SDS100/SDS200: DMR Decode Issue on SDS100E

[Neutrino222]

Hi

Sometimes SDS doesn't decode DMR, but sometimes it takes a few seconds to decode. The signal is coming in sharp and clear with a low Db value. Is this considered normal operation for this unit?

As a side note, my Whistler TRX-1 decodes every DMR instantly, but I'd like the SDS to be my mobile unit - it's not great when DMR tones rattle my cranium through the headphones. Perhaps I'm missing a setting?

Cheers

 

[Ubbe]

Sometimes SDS doesn't decode DMR, but sometimes it takes a few seconds to decode. The signal is coming in sharp and clear with a low Db value. Is this considered normal operation for this unit?

Unfortunately it is. The firmware code are being reused between scanners, why invent the wheel again, and the code where compromised some years ago in the BCD536 when a feature where added, I think it was RAS handling for DMR. The code have been broken since then.

When RAS are used in a system the forward error correction cannot be used in a scanner to correct bit errors that almost always occur when sending information using airwaves. My guess are that there is a bug and the code are always used that doesn't use error correction even when there's no RAS in use. I hope that JoeBearcat can convince Uniden to take a look at it when they are not busy fixing firmware to support new hardware and new scanners, that brings in money contrary to fixing bugs for free in existing scanners.

/Ubbe

 

[Neutrino222]

Unfortunately it is. The firmware code are being reused between scanners, why invent the wheel again, and the code where compromised some years ago in the BCD536 when a feature where added, I think it was RAS handling for DMR. The code have been broken since then.

When RAS are used in a system the forward error correction cannot be used in a scanner to correct bit errors that almost always occur when sending information using airwaves. My guess are that there is a bug and the code are always used that doesn't use error correction even when there's no RAS in use. I hope that JoeBearcat can convince Uniden to take a look at it when they are not busy fixing firmware to support new hardware and new scanners, that brings in money contrary to fixing bugs for free in existing scanners.

/Ubbe

That's very naughty of them, especially since one must 'pay' for the DMR feature, but it comes with a broken code and thus patchy decodes! Had I have known beforehand!

 

[Neutrino222]

Unfortunately it is. The firmware code are being reused between scanners, why invent the wheel again, and the code where compromised some years ago in the BCD536 when a feature where added, I think it was RAS handling for DMR. The code have been broken since then.

When RAS are used in a system the forward error correction cannot be used in a scanner to correct bit errors that almost always occur when sending information using airwaves. My guess are that there is a bug and the code are always used that doesn't use error correction even when there's no RAS in use. I hope that JoeBearcat can convince Uniden to take a look at it when they are not busy fixing firmware to support new hardware and new scanners, that brings in money contrary to fixing bugs for free in existing scanners.

/Ubbe

By the way, is there a thread on this issue here on RadioReference?

 

[zerosix]

I also recently switched from a TRX-1 to a SDS100E and I have noticed it as well. Can it somehwat be remedied by changing settings? (there is some digital dwell timer or something that is on 400ms as default if my memory serves me right.)

But from reading the other comments, it seems like the decoder has a software problem.

I paid for all the extra upgrades and I truly hope Uniden will step up its game and continue to improve the SDS line in terms of software.

 

[Neutrino222]

I also recently switched from a TRX-1 to a SDS100E and I have noticed it as well. Can it somehwat be remedied by changing settings? (there is some digital dwell timer or something that is on 400ms as default if my memory serves me right.)

But from reading the other comments, it seems like the decoder has a software problem.

I paid for all the extra upgrades and I truly hope Uniden will step up its game and continue to improve the SDS line in terms of software.

From what I can gather, the fault only seems to be prevalent in the lower bands (UHF - lower, and VHF. Here in UK, most of the comms take place in those bands, so it makes DMR impractical to use. I've only had it a couple of days, but I already notice my TRX-1 is making it look rather stupid.

When I have time, I will methodically trawl through the settings (dwell time, too) to see if I can increase its performance. I did have a play around with building a simplex DMR system today, but still the issue persisted. I also took it out in the town with a supergainer and experienced the same issue. At home, it runs on a yagi tuned between 400 and 470.

If you come up with anything, please feel free to post it here.

EDIT: I've just read a post confirming it happens in the upper part of the spectrum, too.

