How do you know what Slice is assigned what SCU??

K3SF

Basic question is how do you know what slice is assigned to what SCU in a multi-SCU flex radio ( aka 6600 and 6700's series).

I have not be able to find any indicator on a slice to see what SCU is generating its input.

There are controls that affect the SCU only like RF preamp/Attenutor settings.
But they also directly affect associated slices>

One of the possible configurations of the 6600m gets me thinking.

The config is one panafall with four slices all on 20m with only 500khz showing in the panafal.

One slice for cw set with daxiq with cws, second slice on for FT8 with dax audio
third slice for rtty with daxiq for rttyskimmer and fourth slice up in ssb portion
all on ANT1.


There is no way of knowing which slice is generated by which SCU
so now if i wish to attenuate the FT8 slice what other slice is affected.

A similar thing would happen if one of the four slices is assigned to ANT2.
but
here i would expect the other slice on the same SCU i would now see its receive antenna change to ANT2.

I know of NO indicator of slice to SCU..
I have read and re-read the manuals and found no answer there.


I believe the guys with the 6700 have same issue.

i do know when you set one slice to ant1
and
the other slice to ant2 they each have separate SCU...
BUT
that is the only indicator.

Paul K3SF

Marc Lalonde

the "control" affect Slice and ANT  ,the SCU is totally hidden to user

so if you add pre-amp to ANT1 all slice on ANT1 got pre-amp 
but you never knot if ANT1 is SCU1  or 2  and in reality  ,you not need to knot until you have issue whit one of the SCU 

Mike va3mw

However, I just tested this by opening 2 slices on 80M on the antenna assigned to the same SCU.  This is on a 6600.

As I varied the gain, you can see that is also varied on other panadapters on the same SCU.

From the manual:

6.2.1 Spectral Capture Unit (SCU)

Because the notion of a receiver is firmly established in both the amateur community and possibly the rest of the world, it seemed inappropriate to describe the functionality of a wide-band sampling system simply as a “receiver”.

Instead, in the FLEX-6000 world there are one or more “Spectral Capture Units”, or SCUs, that are responsible for the collection of wide-band data from the RF spectrum. The SCU components are: an antenna input, an optional set of receive preselectors, and an analog-to-digital converter (ADC).

Each SCU in the radio system can be connected to only a single antenna at a time, but due to the sampling architecture it may support any number of receivers and spectrum displays. The SCU enables listening to multiple bands at the same time on the same antenna.

A hardware platform with multiple SCUs such as the FLEX-6600M allows for monitoring multiple bands on different antennas or the ability to perform more complex noise mitigation techniques that are available in multi-antenna systems. 



Mike va3mw

Ted_Spiegel_NX6C

Paul,
   Your going to love the responses you get!
Here is mine:


In your example you mentioned one Panafall with 4 slices.  For 20M only one SCU needs to sample the RF and provide you with all the data for each of the slices which then provide you with your desired modulation scheme from each slice.

In this example you will have a spare SCU to possible be ready for another antenna system.  

What is your desire?

2M transverter, 6M skip opening, 15M DX ?? If you’ve got the antennas and transverter ready to go you could just use a memory software ’Load’ for one of those possibilities and choose one of your slices to switch to your choice.  The memory software operation ‘load’ will switch your frequency, mode, antenna and many more parameters.

The other 3 slices will continue on 20M with your 20M antenna system on ANT1.

The newly changed slice will then use your ANT2 RF system.

Simply put, yes, the only indicator of the SCU is ANT1 or ANT2.  The amazing thing that makes the FlexRadio so much fun is in the ‘Flexibility’ of the station.  

There are Lots of things the operator can manage and change with memories, profiles and the two SCUs with their totally independent RF measuring capabilities.

I love the FlexRadio concept.  Can you tell?

Ted 

NX6C

K3SF

hi mike

yep that is exactly what i am talking about, hat making an adjustment on one slice affect another...

now with a 6600 if you put 8 slices on 80m ant all on ANT1, would you know what slices would be affected when you adjusted the RF gain of just one slice.  

Cause with 8 actives slices i think you would be using both SCU's 4 on each SCU.

i setup my initial example bases on 6600 have 2 SCU but only 4 slices.
and
assume two slices per SCU..

i have on order a 6600M and been doing brain games while waiting ;-)

Paul K3SF

Mike va3mw

Hi Paul

There is only 4 slices on a 6600.  Based on what I just tested, all Slices on that antenna port would be affected.  If you had 4 slices on your 80M inverted V, they would all be on that SCU and affected by the common gain/attenuator.

