Where did the 100 KHz go?

What would be nice (MK2) is no matching at all but a user output to select appropriate filtering/,matching circuits to cover the band in use. Generally a wide band front end will be totally overwhelmed by local strong signals, no matter how good it is. I know this because for example my Ettus B200 which has similar dynamic range, can’t be used without a filter due to a pager site a mile away. Everyone in urban areas is likely to have these problems, even in the US. So filtering is needed for any serious use.

HF only use of this device to me seems pointless given there are so many options out there, e.g. RFspace that have great performance, filtering and support. It is probably better not to use a silicon tuner at all. HF is baseband in effect. Using a direct connection to the ADC/DAC could give coverage down to DC that would avoid spurious responses associated with a direct conversion to 30MHz.

Finally - again in MK2 can we please use SMA connectors and put them on the edge of the board so they can be used directly like everyone else does?

Mike

1 Like

@andrewback
In my opinion, the proposed mod is a half of the fix.

RXn_L are low frquencies inputs. if RX2_L is unmodefied, this port is useless because higher frequencies are avalieble on RX2_W and RX2_H. More over, channel 2 will not work on HF.

If the mod consists in removing one component, why not propose to remove one more to get something more coherent ?

This is not an issue for those who can fix their boards by themselves, but not everybody will be able to do it.

Why the fix is not applied on the channel 2 ?

“high” is relative and the _L ports are by default optimised for frequencies lower than the _H ports. The _W ports are by default wideband, but at reduced performance and not suited to use in certain key applications.

It would compromise other use cases.

If you require HF optimised performance on two receive channels and are not prepared to make changes to the RF matching networks yourself, all I can suggest is that you request a refund.

@SimonG4ELI @martywittrock @nn4f_radio
This a screen shot of a local 2m beacon at 123 miles from me.
I have tried to set the display on SRC Console to be as close to identical as possible and they were received on the same computer running two instances of SDR Console.

The left hand side is the Lime the right hand side is the FCD+. I use a FCD+ as my test RX as IMO it is a very good weak signal RX on 2m. the only better RX I have is my Kuhne but I could not easily set that up for a side by side.

Settings…
Antenna is a M25WL on a 8ft test tower and NOT aimed at the beacon but slightly off to reduce the signal. The coax goes to a WA3LBI 6 pole 2m BPF and then into a Mini Circuits 2 way power divider feeding the Lime and the FCD+
Lime settings are built in LNA MAX External ARR LNA.
FCD+ built in LNA 0 and no External LNA.

This is not a lab test but just a real world test. I think I managed to get them pretty close on settings by eye. Anecdotal comment. The lime in this picture is about as good as I remember when comparing a TS2K to the FCD+. If I recall correctly I had to have and internal Pre amp and the same external LNA in inline for the TS2K to be equal to the FCD+ on the SDR Console waterfall. Again that is from memory and not a lab test. My view is based on what I think the average Ham will hear (or see) when connecting a radio up. I will leave the lab reports to those who can perform them better than I.

Paul
N2EME

2 Likes

Paul,

Looks good and it’s consistent with other non-scientific testing I’ve done, too, with the Lime on higher frequencies. I really couldn’t tell the difference between my Lime’s performance and that of my Yaesu 857 transceiver on 2m and 70cm…Equally sensitive and meter readings that were comparable…So your testing, while even better than mine, confirms what I’ve found with the Lime above 6m.

Thanks for passing that info along - that should reduce some of the noise based on posts that are requesting more info on receive performance across the band.

73 de Marty, KN0CK

Does anybody know what the effect on other frequencies (i.e. 1GHz - 3.8 GHz) of the modification is? Charts on the blog post at https://myriadrf.org/blog/optimising-limesdr-matching-hf/ only show the improvement over 10MHz to 1GHz. But testing of higher frequencies is absent. Thanks.

I’ve been using an AD6IW wide band preamp for testing through 3456. So far, things look OK. I wouldn’t call them great. I need to verify the noise head ENR that I’m using (it’s homebrew). Stay tuned.

The only band so far that appear have an issue is 3456. NF is pretty bad, but my preamp is not nominal up there. The preamp is the key to making the Lime work.

I am using the the LNA_W port with inductor removed for these tests.

I am using SDRConsole to measure NF. I want to try HDSDR with the EXTIO as there are a few settings in that software that could help with the measurements.

