This old (Bluetooth connected) Phone: Part 1

I inherited an old Western Electric candlestick phone from a relative, who in a previous life, converted to a nice decorative little table lamp.   I wanted to restore some of the intentional functionality of this old device  but with a modern twist. When I am at home in my office/mad scientist’s lair, I often miss calls on my cell phone because, well, I have it on silence or vibrate. I thought it would be fun to sort of “Steam-punk”  this old phone to have Bluetooth on it and have an old time RIIINNNNNG coming from an incoming call.

 

My first step was to test the Receiver ( the cone held up to the ear) with a  signal generator. I disassembled the base and removed the guts from the candlestick base. Using a signal generator, I was surprised to hear this old ear piece still worked as it did.

The next thing to test was the transmitter piece.  Unfortunately, this part had seen better days. I couldn’t pick up  a sound out of it and I feared the coil and diaphragm were frozen.  No worries, I will just use a modern microphone in its place hidden behind the mouth piece.

There was no bell in the base so that was another part I will need to get and since this style phone typically had a remote bell box, I will have to get creative on how I want to implement this.

For testing and  development purposes, the Purpletooth Jamboree development board makes for a nice break out to try microphones, speaker attachment and GPIOs. It’s based on the BC127  module from Blue Creation. The Bluetooth protocol native to this allows for a Hands Free Protocol ( HFP) as well as an audio device.

 

I was able to connect the earpiece to this development board and pair it with my phone.  I played a song from my phone and I was pleasantly surprised listening to some Little Feat being played through the ear piece.

Next step will be to get the microphone part working so I can actually have a conversation on this old phone.

Receiver Adjacent Channel Rejection Testing

Test Description

The adjacent channel rejection testing is an extension of the receiver weak signal performance. The test configuration is similar to the previous discussion on BER testing basics in my last post.. This time, we will look at  verifying the sustainability of  receiver performance while an interferer signal is introduced.

Test Setup

The general setup is shown in in the lead in figure. Depending on how the DUT is specified, the receiver should have the ability to reject an adjacent signal of a given strength. We are not looking for the same performance as the DUT’s path-loss with no other interference. In this case it’s usually looking for a acceptable threshold of BER at about +10 to +15  dBm from the previous floor. For instance, you may allow your DUT to insure no BER at -75 dBm  while an adjacent channel is interfering at -60dBm in proximity.

In order to emulate this on the bench you still need to take the the cable and connection losses into consideration. With the AP set for 12 dBm, and the losses to the DUT through the system equal -7.5, to have the DUT experience -75 dBm path loss, the attenuator needs to be set to – 70.5  ( AP Tx – Connection losss -75 dBm). This can be monitored on the power meter in the set up.

Initialization parameters

Set the test channel on for the DUT on the Telemetry AP Unit and adjust the variable attenuator to a transmitted power level of –75dBm @ test channel as read on the power meter.  The external signal generator ( ESG)  transmitted power level will be –60dBm @ adjacent test channel frequency. Set the ESG signal generator to the center frequency of Channel +1 and Channel -1  around the test channel and measure it’s power at the meter. Adjust the power at the ESG to read -60 dBm at the meter.

Test Procedure

The Adjacent Channel Rejection of the DUT shall be tested by means of measuring the BER of the receiving signal sourced from the telemetry’s AP Unit in the presence of an interferingsignal provided by an ESG operating continuously using the same modulation scheme. Check to make sure the BER measured passes under the above mentioned condition.

For each channel tested, set the ESG to +/- 1 channel away, testing twice for each channel.

I hope you found this helpful. Feel free to drop me a comment.

 

Avoiding Problems during Receiver sensitivity BER testing for a Medical telemetry device

Since some medical telemetry devices fall under low power low bandwidth broadcast, the FCC has strict rules for such classification especially when used in the 2.4GHz spectrum. FCC 15.231 allows 48 narrow channels of low power and bandwidth for use in this spectrum. Normal 802.11 allows for only 11 channels of 20 MHz bandwidths. Medical Telemetry utilizes this same ubiquitous space with MUCH narrower channel spacing at 1.74 MHz per channel.