Mike


Ted_Spiegel_NX6C

The SCUs are not assigned arbitrarily.
The two SCUs are the ANT 1 and ANT2.  

    Each ANT provides an independent RF  gain for all Slices associated with it.  If you change the RF gain on ANT1 all the slices using ANT1 will have their RF gain changed.  
     
  You use the SCU RF 'gain' to maximize the quality of that particular antenna RF system. 

    If you need to change an individual slices level you can use the DAX  gain control of the demodulated signals.  
    The DAX control panel has sliders for each slice.  It's like having a recording studio mixer panel for all your slices.   There is signal indicator for each of the slices output that immediately show the effect of your DAX channel slider.  In the DAX panel all slices are independently adjustable.

    For some modes you also have the AGC-T control of each individual slice to pull your desired signal out of the noise.

In summary;
RF gain is associated with antennas.  With the 6600 you'll have 2 independent RF gain adjustments associated with ANT1 and ANT2.

Signal gain is associated with Slices. DAX and AGC-T are associated with the slices.  With the 6600 you'll have 4 independent slices each with their own DAX and AGC-T adjustments.


When you get your hands on the controls you'll have total power over everything you need.   Your goal will be choosing the best way to get that elusive signal on each of them with all that power.

You will have 4 Independent High Performance Receivers.  Lots to have fun with!

Ted
NX6C
   

K3SF

hi Ted

Yep...i am anxiously awaiting a 6600M. I had a 6300 which i traded in and loved how it performed so had no doubts about upgrading.

When i read thru the manual i got the impression that the SCU's were not tied to an Antenna port. But you are saying they are and such that SCU1 would be ANT1 and SCU2 would be ANT2.

The Manual did say RXA is SCU1 and RXB is SCU2 but i didnt see the ANT1 and ANT2 being fixed in a similar manner.

Thanks for reminding me about DAX level controls. I had forgotten about them cause i was so focused on how to manage the SCU utilization.

Guess i will have even more fun when i get may hands on one.

I do like to experiment and find the outer-limits and array of configurations.

"see" you on the bands

Paul K3SF

Al K0VM

Another way to think of it... 
On the 6600 or 6700, If a slice is connected to ANT1, RXA or XVTRA for receive it is using SCU0,  
If a slice connected to ANT2, RXB or XVTRB for receive its using SCU1, 
 All slices using SCU0 share the same antenna selection, PreAmp control setting and Band/Wide filter.,  
All slices using SCU1 share the same antenna selection, PreAmp control setting and Band/Wide filter., 
 If all slices for an SCU are on the same band, the filter for that band is used. otherwise the Wide filter is used..
AL, K0VM

Tim - W4TME

In the radio, an SCU is connected to one and only one RX antenna.

The SCU digitizes the RF and from that connected RX antenna generates the foundational component of the panadapter(s).  All panadapters that use the same RX antenna are connected to the same SCU

Slices are resources created within the context of the panadatpter - meaning a slice is not connected directly to an SCU, it is associated with a panadapter which is in turn associated with an SCU that is defined by which RX antenna is used.

SCU resources are allocated dynamically when connected to an antenna with one exception.  If you select RX-A for the RX antenna, it will always connect to SCU1 and RX-B will always connect to SCU2.  VXTR (A/B), ANT-1 and ANT-2 can be assigned to whatever SCU the software determines; first come, first served.

There is no way for the user to know which SCU is dynamically allocated to a set of panadapters unless your RX antenna is RA-A or RX-B.  And you really don't need to know as you can tell which panadapters are on the same SCU by looking at the RX antenna for those panadapters.  

If you change the SCU RF preamp setting, all panadapters defined on that SCU will change accordingly.

wb7ond

My help question led me to here, the community search works very good, but, I would like to suggest that the above information be encapsulated and included in the software manual.  I feel that I have to "connect the dots" rather than read the "plain-text" information in the manual to come to the conclusions explained above.  The antenna 1/2 is the key to which scu is being used.  I recently upgraded from a single scu to a dual scu, and my first question was "how do i enable that second scu".  I tried to find the answer by scouring the manual, but, only by experimenting with the software did I figure out that the antenna selection was the answer.  The manuals definitely do provide valuable information and are well laid out. 

wb7ond

K1DBO

The diagram at the wiki seems to address some of this



While it doesnt address the orginal question, it does appear to contradict some of the discussion.  It looks like RX-B goes only to SCU1 and RX-A can feed either SCU0 or SCU1 via the splitter.  But it looks like ANT1, ANT2, and XVTR can feed either SCU0 BPF or SCU1 BPF.  Am I reading this right?

