"TX Out" RCA active during bootup process of 6300 - resolves after starting SmartSDR

Howard -W6HDG

I had previously posted a thread about "desense" of my Yaesu FTdx3000 during the bootup process of the Flex 6300. See https://community.flexradio.com/flexradio/topics/6300_before_starting_smartsdr_causing_desense_of_ne...
It turns out my KPA500 amp is quietly tripping during Flex 6300 bootup and my antenna is being disconnected.  The KPA500 has no sign of being in transmit mode since it is not being driven.  I just happened to pull the TX RCA cable and voila, the FTdx3000 comes back to life.

Once SmartSDR starts or the 6300 is shutdown, the "TX Out" RCA is inactive again.  Is this just my radio or a general 6300 issue?

Steve-N5AC

This is likely an issue with all FLEX-6300s, but it is a software issue and can be remedied.  I've entered this in our defect system as #356 for resolution.  Thanks for the report, Howard!

Steve-N5AC

Howard, a few questions: 1) are you using the TX Out in an active high or active low configuration?  2) Do you have anything else connected to the TX Out line or does it go directly from the FLEX-6300 to the KPA500?

Howard -W6HDG

Steve,
I have the RCA 75 ohm cable provided by Elecraft going from the TX RCA on the back of the 6300 to the PTT IN on the amp.  The amp is expecting the signal to go low (ground) in order to trip the amp.  Isn't that what the TX jack on the 6300 doing?   Is there a high vs low setting for the Flex?

Do you notice any voltage changes on the TX line if you measure during 6300 boot/start and before SmartSDR is started?  The change may be subtle but it is apparently enough to confuse the KPA500.

Steve-N5AC

The device we use for the TX Out is a TPIC6B595, an industrial high-voltage switch.  We have tied the !G line (pin 9) to GND and on power-up the device should float.  I'd probably recommend a pull-up on the line that is at least 2.5k to flow 5mA.  While I've not measured on a 6300, this is the same mechanism we use on most of our radios and I suspect I would have heard if this is a common problem (but not always).

I was guessing that you were going to say that you were expecting it to go low on power up and only go high on key down.  There is a time period while the radio is booting that this would not work and I figured this was the issue.  Now I'm a little more suspicious of your pull-up.  What value are you using or is it in the amp?  I loaned out my 6300 and should have it back in a day or two and I'll verify with the hardware if you don't discover a resolution on your end before that.

Howard -W6HDG

Steve,
I didn't do any current measurements, but the PTT out of the Flex measures about 50 mv when SmartSDR is running in receive.  In transmit, this goes low and drops to about 1 mv.  

On cold starting the Flex 6300, and before starting SmartSDR, there is about 1.5 - 2.0 mv at the PTT out. This jumps to 50 after starting SmartSDR.  The KPA500 has a very sensitive PTT since it is silent diode switched for QSK.  That is why I missed the fact that the amp was switching in and disconnecting my antenna to my other radio, a Yaesu FTdx3000.  Others may not notice this issue if they don't have a dual rig setup with a coax switch that can select the radio attached to the amp.  My Yaesu FTdx3000 is connected by a separate data cable with a PTT connection to the KPA500 so it does not share the PTT amp phono per se (but may be in parallel and having some effect).  But the bottom line is that all works well with both radios once SmartSDR is brought up.

Is there any way to set the PTT high on rig bootup as if SmartSDR was running?

John n0snx

Same problem here Steve... 6300 into an Ameritron al800

Steve-N5AC

Guys I believe a pull-up will solve your problem.

Stan VA7NF

In a case like this there are two pieces of equipment with standard connectors joined by a wire.  The user should not need to supply a "pull-up" signal. 

The question is: Is there a standard on which machine (sender or receiver of control information should ensure line levels during initialization; or both.

The amp, during power-up should pull up the signal line to its default level; this ensures the line is biased before it checks status.

The Flex, during power-up should also pull up the line from its float level; this ensures it doesn't give false signals.

Sounds like the problem is a result of neither end biasing the line.  Hardware change?

Stan

Howard -W6HDG

The KPA500 amp behaves as it should and I can't find fault with it.  