With such a narrow channel in an overused broadcast spectrum, accurate Bit-Error-Rate (BER) testing becomes even more important. Measuring Bit Error Rate (BER) in a digital system is the equivalent of measuring the Signal-to-noise ratio (SNR) in a low noise amplifier.  Basically, the BER test measures the receiver’s weak signal sensitivity.

This article looks at the basics of conducting BER measurements for lowest RSSI (Rx Signal Strength Indication).

Test Setup accounting

The general setup is shown in the BER test configuration diagram. Before any measurements are taken, one needs to measure the path losses introduced into the system. This may sound rudimentary as it’s RF test 101 practices.  This should include cable losses, splitter attenuation, and especially taking into consideration any cabling conversions ( like Type F to Type N or TNC  at the power meter and SMB to Type N at the DUT).  In my setup, I prefer to use  similar sized cables from the splitter to the DUT  and from the splitter to the power meter. Keeping  the cable conversion losses of the interconnects for both as close as possible  allows for a matched power reading at the DUT as read on the power meter!

In this configuration the total path loss from the AP to the DUT will be -6.5dBm when the variable attenuation is set to 0dBm.

Verifying the loss calculations

Connect to the access point (AP) through your server (not shown) to control the transmit power and channel access at the telemetry’s AP. Select the first channel and set the Tx power to 10dBm. In my set up, I accessed the AP through one of the diversity antenna ports. I had to configure the AP to only transmit out that port.  Also, since I removed the antenna, I had to take into consideration the +2dBm antenna gain is not in effect. Thus, in order to emulate a +10 dBm Tx from the AP, I had to set the power level to +12 dBm.

Setting the variable attenuator to 0 dBm, I was able to get a reading on the power meter of about  5.48dBm. This is a provided a good indication that I accounted for all the power losses due to the system setup.

Provided you have a command line interface to the DUT and an allowable access through a wired port ( USB, SPI etc), the DUT controller should allow you to establish a link to the AP,  read RSSI and channel data. The power level on the DUT was at +5.43 dBm. Since this is close to what is measured at the Power meter and within 0.1 dBm of the calculated and measured losses, I can be satisfied that these measurements are accurate enough to continue. I  scan through all 48 channels to insure a flat response from the AP and keep track of any variations.

Test Procedure and BER calculations

Start by establishing a link with the DUT while you have the variable attenuator set to 0 dBm. Verify you link by running the BER software at your AP.

The telemetry AP should also have a command line access interface either from the server or through a wired serial port.  Depending on the setup, the low BW low power application may have a time division multiplexing scheme where calculating BER needs to take into consideration the slots.

The AP software  should be capable of reporting  the number of bad bits, total number of slots that where sent and the number of bad slots.  A bad slot is when there is a bit error in the sync word or when the total number of bit errors exceeds more than half the total number of bits in the slot.

The equation for calculating the BER for the AP Telemetry is:

Bits in error  = Be

No of slots = Ns

No of bad slots = Nsb

n = one line of reported data (typically  100 slots / frames)

n-1= the previous line of reported data (typically  100 slots / frames)

b= # of data bits ( 8)

Slts= # of slots plus any framing slots

BER = [Be(n) – Be(n-1)] / [[[(Ns(n)-Ns(n-1)]-[Nsb(n)-Nsb(n-1)]]*b*Slts]

Once you are sure you have a good link to the DUT and the AP is reporting bit error correctly, start attenuating the signal. I tend to use a coarse adjustment to  increase the attenuation until I just start to see the point where BER is reported.  Then using the fine adjustment, I record where the “knee” starts to occur and look for the 0.001 threshold.

Sample Test Results

In the setup I used, I compared the receiver’s sensitivity threshold using both a USB power supply connected to the DUT ( switching power) and just Batteries running the DUT.  The chart below shows the results averaged over the 48 Channels for each case. The .001 BER threshold was -90.5 dBm for the USB power connected DUT but with a batteries I was able to push the sensitivity of the telemetry receiver to -94 dBm.

  

Next time

Next time I will discuss how to expand this set up for Adjacent Channel rejection testing.

If you found this helpful or if you have any suggestions, please feel free to leave me a comment.