two to three DB to your antenna accepting the consequences of what that does to the isotropy
Addendum addendum?
Might not have been that setting. It went down again where the Pi is on/LED is one but that service and SSH is down.
Changed the power source and now it seems to be reliable. So the 3 port USB wall charger from Amazon bosting it has enough juice for modern phones and more than enough for a pi-zero either doesn’t or the output is so dirty it locks up the Pi. Some rando single port one I have it on now that I’m pretty sure just comes shy of the recommended Pi’s requirement has been reliable.
1 Like
Alright really digging the T1000-E.
Its TX/RX’ing from inside the house and hitting repeaters and receiving just as good as rooftop client with large alpha 915.
Form factor is amazing. Lots of cool cases, if you are an iphone user there are magsafe cases.
Since I’m not in some rural sprawl my build philosophy is massively changing to small builds with small antennas.
1 Like