b66 and n71 TMHI Trash Can Hack

[McPierson]

Hi,

I'm new to the Antenna life, terminology, and best practices. I'm needing some guidance on next steps.

So I've also Hacked the T-Mobile issued 5G home internet device and installed 4 log periodic antennas (600mhz-4000ish) on the roof in a 4x4 MIMO ie two pair of antennas that are separated by about 8 feet.

It seems like I'm not able to hold onto the "Primary" b66 signal and that is kicking me off of "Secondary" n71 and into "Primary" B71, which is the home of bad speeds.

I think this is an aiming issue, but I'm not sure how to test that without spending a lot of time trial and error on the roof during a heat wave. So I have some questions about aiming.

Should I try to aim all of them at the same time or should I do two at a time?
Since I'm shooting through the middle or top of large trees, should I be pointing them up? If I don't point up some I can't get B66 at all (I could with one panel and two log pointed directly at the tower instead of up).

The antennas just barely clear the ridge of my house (not a metal roof). Do the antennas need some additional height above the ridge to work correctly?

Do the trees nearest my house make a bigger difference relative to trees away from the house? I have a good chainsaw.

Oher relevant info:

I also have a 2x2 MIMO panel that is not in use.

My nearest tower is 2.5 miles away and I would have line of sight but for the last 1/8th mile of Oak/Savana (according to a Google Maps Traverse).

I have determine the heading of the tower based on Google Maps and Cell Mapper.

TESTING:
No antenna:
I was able to get 250/50Mbps down/up sitting in my car looking at the tower a mile away.
I got 130/14 in the car 2 miles away with line of sight.
I got 20/25Mbps outside my house.
Inside with the window open I could also get 30/17 Mbps
With the window closed it would get 10 or 11 depending on traffic.

With two OEM antennas and 1 2x2 MIMO Panel (600mhz-4500ish):
(B71) 26/13 Mbps and stable 15 UP

With 1 2x2 MIMO Panel and two log periodic (no real aiming, but a best guess).
(B66/n71) 70/50 Mbps and apparently stable (I was pretty optimistic at this point and ordered a second pair of log-periodic. Limited testing).

With 2 pair of log periodic (exact as possible aiming based on a compass).
Best I can get is (B66/n71) 55/24. That's if it stays on B66 and n71.

More often it it drops back to B71 and I only get 10/10.

Thanks for taking the time to read this far.
Matt

[gscheb]

It seems like I'm not able to hold onto the "Primary" b66 signal and that is kicking me off of "Secondary" n71 and into "Primary" B71, which is the home of bad speeds.

There apparently is a known issue with this dropping 5g read somewhere and a new firmware update is suppose to work.
The old current firmware is 1.2003.03.0178

I think this is an aiming issue, but I'm not sure how to test that without spending a lot of time trial and error on the roof during a heat wave. So I have some questions about aiming.

I look at my phone or tablet device on the roof and look at the signal stats to tune it in. Logging onto the user interface 192.168.12.1 Can use this Signal graph to get an idea about what is good.

Signal Metrics.PNG

The antennas just barely clear the ridge of my house (not a metal roof). Do the antennas need some additional height above the ridge to work correctly?

Do the trees nearest my house make a bigger difference relative to trees away from the house? I have a good chainsaw

Yes being very close to the roof "could" cause some reflections. Specially if going across the roof too.

Yes. Close trees to the antenna cause more issues than trees farther away. Wouldn't go cutting things down step one though. Try some other tweaks. It might not be it.

After all this also what kind of cable and stuff is ran with this?

[Dr-BroadBand]

New firmware

See

https://tmo.report/2021/08/firmware-mr4 ... vpn-issues

[mtl26637]

I do not use the trash can setup, however, I'm working with the same Tmo 5G bands. You are correct about the bands, on my setup the go to bands are 4G B66 primary and 5G n71 and when things are working perfectly it will pick up an extra band of 4G B71 for CA.

If I was to guess it seems you need better signal on your B66 primary. I do not know the trash can antenna port layout but I do know that finding an antenna that has good signal at B66 (1700-2100Mhz) and also n71 (600-700Mhz) is very difficult. I have also tried various antennas and I ended up trying to stay away from the LPDA sytle since they are meant to cover a large signal range and 'okay' strengths. I think you need to look into the antenna ports and what signals are routed to each and then use something like a yagi to get as much gain from the single bands. Panels are sometimes good too so It's just all trial and error at finding the right combos to the right ports to get the most out of those 2 bands above. Speeds drop drastically with signal quality, that I can tell you for sure so most definitely the higher you can go the better.

[McPierson]

Thank you for these replies.

I'm still on the old/current Firmware.

The cable is 30' of LMR-400 with two adaptors, one from N to SMA and one from SMA to the Trashcan.
I don't know anything about grounding the antennas (they are not). I have lightening arrestors that I need to install and ground, but I've not gotten to that step yet.

My signal stats are TERRIBLE. I've never gotten more than 3 bars. It just jumps bands when the signal gets "Good" so that's part of the challenge for me.

