Page 3 of 4

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 31 Dec 2016, 10:07
by Admiral
radiosification wrote:
FakeJake wrote:It does do tier 2, heard someone on UK wide using one via his local repeater. It may still be transmitting on both slots though, I suspect. Needs further testing.
https://www.youtube.com/watch?v=BQkzq0H ... nel=spymsn
As you can see in this video, it transmits on both timeslots therefore it's not a proper tier 2 radio. It isn't even a proper DMR radio to be honest because even tier 1 radios are supposed to transmit in only one slot and not both. It never should have been sold as a DMR radio.
Does that mean that it will also receive both time slots simultaneously on a repeater?

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 31 Dec 2016, 17:55
by radiosification
Admiral wrote: Does that mean that it will also receive both time slots simultaneously on a repeater?
No. It will only receive and transmit to one slot, but because it's incapable of turning the transmitter on and off fast enough, the transmitter just stays on during the other timeslot, jamming it for everyone else.

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 31 Dec 2016, 22:06
by troffasky
radiosification wrote:You guys just inspired me to write this blog post on the topic
http://radiosification.blogspot.com/201 ... avoid.html
This seems like a sensible approach:

http://www.trilogy.org.uk/?page_id=252

Low-cost Chinese hardware running software tweaked and debugged in the UK [I say that not having touched one, of course]. Doesn't seem to extend to the manual though:

https://dl.dropbox.com/s/ldsps0o7g0jm86 ... 0.pdf?dl=0

"To make sure the radio's best performance and prolong its working life, please be acquainted with the following contents, so that you can do well maintenance and clean for the radio"

Is the Kydera DM-880 any good to start with? £89 seems reasonable.

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 31 Dec 2016, 22:27
by Sgfcultrajon
troffasky wrote:
radiosification wrote:You guys just inspired me to write this blog post on the topic
http://radiosification.blogspot.com/201 ... avoid.html
This seems like a sensible approach:

http://www.trilogy.org.uk/?page_id=252

Low-cost Chinese hardware running software tweaked and debugged in the UK [I say that not having touched one, of course]. Doesn't seem to extend to the manual though:

https://dl.dropbox.com/s/ldsps0o7g0jm86 ... 0.pdf?dl=0

"To make sure the radio's best performance and prolong its working life, please be acquainted with the following contents, so that you can do well maintenance and clean for the radio"

Is the Kydera DM-880 any good to start with? £89 seems reasonable.
I'd go for the md-380 or the rt3 like the admiral says

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 31 Dec 2016, 23:15
by paulears
I'm a tad confused here by the the notion it's a hardware issue rather than a firmware curable one. Multiplexing data isn't exactly new, and all this slot 1 and slot 2 business is under software control. If the unit was physically switching between receive and transmit, then that could be a hardware function, but most simple devices that work in time slots, and 2 is pretty simple stuff compared with the timing needed for other devices with far more channels being transmitted sequentially. The critical issue for Tier 1 devices is timing accuracy, and since we started digital, quite a few radios do cause packets to collide with those from other users, and in most cases, updates to the firmware can sort this to a degree. Repeaters have to be able to re-clock the input signals to match the clock of the output, but two radios each set to a different slot, but on the same frequency will still cause packet clashes unless the radios are duplex and can listen, and sync to the output - which they cannot. Lots of claims get made, but two radios on the same frequency, working independently will cause packet loss at the receiver. Do these Baofengs actually completely weak the DMR spec by transmitting in both slots on purpose, or do they simply have rotten timing stability and appear to be in both slots? If timing is a hardware issue, then they're useless - but if the sync stability is a firmware function, then possibly it could be fixed? I may be wrong, but DMR doesn't turn the RF on an off in the traditional sense, as in switching between transmit and receive - it just switches the thing to transmit and then pulses the output. Modern digital devices can run at much, much faster speeds than DMR uses. I'd also have to wonder if Baofeng actually designed a totally new device, and probably use an existing one - again, another clue that this could be fixable.

