Call waiting and 3-way calling

[Kjaslow11]

I'm using an SPA-112.  I have used this unit for a few years, and while this problem has always existed, I usually run out of energy trying to figure this one out...

I am unable to make conference calls, nor am I able to receive call-waiting calls.  I can hear the call-waiting tone, and I've confirmed with my VoIP provider that my account supports conference calling and call waiting.  They are convinced that my ATA is the problem.

I have "CW Setting" set to 'Yes.'  My guess is that the ATA is not responding to the phone's flash button during a call.  I cannot get a dial tone when on a call and after pressing the flash button.  Does anyone have any idea what setting I should look for to understand why the flash button does not give me a dial tone to initiate a 3-way call?

I've also tried turning on debug logging, but the log is undecipherable to me.  I tried making a call, then opening and clearing the log, and then attempting to use the flash button to initiate a 3-way call.  The log gets the following two messages:

Jan 25 08:39:26 VoIP Phone user.notice sprvoip: cmd - 20, response 42
Jan 25 08:39:26 VoIP Phone user.notice sprvoip: cmd - 20, response 43

Does that provide any useful information?  Thanks for your advice.

[R0g22]

Just for the sake of it, have you tried with a different phone ? Can you grab a log from SPA-112 for a fail scenario ? Set the priority to Debugging.
https://www.cisco.com/c/en/us/td/docs/voice_ip_comm/csbpvga/spa100-200/admin_guide_SPA100/spa100_ag/administration.html#61544

[Kjaslow11]

Thanks.  In the attached log, I dialed into a conferencing service.  Then at 9:20:30 I attempted to initiate the 3-way call.  I watched the log viewer as I did this.  The log was not appended when I pressed the flash button, but did append with each number that I dialed.  I attempted the 3-way call twice, and then hung up.  Do you spot anything?  Thx.

[Kjaslow11 · Accepted solution]

In case it helps anyone else, this ended up being related to my phone hardware.  I bought the phone while living outside the US.  While the phone works in every other way, the flash signal is not recognized by my ATA and there's no way to modify the duration of the flash generated by the phone.  Ended up replacing both the phone and ATA with a SIP phone.

[David30]

---

@Kjaslow11 wrote:

I'm using an SPA-112.  I have used this unit for a few years, and while this problem has always existed, I usually run out of energy trying to figure this one out...

I am unable to make conference calls, nor am I able to receive call-waiting calls.  [...]   My guess is that the ATA is not responding to the phone's flash button during a call.

---

I know the "running out of energy" feeling, and your guess is quite correct.

For the benefit of others with this problem, the recommended & default values for "Hook Flash Timer" (min & max) are incorrect, as noted in various places, e.g. https://support.8x8.com/

The values 0.02 (1.2 sec) & 0.2 (12 sec) for minimum & maximum, respectively, work fine.  It seems the ATA assumes the units are minutes and multiplies the specified values by 60, also contrary to the documentation.  Furthermore, the minimum value apparently has to work out as >1 second, and the ATA doesn't enforce the documented ranges (maybe as a bug-fix!).

It would be nice if Cisco updated their Admin Guide as this is thoroughly confusing.

DL.

[Dan Lukes]

---

the recommended & default values for "Hook Flash Timer" (min & max) are incorrect

There's nothing like "standard plain old analog telephone lines interface".  Line parameters including hook flash times are per-country and sometime even per-operator specific. Side note - most of European public telephony knows no flash at all, thus flash is not part of national standards and is pure implementation specific signal. Not surprisingly, default values of parameters can't comply with all environments. It's why they are configurable.

In short, configured Hook Flash Timer ranges must follow specification of the flash signal generated by phone connected to SPA112.

The values 0.02 (1.2 sec) & 0.2 (12 sec) for minimum & maximum, respectively, work fine.  It seems the ATA assumes the units are minutes and multiplies the specified values by 60, also contrary to the documentation. 

Doesn't sounds plausible.

Flash hook can be described as quickly depressing and releasing the plunger or the actual handset-cradle to create a signal indicating a change in the current telephone session. It's the same kind of signal causing hang-up. A flash hook is detected when the hook switch is pressed for a shorter time than would be required to be interpreted as a hang-up.Even this is not standard, but it's most common implementation of hook flash signal I know.

In short - on-hook with length beyond min hook flash timer are considered disturbance and ignored. On-hook with length within min-max hook flash timer are considered hook flash.  On-hook with length above max hook flash timer are considered hang-up.

Min/Max values either can't mean minutes or your values (0.02 = 1.2s, 0.2 = 12s) are so high to work as hook flash. But 0.02 = 20ms and 0.2 = 200ms sounds familiar to me.