In the User Interface:
B66 Primary two bars RSRP -116, SNR -4, RSRQ -19, RSSI -79
n71 Secondary three bars RSRP -96, SNR -5, RSRQ-16

OR
B71 Primary three bars RSRP -94, SNR 4, RSRQ-15, RSSI -62

[Dr-BroadBand]

Cable Loss, & two adaptors, one from N to SMA loss

[McPierson]

Yes being very close to the roof "could" cause some reflections. Specially if going across the Roof.

A6C11848-90BF-40F9-A625-34B5959F0C77.jpeg

[McPierson]

Should I just move the antenna to the other side? I hate more holes in the roof, but if the signal doesn’t improve I’ll probably take it down.

My Panel had SMA connections so that’s what I put on the modem, but if you’re saying each is 3dB loss maybe I should switch to N pigtails.

That would prevent me from using the panel at all though so im resistant.

In some level im thinking the panel might be required.

[McPierson]

Grounding: I currently have each pairs shield connected together (metal bracket) and that it not being connected to ground. Could that be part of the problem?

[gscheb]

In real world testing I have never noticed high loss on connectors.

[gscheb]

Wouldn't go making new holes and all that. Small steps first. Do you have enough length to just raise it up above the roof a little? Even a foot or two.

[gscheb]

Another small test is to play with what Antenna lead is to what modem lead on the can. Would start with one pair going L1 and L2. The other to L3 and L4.

[McPierson]

Wouldn't go making new holes and all that. Small steps first. Do you have enough length to just raise it up above the roof a little? Even a foot or two.

I don't have any length on the j-arm, but I think I could make something out of PVC to temporarily test that. Do you think 2 feet higher would accomplish anything?

[McPierson]

Another small test is to play with what Antenna lead is to what modem lead on the can. Would start with one pair going L1 and L2. The other to L3 and L4.

I do have it setup that way now. I wondered if I should consider L1 and L2 as the "Primary" antenna?

Is there ever a world where I point the two pairs in different directions? (one up and one straight for example)

[gscheb]

I don't have any length on the j-arm, but I think I could make something out of PVC to temporarily test that. Do you think 2 feet higher would accomplish anything?

It is possible. Just to get it away from the reflections of the roof. Can read this I posted along ago. Antennas can be finicky.
https://wirelessjoint.com/viewtopic.php?f=15&t=588#p3516

[gscheb]

I do have it setup that way now. I wondered if I should consider L1 and L2 as the "Primary" antenna?

Is there ever a world where I point the two pairs in different directions? (one up and one straight for example)

Would think the same direction but you never know.

[gscheb]

I do have it setup that way now. I wondered if I should consider L1 and L2 as the "Primary" antenna?
Is there ever a world where I point the two pairs in different directions? (one up and one straight for example)

Wouldn't hurt to try to switch one are around here and there. See if something just works out with it. It did one time on one of my antennas setup.

In my testing I firmly believe that L4 is the primary antenna.
https://wirelessjoint.com/viewtopic.php?f=32 ... 120#p18270

[McPierson]

Thank you for that link, I may have read it before, but I've learned so much that it was very helpful. Reading that thread it seems that I won't be damage my gateway to leave one or more antenna disconnected. Any concerns with disconnecting antennas? I seems like it might be easier to aim two antennas at a time.

[gscheb]

Reading that thread it seems that I won't be damage my gateway to leave one or more antenna disconnected. Any concerns with disconnecting antennas? I seems like it might be easier to aim two antennas at a time.

Far as I know there is now damage of concern. When I say it was disconnected. Mean there was SMA female pigtails are still connected to the can. Just don't have the outside antenna connected to it. Just an unconnected SMA female connected.
Really would try to switch those around and looking at the signal strength and quality. It is an easy shot in the dark.

[McPierson]

I agree. I'll try that first. It may be a few days before I can really get back into it... it's just too hot to be on the roof for more than 20 min.

[gscheb]

Switching around should be able to be done indoors would think. That is how I was doing it.

[McPierson]

So I’m back at it. I cut down 3 trees (for unrelated ish reasons).

With the leaves off the trees and with the new firmware I have better speeds and am now regularly connecting to B66 and n41 home of 100M download and 0.2M upload.

It seems like n41 is the preferred band for the modem, but with a bunch of dropped packets it will switch back to B66/n71 where I see 70 down and 40 up. This is with two 2x2 LP setups on the ridge of my house.

In an effort to make the best out of n41 I’ve raised two of the LP antennas and put back my Panel.

Now I’m seeing lower speeds, but for a bit it held onto B66/n71 and it was like 60down 25up. That’s acceptable, but after a bit it locked onto n41 and the suck resumed.

I’m currently experimenting with port combinations. My thoughts at this point is the gateway doesn’t really care which port the antennas are connected. This contradicts others work, but I’m still in this process.

I’m going to keep messing with ports then I’ll switch back to 4 LP and see if raising helps.

I’m interested in a filter to be able to block n41. I read about this in another thread.