The video link, despite my poor language skills doesn't say it transmits on both time slots, just that it is received as doing so - that's a vital difference, as poor timing in digital systems often 'overflows' into the next slot - but not every time just in multiple ones as the two sync speeds drift in and out. Until somebody sticks one on a scope and looks at the waveform, we cannot know with certainty. I remember people complaining when analogue FM moved to 12.5KHz spacing, and some makes got slagged off for transmitting on the wrong channels, when they were simply set up with too much deviation, and took out the next channel under some circumstances. The problems with the Baofeng might be a major cock-up in the design, but I suspect it's just loose sync, which could be a code problem and fixable, or worst case, a rubbish component that isn't up to the spec - but would need a mod costing more than the radio, if done here. We will have to wait and see what the technical boys come up with. Somebody must soon stick one on a scope and see if the timeslots are there, or not. My money is on them being there, but with poor timing stability.

The tests they do don't really say much at all. If the baofengs are Tier 1 only then even if the ETSI spec allows for the two time slots, in practice they will interfere as there is no timing possible with every radio having it's own sync?

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 01 Jan 2017, 00:06
by troffasky
paulears wrote:I'm a tad confused here by the the notion it's a hardware issue rather than a firmware curable one
In the case of the MD-398, looking at the specs I would guess a dumb PA has been welded onto an MD-380, and it can't handle the whole TDMA thing. For the Bumfeng, have they just stuffed a DMR "modem" into an FM transceiver, and again, the PA can't switch fast enough?
two radios each set to a different slot, but on the same frequency will still cause packet clashes unless the radios are duplex and can listen, and sync to the output - which they cannot.
This is referred to as "TDMA Direct Mode" in the DMR spec [TS 102 361-1]. It sounds to me like it would be prone to going wrong if one radio misbehaves. [If you want 6.25kHz channel equivalence, just use 6.25kHz channels!]

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 01 Jan 2017, 14:51
by JohnB57
TDMA only works through a repeater or a compatible proxy, such as the Motorola 4000 series, that is able to transmit a universal time synchronization signal. Without the sync signal, mobile radios do not transmit and receive in time-slots even though one is specified in CPS; the function is only active in repeater mode and is wholly under the control of the repeater. There is no point in TDMA for simplex transmissions as, to function at all, TDMA obviously has to rely on a single source of synchronization.

A non-TDMA DMR radio would not be able to sync with the repeater and would therefore be incapable of registering on the system, so there is no possibility of transmitting through the repeater on both slots; the repeater would simply reject the transmission. Input frequency could, obviously, be blocked or jammed, but that's just the same capture effect as with a normal FM signal.

Because synchronization is within the control of the repeater, the concept of "packet clash" on TDMA is only an issue where there is a large distance between the repeater and mobile station and the transmission cannot be received within the slot. In practice, any transmissions that don't comply with the timing constraints are rejected.

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 01 Jan 2017, 15:52
by radiosification
JohnB57 wrote: A non-TDMA DMR radio would not be able to sync with the repeater and would therefore be incapable of registering on the system, so there is no possibility of transmitting through the repeater on both slots; the repeater would simply reject the transmission. Input frequency could, obviously, be blocked or jammed, but that's just the same capture effect as with a normal FM signal.
There are actually people who have used non-TDMA radios on repeaters and they have worked. It's not that they're trying to send data in both slots, they try to use only one slot, but the radio still transmits in the other slot. I suspect it is just transmitting the carrier in the other slot.

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 01 Jan 2017, 16:58
by paulears
Hence why the videos that clearly let you hear the audio difference let you determine crudely that it is a non-Tier I transmission. I wonder if the better radios simply are better at the timing when they 'hear' the repeater sync, and see the radio synced read for TX. I'm still thinking that with solid state TX/RX switching, then software/firmware upgrades should be able to get the sync in order. I wonder if the timer for TX length is also something to do with the sync process - if transmissions are able to last longer between an RX sync up, that would cause clashes. As maximum TX time is a user adjustment, has anyone done any tests to see how long it takes before time alignment drifts outside the window and the data is corrupt? As for using the wrong data format, surely the repeater should reject it - a properly clocked signal is still as continuous signal. A repeater with a carrier on it's receive frequency cannot receive two at the same time - just the strongest with or without data content. The dPMR system wins here. doesn't it with two parallel channels on different frequencies, DPMR can receive two channels at once and software select which one to hear, but surely on a repeater input it is one channel at a time unless sync accuracy allows two different sources to interleave?

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 01 Jan 2017, 17:18
by Admiral
Thanks for the technical stuff, most of it over my head, as much as I would prefer a dualbander for general 'experimentation', I also don't want an $80 paperweight, so personally for now will use two known good single band sets, one for VHF and one for UHF.