On the transmit side, I’m working with LimeSuite. I’ve been doing basic phase noise testing. Numbers are running in the high 80’s for dBc/Hz. As I look at the specs for the LMS7002, I should be about 10 dB better. I must be doing something wrong in the setup of LimeSuite. I am working in SXT, TBB, ClkGen etc.

I am using the TX1-1 port.

As reported earlier, External 10 MHz Ref appears to work and can be set up using LimeSuite in the ADF4002 Module.

Mike

At 100 kHz the reiceiver sensitivity is 80dB down below the specifications. Disappointing…

This in practice is not quite as dramatic as it sounds. At low frequencies the reveived signal levels are quite high. So a receiver must not have the full sensitivity - as long as you use a very good (efficient) antenna. An antenna always is good enough if its noise level is a couple of dBs (>6dB) above the receiver noise.

The background for the problems below 30 MHz is a small internal DC blocking capacitor in front of the LNA. It gives a high pass characteristic with a cutoff frequency of 30 MHz.
Following Fano’s rules, the high input impedance can not be matched to 50 Ohm over a wide bandwidth. There is no economic way to solve the problem. The LMS7002 should not be specified to operate below 30 MHz or even down to 100 kHz in the receive mode.

JMG,

When you were testing in the 100Khz to 10Mhz range …

lets say – on 3.970Mhz, Did the LNA gain bring in more noise floor or unwanted signals than the desired signals as the gain was increased?

My setup is far from ideal and when i pass about 5db of gain (gqrx) i start to see siganls that are not in that area. At that frequency (and lower) iv been setting the LNA to its lowest setting (in gqrx) for the best gain/SNR. TIA maxed and PGA about 15db.

I was about to post something similar, csete - but refrained from doing so on j4mbob’s post because it wouldn’t be fair to single his post out… As an M6 (Foundation) licence holder he doesn’t have access to the bands above 70cm (with the exception of 10GHz, which isn’t on offer from a (barefoot) LimeSDR.

However, in general, I agree with your comment - many posters seem to assume/think that Amateur Radio is purely about the 100kHz-30MHz section of the spectrum. That was not the attraction of the LimeSDR for me, as that slice of spectrum is well served by other SDR boards - the interesting thing about the LimeSDR is the upper end - where there is currently very little gear available at the same performence and price point.

I’m curious to know (if anyone who is affected by this problem of reduced performance at the bottom end would like to comment) - what was it prompted you to go for the LimeSDR rather than one of the other SDR offerrings that cover HF? I’m asking because I genuinely am curious - HF is not my thing, so I don’t know whether the LimeSDR promised something exciting tat you couldn’t get elsewhere.

73
John

I got LimeSDR because, what I read about it last year made me think this could be the one for all device with no gaps in receive frequency range. I thought I could use it for all Bands for receive and transmit. Maybe this was a bit starry-eyed, but as I am just getting in to this I was not able to interpret all of the specifications for my own.

@hf_user Do not be too despondent. I am seeing HF signals on the Lime with SDR Console. YES you will need a LPF or BPF but that was pretty much a given as obviously the front end had to be wide open. I have not run a comparison yet between the FCD+ and the Lime on HF but I was certainly seeing a lot on 20m. That said I am a VHF op so more than one station on a band at once can seem A LOT :slight_smile: We will know more about the TX side as soon as more software becomes available but as far as RX goes start playing and let us know what you see.

I was planning to use the TX1 output to drive a modified Pennywhistle
amplifier at 850 kHz to drive an AM broadcast transmitter. I was
disappointed to see on the WIKI that the TX1 output is woefully low at
this frequency, and is unlikely to provide adequate drive to the
Pennyswhistle.

It appears that the T1 is inadequate for this frequency. It looks as
though replacing this with a Mini-Circuits TC1-1T+ should correct this
problem for me. The card will not be used for anything over 30 MHz.

The TC1-1T+ has the same footprint and is rated to 0.4 MHz vs the TC-1-13MA+ used on the TI-1 output is only rated to 4.5 MHz. C6 is listed as 100pF on the schematic but 100nF in the BOM. If it is the former, that also needs to be changed or shorted. If MN1 and MN2, the bias coils are as marked, they should still be OK at 400 kHz.

Please let me know your thoughts on any other considerations on how to get at least +5 dbm from
this card at this low frequencies. I don’t want to be running -15 dbm
signals around, as the units will be the exciter in a 10 kW transmitter.