 

[Ubbe]

It helps if the scanner are allowed to decode the data for a longer time period. If you set the system hold time to 5 seconds it could start to decode more properly, when the bit error during that time eventually gets better.

/Ubbe

 

[Neutrino222]

I also recently switched from a TRX-1 to a SDS100E and I have noticed it as well. Can it somehwat be remedied by changing settings? (there is some digital dwell timer or something that is on 400ms as default if my memory serves me right.)

But from reading the other comments, it seems like the decoder has a software problem.

I paid for all the extra upgrades and I truly hope Uniden will step up its game and continue to improve the SDS line in terms of software.

Right, I've had a play around with the damn thing, and I've managed to increase the DMR decode success rate to around 90-95%, by simply changing the filter to WIDE INVERT. However, after some research into the filters, they seem to have different results for different radios in different locations, so maybe you can try and mess around with some of those filters. I also switched the DMR frequencies from AUTO to NFM, as one of them kept hanging.

As per Ubbe's post, it looks like Uniden tried to compensate for this decoding error by not actually fixing it, but instead providing some filters to try and manage it. Better than nothing, I guess!

 

[Ubbe]

As per Ubbe's post, it looks like Uniden tried to compensate for this decoding error by not actually fixing it, but instead providing some filters to try and manage it. Better than nothing, I guess!

It's the poor SDR receiver chip that is wide open, some 10MHz wide as it is designed for those kind of wide satellite signals, to all other signals that will create interferences and mixing products in the receiver. The different filter settings are there to try and find a frequency range with as little interfering signals as possible, by moving the 10Mhz wide window up or down in frequency by different amounts of MHz but still keep the monitored frequency within that window, as the filter has a fixed bandwidth.

/Ubbe

 

[zerosix]

It's the poor SDR receiver chip that is wide open, some 10MHz wide as it is designed for those kind of wide satellite signals, to all other signals that will create interferences and mixing products in the receiver. The different filter settings are there to try and find a frequency range with as little interfering signals as possible, by moving the 10Mhz wide window up or down in frequency by different amounts of MHz but still keep the monitored frequency within that window, as the filter has a fixed bandwidth.

/Ubbe

yeah, it certainly seems that way. Totally crazy that they did not opt for the R820T2 or something similar.

 

[Ubbe]

yeah, it certainly seems that way. Totally crazy that they did not opt for the R820T2 or something similar.

They probably consider it but it was on its way to be obsolete and the R836 promised better performance and still just costs 85 cent if you buy one from China. It was soon obsolete and where replaced by R840 with the same poor performance that are now used in both SDS models. They behave exactly as the R820, that also are obsolete and replaced by R826 or something similar, when it comes to internal spurious signals and overload issues.

/Ubbe

 

[Neutrino222]

yeah, it certainly seems that way. Totally crazy that they did not opt for the R820T2 or something similar.

With some more tinkering, I've managed to get 100% decode. I've dropped all DMR frequencies into one favourites list, and by scanning just that one list I get a perfect decode. I'm not able to say why that is. Perhaps you or Ubbe would have a theory about it. Mixing an abundance of various modes may throw the decoding processes somewhat- I really don't know. Still researching.

 

[Ubbe]

I've dropped all DMR frequencies into one favourites list, and by scanning just that one list I get a perfect decode.

There have been some reports that mixing modes can have some issues, and having analog channels scan in succession and then conventional digital separately seems to work better and are also how Unidens database will scan or if you append from it. Some SDS200 owners even gets their analog channels non working and have to power recycle their scanner. There's also a delay when switching bands, especially to and from low-vhf there's a delay in time but the scanner seems to be aware of it and waits for the receiver to settle down, as it seems when looking in the debug file. It's a complicated scanner and are probably impossible for Uniden to find these kind of strange bugs.

/Ubbe

 

[KB8KOJ]

I have and had constant dropouts on the Connect + system I monitor. Mid transmission it looks like it goes back to the control channel. See this both with the SDS100 and SDS200. Signal levels are good. This was always something I had hoped they would address. But haven't seen any other reports of it.

 

[Ubbe]

Mid transmission it looks like it goes back to the control channel.

That's typical when a transmitter on an adjacent channel begins to transmit and the scanner loose sensitivity or gets interfered with a high bit error. If the signal are good then try the attenuator. Then different filter settings and IFX.

/Ubbe