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 01 Jan 2017, 19:12
by radiosification
paulears wrote:As for using the wrong data format, surely the repeater should reject it - a properly clocked signal is still as continuous signal. A repeater with a carrier on it's receive frequency cannot receive two at the same time - just the strongest with or without data content.
Wrong data format? What? The issue is timeslot incursion, not data formats.
DMR works by having two timeslots. There should not be any continuous carrier on the input of a DMR repeater. Ever. Even if two different users are transmitting, using both timeslots. There is still a guard time between the slots to allow for some change of distance of the mobile station and some amount of time inaccuracy. Using a radio that transmits in the other timeslot defeats the entire purpose of a TDMA system.
If a radio is transmitting a blank carrier the repeater will still be affected by this if another station is trying to transmit on that timeslot. The received signal will definitely have a higher average bit error rate (BER), if it can even be decoded at all.
paulears wrote: I'm still thinking that with solid state TX/RX switching, then software/firmware upgrades should be able to get the sync in order.
How can you avoid transmitting in the other timeslot when the PA simply is incapable of ramping down fast enough? Sync is not the issue.
paulears wrote:The dPMR system wins here. doesn't it with two parallel channels on different frequencies, DPMR can receive two channels at once and software select which one to hear, but surely on a repeater input it is one channel at a time unless sync accuracy allows two different sources to interleave?
No. A dPMR repeater uses 6.25KHz channels. That's half the bandwidth of the channels DMR uses. dPMR repeaters can only support one user at a time. The terminals for this type of system are cheaper because they don't need to have hardware capable of TDMA. The network cost is greater though, because instead of buying one 12.5KHz channel and one repeater, you have to buy two 6.25KHz channels and two repeaters (and either an extra antenna or other hardware to allow them to share one antenna) There is no interleaving of two signals on dPMR, it is an FDMA system. Not something like CDMA. A single dPMR repeater cannot receive two channels.
dPMR only wins if you need just one channel per site. Also in some cases, like for purely simplex radio communications, dPMR would be a much more appropriate standard to use than DMR.

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 01 Jan 2017, 20:12
by Metradio
If you want to see the effects of a DMR radio transmitting on the two time slots at the same time. Using an MD-380, key up to gain access to the repeater then quickly de-key and re-key the MD-380, you will be transmitting on both slots!!!

Information from Jerry Wanger, president, Connect Systems Inc.

Mike

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 02 Jan 2017, 18:29
by paulears
Take the point on dPMR, but radar with high PRR has been around for a long time, so certainly possible to switch extremely fast nowadays.

Looking back at some of my dPMR info sheets I got from manufacturers, they don't actually explain the two channels very well at all, and I took their comments that there would be two parallel data streams in one 12.5KHz channel to mean that the repeaters did both. Never even occurred to me that one dPMR repeater is a single channel unit. Whata mistaka to maka!


So do all DMR portables sync themselves to incoming RF? This suggests that repeaters need to be transmitting continuously, or at least regularly. So each radio sync up with the incoming data ready for TX in the right time slot. I used the term format to refer to the data packet format - bursts and frames. I'm familiar with the format of data streams. In DMR the guard bands are very small as a percentage of the total time for each packet - not surprising clashes happen.

So is the problem really sloppy timing or something more complicated. Sync lockup could be a firmware tweak, hopefully.

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 02 Jan 2017, 18:38
by stanogs68
thats basic radio stuff ,ill let you know the answer :)

Re: CHEAPEST way to go digital HELP NEEDED

Posted: 03 Jan 2017, 15:38
by troffasky
paulears wrote:So do all DMR portables sync themselves to incoming RF?
If a receiver's timing is poor enough that it can't distinguish between slots, it's never going to decode anything successfully.

paulears wrote:This suggests that repeaters need to be transmitting continuously, or at least regularly.
That's my understanding of it. For TDMA direct mode obviously you don't want an HT transmitting continuously:
http://cwh050.blogspot.co.uk/2012/12/dcdm.html wrote: The Channel Timing Leader periodically announces the Channel Timeslot Structure using special Beacons (600ms transmission every 4,5 minutes)
4.5 minutes is presumably the tradeoff between the accuracy of the clock in HTs and not having the timing leader's battery empty after an hour.
paulears wrote:So is the problem really sloppy timing or something more complicated.
In my opinion, it's less complicated than that - as radiosification said, crappy PA that can't switch fast enough. Almost certainly the issue with the MD-398 because looking at the specs it looks like an MD-390 that's had a PA stuck on to it.