Any thoughts?

[Dr-BroadBand]

DC-Car-Charger.png

May want to try a "clean power" using a car charger

https://thewirelesshaven.com/shop/power ... ctor-plug/

[McPierson]

So after questing my bad connection to n41 I suspected I might be on a Sprint tower that's about 30degrees away from the nearest known T-Mobile 5G tower.

Still have the log-parabolic on two ports and the panel on 2 ports

BANG! It locked onto B66/n71, but had low speeds, and low sinr... so I started inching the LP antennas back toward the T-mobile antenna and the speeds and signal increased until SNAP! back on N41. So I then switched the ports so that the LP antennas are split between ports 1/2 and 3/4.

Since that switch it's stayed on B66/n71 with 2 bars and something like 60/30.

I'm thinking the antenna is more directional at the higher n41 frequency. So my aim has me in a good gain at low frequency, but poor gain at higher frequency. For others that would be in this situation, it's been successful for about a month.

[Dr-BroadBand]

Thanks for the information, did your SINR also get better??

Most antennas have trouble with the low 600MHz.

[Didneywhorl]

... So my aim has me in a good gain at low frequency, but poor gain at higher frequency. ....

This makes sense because high frequency travels less and penetrates obstructions less, and low frequency travels further and penetrates obstructions better. Thus high received gain at low freq and low received gain at high freq.

Paradoxically the frequency wavelength is also why high freq on antennas is higher gain and low frequency is lower gain on the same antenna. (in general)

The low frequency is a physically wider "beam" and the higher freqs are a smaller tighter more focused beam.


....Or I'm dumb and am rambling after a long day. Either way, I think that's right.

[mtl26637]

So after questing my bad connection to n41 I suspected I might be on a Sprint tower that's about 30degrees away from the nearest known T-Mobile 5G tower.

Still have the log-parabolic on two ports and the panel on 2 ports

BANG! It locked onto B66/n71, but had low speeds, and low sinr... so I started inching the LP antennas back toward the T-mobile antenna and the speeds and signal increased until SNAP! back on N41. So I then switched the ports so that the LP antennas are split between ports 1/2 and 3/4.

Since that switch it's stayed on B66/n71 with 2 bars and something like 60/30.

I'm thinking the antenna is more directional at the higher n41 frequency. So my aim has me in a good gain at low frequency, but poor gain at higher frequency. For others that would be in this situation, it's been successful for about a month.

Let me see if I can make it a little more confusing, lol. Are you sure that it is "n41" that it is locking to and giving you slower speeds? From what I have seen in and around my area is that the old sprint towers have band 41, however, it is the normal width "B41" 4G band (typically 20mhz) and not the 5G "n41". Typically the "n41" 5G signals from the T-mobile towers are 100mhz wide compared to "n71" which is 20mhz wide so I'm very surprised that you get faster speeds from "n71". Maybe the band 41 you are seeing is regular 4G band 41? I'd be curious to know the width of the band 41 you were seeing or anyone's width for that matter.

You are correct in assuming the antennas are more directional at higher frequencies. 5G band n71 is from what I've seen the lowest frequency in use for cell connections using the ~600mhz frequency range. Most LTE antennas aren't designed to pick up anything lower than the ~700mhz range, but they still do just not at great strengths. Finding a great antenna that covers from n71 (600mhz) up to n41 (2500mhz) is gonna be tricky and even at that its gains are gonna be weak more than likely.

Best case scenario would be to know exactly what bands you are after and what modem antenna ports are assigned to what frequencies and then find specific high gain antennas for the correct modem ports. It would be nice to know the frequencies assigned to the specific ports on that device but don't think I've came across that information.

[McPierson]

Well, n41 is mixed. I get 120 down, great, but I get 1 up…. So yeah it’s n41 5g and n71 is better for me. I’m not the only person with this particular problem.

I’m not convinced there are frequencies assigned to a port. Maybe that’s a thing that HAS to happen, but the device itself doesn’t have a “front” so wouldn’t all the ports really need to be able to operate on all the bands?

[mtl26637]

Well, n41 is mixed. I get 120 down, great, but I get 1 up…. So yeah it’s n41 5g and n71 is better for me. I’m not the only person with this particular problem.

I’m not convinced there are frequencies assigned to a port. Maybe that’s a thing that HAS to happen, but the device itself doesn’t have a “front” so wouldn’t all the ports really need to be able to operate on all the bands?

Pretty sure n41 is download only and would explain why your uploads are slow. Your uploads would only be coming from your primary 4G band. Regarding ports, typically the RF front end of the device will indeed be tied to specific frequencies on specific ports. At least everything I've ever looked into does.

[McPierson]

Well, I'm using B66 for 4G when I'm connected to either N71 or N41, but it could be that N71 does upload....

I'll say that when I was connected to only B71 it had similar upload speeds to when B66/n71, so that might make sense.

Reguarding ports... I'm using both a 2x panel and 2 log P antennas, and I can't really see differences as I'm moving the ports around... Could be the device is compensating...