I really do not know how do you plan to squeeze the +5dBm from the device that is rated for a max output of 0dBm for a continuous wave ? If the LMS7002M datasheet says 0dBm you can expect only lower and not a higher level. If you add the losses in the matching network and transformer the -15dbm at your frequency of interest is more realistic approach. Frankly speaking, I exaggerate a bit, but -10dBm should be max you can get.

This can be solved with some external boards. As I am the guy behind one of the project, solution presently used to cover that HF problems, it will be not fair to advertise it from my side.

1 Like

Ok got my LimeSDR two days ago. I populated the wideband solution in RX1-L. The inductors are bad for soldering by hand, they have only contacts on the bottom, but it works now. I went outside and connected a PA0RDT Mini Whip active antenna to my Lime and got this:

I have no other SDR so I cant compare it to something. But on my (very old) Lowe HF-150 the Stations sounded the same. So lime is definitely not completely deaf on HF. For me (and I am no spcialist at this) HF works quiet well.

Unfortunately the discussion about getting HF to work well on Lime is split up in several threads and disrupted by unproductive posts in other threads. Maybe we could make a new thread where we post solutions that have been tested. Does someone have an idea how we can make results comparable?

2 Likes

@hf_user @martywittrock @W7BOT et al…

Firstly those pics look pretty good to me BUT to me anything below 6m is DC and I have very little knowledge so I will leave those who know to comment.

That said as you suggested I did set up a link on the Amature Radio section of this forum to discuss HF settings and improvements. There is already one there for 6m and up…

hf_user Was that the EasyFix1 or EasyFix2 that you did?

Neither of them. I erroneously wrote: [quote=“hf_user, post:199, topic:690”]
I populated the wideband solution in RX1-L
[/quote]
But I meant the Broadband solution. It is mentioned here https://wiki.myriadrf.org/LimeSDR_HF_Performance in the Table A at last. If someone needs help to populate this on his lime I can offer to help with that (Location 60km from Hamburg, Germany)

@hf_user,

I wouldn’t be despondent either and I’m having a real hard time believing, as a long time beta tester with the LimeSDR over the past year, that the LimeSDR with the coil removal modification is deaf. I’ve been running my V1.2 and my V1.4 LimeSDRs since I did the mod almost 3 weeks ago and it’s been nothing but comparable to some of the best receivers I have in my shack. I have a Flex 5000A and I have a Collins KWM-2A (and many more transceivers) that I can compare to and I cannot see where the LimeSDR cannot perform acceptably and on-par with any of them. In fact, it’s hard for me to tell the difference between the Flex-5000A and the LimeSDR - on a net that I participate on Saturday mornings I’ve purposely switched between the two and I cannot tell any difference in sensitivity on the 75m band in the early morning hours (6 AM). And for that matter I cannot tell the difference on 40m or 20m in the afternoon when those bands are active. And that’s comparing it to my V1.2 and my V1.4 LimeSDRs. Now admittedly, I have my LimeSDR with a LP-30A Kenwood Low Pass Filter to contain the LimeSDR front end to 1 - 30 MHz - that’s the only difference.

This whole discussion is pointless. The Lime has a wideband receiver that really needs a good preselector such that high-frequency content doesn’t interfere or degrade the HF performance. That’s the only rub the LimeSDR has (and we still need to check out transmit). Think about it…What receiver do you have in your shack right now that has the capability to receive from, realistically, 0.5 MHz to 3.8 GHz? Once you do the mods, that’s your tuning and receiving range with RX1_L, RX1_W and RX1_H. No kidding - what receiver do you have that can do that? Your LimeSDR can…!

Everyone has to stop all this whining about the Lime being deaf. It’s a work in process and eventually you’re gonna love its performance for receive and transmit. If anyone thought we were going to get it all right out of the chute, they’ve been smoking too much funny weed. If anyone but their money down thinking that the LimeSDR was an ‘appliance rig’ then they’ve lost their mind. It’s like the Red Pitaya - a development system where YOU get to call the shots on how it’ll end up. No one is kicking the Red Pitaya vendor in the pants about their SDR…That one is TRULY a development system - - I know, I bought TWO.

Okay - enough soapbox. Start getting happy about your Lime because like they say ‘downtown’…YOU AINT SEEN NOTHIN’ YET.

#EndRant de Marty, KN0CK

4 Likes

Marty, I picked up a Red Pitaya a couple of weeks ago. I was reading all the things Pavel Demin did and I needed to have one. That board makes a pretty serious SDR. I love these devices, LimeSDR included.

2 Likes