If this issue does require a hardware change to the 6300, also fix the misplaced TX (PTT out) RCA jack which should be paired with the ALC jack.  I'd bet most people make the common mistake discussed here: See https://community.flexradio.com/flexradio/topics/flex_6300_to_hl_1_5kfx_connection

Gerald-K5SDR

Stan, we use an open drain output so that you can drive many types of devices from the same port.  That means with the proper external terminations, you can drive 12V relays, 24V relays, 5V TTL, 3.3V LVCMOS, etc.  A pull up would eliminate this versatility.  Standard engineering practice puts the pull up at the digital input.  This is what we provide on all of our digital inputs such as PTT and TX REQ.

Paul Christensen, W9AC

The open drain is treated much like an open-collector and to Gerald's point, it should definitely not have an internal pull-up resistor when so many user-defined external voltage and interfacing variables are involved. 

I think Steve indicated that the DMOS device has a master inhibit data line (/G).  As long as the software takes advantage of that -- and as long as the switching DMOS gate is always terminated and never floating, it should not cause erratic operation with the user's equipment.

If having an open drain line is causing a problem, the problem is owned by the user and a pull-up resistor is the obvious solution.  For those who are not comfortable with pull-up resistor technique, then consider purchasing a key line buffer/interface.

Paul, W9AC 

  

Howard -W6HDG

The problem is owned by the user? Absolutely, as long as Flex Radio's target users are digital engineers. I'm no appliance operator, but if Flex wants to sell ham radios to hams, make your radio work like all other popular transceivers. New paradigms are great - but not for PTT out! How hard is it to ground and unground a line?

Paul Christensen, W9AC

It looks like Flex is investigating.  See Steve's comment above about entering the issue into their "defect system."  I'm not sure if that was entered before or after most of the discussion here.

If we're going to follow the lead of other manufacturers, that means using relay switching in place of solid-state FET switching.   In the few transceivers that use FET amp key line switching (e.g., K3 and Icom 7800), they have issues with key line activation during boot-up.  I know my K3 still does.  I just ensure the amp is not brought on-line until the K3 finishes booting.

If we're expecting Flex's key line to work with any amp, including those made in the 1950s that use 120 VAC key lines, then they would need to go back to relay switching (K-logic), or use balanced MOSFET outputs like the type shown on my QRZ.com page.  That's a bit much to ask -- and in the case of using old amps, its reasonable for the end-user to either build or purchase his/her own buffer/interface.

As long as Flex is inhibiting the gate function on the DMOS switch, then grounding and un-grounding a line is as simple as you suggest.  Hopefully, Flex can confirm if the inhibit function is working properly in SSDR.

Paul, W9AC

  

Gerald-K5SDR

Howard, I need to apologize.  I went back and read your posts more carefully and it prompted me to go make some measurements with an external pull up.  It turns out that we are not initializing the output until SmartSDR is loaded.  I will as Steve to reinstate the defect so that they can correct this in software.

My statement is still correct that pull ups do not belong at the open drain output inside the radio.  Pull ups for digital inputs always belong on the input side of a keying circuit.  That would eliminate the ability for the user to define the voltage of the external keying circuit (3.3V, 5V, 12V, 24V, etc.).  

I am concerned about the fact that you see only 50 mV on the circuit when the amp is not keyed.  Even though it is not then in transmit that indicates they may not have a pull up on the input which leave its input floating.

Howard -W6HDG

Hi Gerald,
Thanks very much for confirming the issue.  I was not all that clear in my subject line.  The 50mv vs 1mv was the open voltage as tested at the TX jack of the Flex during Receive and TX respectively.  Probably a meaningless value except to show that "TX type conditions" existed during the bootup process and before SSDR was invoked.  

The KPA500 is a 5v keying circuit and functions as expected.  I will look forward to testing the fix.  

Gerald-K5SDR

If the KPA500 keying input is at 5V with SSDR loaded and in RX, then you should be OK after we get the software reset in the code.

Steve-N5AC

OK we have found and fixed this issue and it will be in v1.3.  Without getting into a lot of technical detail, the issue was some relays flipping that did not need to flip during startup.  They are now not being flipped and the result, in addition to fixing the core issue that TX1 goes low at startup, is that you will no longer hear a "ca - CLICK" sound at startup.  Now you hear one relatively quiet click. Thanks for pointing out the issue.

Steve

Howard -W6HDG

Steve, may I ask if this was a 6000 series issue or just a 6300 issue?

Steve-N5AC

The problem existed in the 6500/6700 also, but it is benign -- it does not affect the TX1 output. All radios will receive the update.