Hook flash troubleshooting have two parts. First, hook signal generated by phone needs to be recognized by SPA112. Use Debug and syslog Messages from SPA1x2 to verify the flash is recognized by SPA112. Second, (recognized) flash signal affects established SIP session. It may or may not be allowed/properly recognized by upstream call control device (VoIP provider). Use debug and syslog messages and captured SIP packets to troubleshoot this.

[David30]

---

@Dan Lukes wrote:

---

the recommended & default values for "Hook Flash Timer" (min & max) are incorrect

[...]  Min/Max values either can't mean minutes or your values (0.02 = 1.2s, 0.2 = 12s) are so high to work as hook flash. But 0.02 = 20ms and 0.2 = 200ms sounds familiar to me.

---

The Admin Guide states:

for the minimum:  "Range: 0.1–0.4 seconds.  Default setting: 0.1"

for the maximum: "Range: 0.4–1.6 seconds.  Default setting: 0.9"
However The values recommended by 8x8 Support at https://support.8x8.com/equipment-devices/phones/cisco/calls-dropping-while-using-cisco-linksys-spa122-ata-adapter are 0.02 (min) and 0.9 (max).

I have to admit I can't remember exactly how I concluded the ATA was treating these values as minutes, so I withdraw that.

However, there's obviously a problem.  For example, with my values of 0.02 / 0.2 the "flash" function works fine when the key is held for much longer than 0.2 seconds.  It doesn't work with values 0.005 / 0.5.

[Dan Lukes]

I'm not disputing the 8x8 recommendation - but it can't be generic. Unless conditions are mentioned, it have limited usability only.

The "original flash" has been fired manually by hand pushing the cradle for a moment. It produce "flash signal" of various length. The defaults fit it most of time. Some phones have dedicated flash button - while it may work like "second cradle" (e.g. length of signal depends on the time the button is depressed), it produce fixed length of hook signal most of time regardless of the length of press. But other behavior is possible as well. Those phones may produce flash signal shorter than 100ms and SPA112 defaults doesn't fit.

Unless we know specification of the flash signal produced by your phone, we can just blindly guess the issue cause. The major rule is - SPA112 configuration needs to follow phone's specification. Generic advice like the one from 8x8 may or may not apply for the particular case (phone).

---

---

For example, with my values of 0.02 / 0.2 the "flash" function works fine when the key is held for much longer than 0.2 seconds.  It doesn't work with values 0.005 / 0.5.

---

May be "just tap" on flash button is not considered "press" by your phone and no flash signal is generated at all. I'm just blindly guessing here - as I know no behavior of flash button on your particular phone.

Please note - I'm not disputing your solution - it just works for you.

I'm trying the explain hook flash background to help others - your solution will help to someone with similar analog phone, while for someone with dissimilar one is rather useless. The same apply for 8x8 advice. As neither you not 8x8 disclosed the phone we are speaking of,  it's no way to distinguish those cases.

[David30]

---

@Dan Lukes wrote:

The "original flash" has been fired manually by hand pushing the cradle for a moment.  [...]  Some phones have dedicated flash button - while it may work like "second cradle" (e.g. length of signal depends on the time the button is depressed), it produce fixed length of hook signal most of time regardless of the length of press.

---

That's a good point.  My 'phone in this case is a Uniden FP1200 and I suspect the flash key produces a fixed-length pulse of around half a second, judging by the interruption to dial tone when it's operated off-hook (I think I tested hook-flash operation using the cradle switch).  But 0.5 sec is still greater than 0.2 sec so I might increase the maximum time just in case 0.2 sec. is marginal.

I think the term "hook flash" originated in the days of manual switchboards and switchboard operators.  Jiggling the hook switch flashed a light on the operator's board to attract their (usually her) attention, and the operators tended to know everything which was going on in the area!

[Dan Lukes]

You hit. Well, I have no personal experience with such kind of switchboard - the oldest switchboard I has been operator of has been relay based - no semiconductors, no vacuum tubes, just relays, control lamps ant mechanical ring ...

[David30]

Thanks for that discussion Dan, I now have 'flash' working well and can switch back & forth between conversations just by operating the flash button.

However I have no idea how to set up a 3-way conference call where the audio is mixed locally by the SPA112 (not externally); would you please point me in the right direction?

I notice there's a parameter "conference activation code" parameter but it's blank by default, which suggests it's optional.

On the subject of telecoms history, I think manual exchanges were still in use in some small, remote towns in Australia until the 60's.  I can remember seeing a step-by-step exchange in operation when I was a boy (I think) and they were an impressive sight, but partly electronic crossbar exchanges were also becoming common by then.

[Dan Lukes]

Standard sequence is: Flash (fist call put on hold), you got dial tone, dial number with # on end, wait until second call become established, flash. 3-WAY done. But upstream support is still required. Voice providers may refuse to establish second call while the first one is still active - and some real providers are known to enforce such king of limit.

[David30]

Thanks, Dan, it's all working beautifully in a few installations!

DL