This has me wondering if the API provides sufficient information to show the state of the 13 relays in the diagram?  I'd really enjoy an active view of this diagram to pop up as needed.  

Tim - W4TME

While it doesn't address the original question, it does appear to contradict some of the discussion.  It looks like RX-B goes only to SCU1 and RX-A can feed either SCU0 or SCU1 via the splitter.  But it looks like ANT1, ANT2, and XVTR can feed either SCU0 BPF or SCU1 BPF.  Am I reading this right?

Mostly correct.  RX-A is always connected to SCU0

rfoust

Probably a dumb question but what's a BPF?  (edit: band pass filter?)

Tim - W4TME

BPF = Band Pass FIlter (preselector)

Eric-KE5DTO

ANT1 is not necessarily tied to SCU0.  It can actually be routed to SCU0 or SCU1.  Same with ANT2.  This is necessary to enable the antenna selection flexibility the FLEX-6600/6700 radios allow.

RX A and RX B are indeed tied to a single SCU.

The software was designed so that you do not have to care which SCU a particular resource (Slice, Panadapter, etc) is using.  It should be sufficient to just note which RX Antenna is in use for that resource and that should be all you need to know.

Will the SCUs be swapped if you open another Panadapter and change the antenna?  Maybe.  Maybe not.  We take care of those decisions and ensure optimal operation so that you can just worry about which antenna you want connected to those resources.  This is how we have designed SmartSDR -- to do whatever we can to take some really complicated stuff and present it as simply and intuitively as possible.  Isn't it cool? 

K3SF

hi eric

i am a little disconcerted by you statement

Will the SCUs be swapped if you open another Panadapter and change the antenna?  Maybe.  Maybe not.  We take care of those decisions and ensure optimal operation so that you can just worry about which antenna you want connected to those resources.  This is how we have designed SmartSDR -- to do whatever we can to take some really complicated stuff and present it as simply and intuitively as possible.  Isn't it cool? 

It is the maybe ..maybe not that bothers me...i like consistent behaviour.

As explained earlier in this thread by TIM the selection of ANT also selects a set of ANT resources (pre-amp and bpf) and became locked to an SCU  once an ANT is  selected and now you are saying they are not way and that it is unknown to the the user until something doesnt work as expected...

The diagram provide some insight with 13x13 = 169 combinations possible
and the s/w selects an operational subset.

i agree K1BDO, would be nice to know what was selected if only thru and an API
and
i think you most likely have a state table that was encoded to make appropriate switch settings decisions.

i have an 6600M soon to arrive
so thinking with 4 slice a possible scenario for my antenna farm is
   Slice 1 on ANT1
   Slice 2 on ANT2
   Slice 3 on RXA
   Slice 4 on RXB

How would pre-amp and bpf be assigned to each slice and which SCU is assigned to a slice?

or
is this an invalid configuration?

i play these brain games a lot...it is a carry over from my previous life



Paul K3SF

Tim - W4TME

i have an 6600M soon to arrive so thinking with 4 slice a possible scenario for my antenna farm is

   Slice 1 on ANT1
   Slice 2 on ANT2
   Slice 3 on RXA
   Slice 4 on RXB

This is not possible because there are only 2 SCUs to digitize RF from an antenna port.  2 SCUs = 2 antenna ports available for use simultaneously 

And antennas ports are not bound to slices, they are bound to panadapter which is a foundational component where slices are created.  If a pandapter has the RX antenna = ANT-1, all slices created in that panadapter will have the RX antenna as ANT-1.  If you change one to say ANT-2, the panadapter's RX antenna changes to ANT-2 along with all of the other slices in that panadapter.

The antenna switching logic in SmartSDR is very complex because it has to deal with all of the possible use cases.  The example above is just one use case, but add a second panadapter with an RX antenna on a different port/SCU, and the logic gets more complex.

We realized this very early on in the development of the radios and is why the construct of an SCU is obfuscated because, from a pragmatic standpoint, you really don't care what SCU is associated to what antenna port; you just want to configure the slices and have it work.

K1DBO

Paul, 

I'm sure the behavior is consistent, the problem is that it's difficult to describe and also possible that it would change between software release (part of the coolness Eric mentioned). 

It's difficult to describe where the next slice will end up because it depends heavily on the current configuration.  From the diagram you can imagine having say four slices all on SCU0.  If you switch the antenna on one of those slices from ANT1 to ANT 2 it's necessary to move the slice to SCU1.  But if you switch the antenna on that slice back to ANT1, it's not necessary to move the slice back... though the software could decide to based on some optimization or another.  Instead of moving a slice, consider creating a new one.  

Now imagine you have eight slices open spread across the both SCU0 and SCU1.  Pick one of them and switch it from ANT1 to RX-B.  That move would require the slice to sit on SCU1.  But where is now?  Who knows? But we do know where it needs to end up!  And, if there is an advantage to evenly spreading the slices over both SCUs an optimization might take place at that time... or could.  Now, for fun, think about the BPFs, what's their current state and what is the frequency associated with each of the slices.  Now where would you like your new slice to appear?

While I would really enjoy designing that algorithm, I would hate to have to explain it to anyone And, I would certainly want to leave my options open for future improvements.

Steve-N5AC

Lots of good comments here and most are accurate.  As Eric mentioned, the number of SCUs in the radio determines how many antennas you can listen to at once.  The 6300, 6400 and 6500 have one SCU and the 6600 and 6700 have two.  Because the preamp and attenuators in all radios are part of the SCU, anything that uses the SCU "inherits" those settings.  Each panadapter, waterfall and slice are attached to a specific SCU.  If you open a panadapter and put it on ANT1, you really don't know which SCU you are on because ANT1 & 2 can be on either SCU, but it doesn't matter.  What matters is that the settings you make in the panadapter are reflected in the panadapter.  

Also any slice you create in a panadapter will inherit the SCU and will be on the same SCU -- but diversity slices will necessarily be on the other SCU.  In the 6300, 6500 and 6700, the XVTR port could go to either SCU.  But in the 6400 and 6600 XVTA and XVTB are tied to their respective SCUs (just as RXA and RXB are).  This does impose some rules -- for example, in the 6600(M), you cannot listen to XVTA and RXA at the same time nor XVTB and RXB, but most any other combination is possible.  As a result, the radio needs some control over the SCU.  For example, if you select ANT1 for the first panadapter, it will generally assign SCU1.  If you then add another panadapter and select RXA (on a 6600 or 6700) the radio will "swap" the SCU on ANT1 and move it to SCU2 so that RXA can get SCU1 (since that's the only place it can go).  When it does this, all the SCU settings are swapped also -- all slices, panadapters, waterfalls, preamp settings, etc. are moved.

We considered all this carefully and felt like showing you the SCU really didn't buy a lot (and probably adds confusion).  The only time we (at FlexRadio) really care is if we are troubleshooting a problem.  Without opening a radio, if we think there's a preamp problem we could look on RXA then RXB to see if it follows the SCU.  By selecting a "bound" antenna port, we know which SCU we're using.  

Finally, for the geeks in the crowd, it's maybe worth mentioning that the rules that govern the SCU for any given antenna combination are interpreted in a Mealy state machine with about 50 states in a 6600 and about 80 in a 6700.  The state machine understands where you were last (previous state) and what you have just expressed as a preference (new antenna selection) and does its best to get you to either your exact selection or something as close to what you have selected as possible, keeping in mind the hardware must end up in a state that actually works.  To achieve this, we invented our own methodology for expressing the states and the possible transitions.  The actual state machine logic is generated from our language and is many thousands of lines of code.  We went this route because we felt that some of the rules in radios we had used in the past were too rigid.  We wanted to make the radio capable of doing as much as it could in hardware, even if we couldn't think of a real reason to do that on day 1 -- we figured some of you might find that reason and want the hardware in a particular state.  Our name has Flex in it; we figure we should be flexible ;-)

ctate243

Ouch.  I thought I was a geek but I think Ill stick to Star Trek. ;-).  and now.. to start reading and googling.

John WA7UAR

Great explaination Steve! I looked terms up in Wikipedia and learned something new. Flex programmers really are among the most cutting edge developers. R-E-S-P-E-C-T ✊

wb7ond

Thanks Steve, I printed that and put it in my SmartSdr Software manual..  Good info.

David Salomon

Speaking of antennas and diversity, can diversity be done with one of the RX/TX antennas and a receive only antenna?

Steve-N5AC

Yes, absolutely

David Salomon

Thanks Steve.

David Orman

Yes, it works. Not fun to adjust with just the front panel controls on a M model, but can be done by pulling open the menu from the top and mucking with the audio levels there.