PECL measurement with oscilloscope probe

[chinito]

how to measure pecl with oscilloscope

Once I terminate PECL device with 50 ohm to VTT, can I measure the output using 50 ohm oscilloscope probe? Where can I find info about ECL/PECL/LVDS etc.

Thanks.

 

[flatulent]

Re: PECL measurement

The probe in parallel with the existing 50 ohm resistor will produce a 25 ohm load. This will affect the performance of your circuit and make false readings. There are some high speed probes that are active with high input impedance. There are other probes with a few hundred ohms input impedance. These should be used.

Motorola and other manufacturers of these parts will have information on their web sites.

 

[Okkam]

Re: PECL measurement

Hi

You can find information on ON-Semi ( previously Motorola ) web-site
www.onsemi.com

Or similar product line from Synergy-Micrel www.micrel.com ( High-bandwidth products SY100-SERIES ).

Best regards.

 

[krashkealoha]

Re: PECL measurement

In order to look at the output with an oscilloscope, the 50 ohms to ground of the scope will terminate your output. Since Vtt is at ground, Vcc or Vdd is at +2V. The reason for this is that Vtt = Vcc - 2V. You will then force Vee or Vss at Vsupply - Vcc(Vdd). Therefore, if you supply voltage is 3.3V, then Vee(Vss) will be at -1.3V. This method is called "split supply."

 

[rfmw]

Re: PECL measurement

There is another way of termination a PECL device with 50 ohm oscilloscope. The only limitation or drawback is that the oscilloscope is AC-coupled to the PECL device.

So one could use a pull-down resistor (value around 200-270 ohms) at the PECL output pin to Vee and insert a coupling capacitor in the 50 ohm microstrip line which is connected to the oscilloscope. Another way would be connecting a rf choke plus series resistor 47 ohm from the 50 ohm data line to the Vtt. Again the coupling capacitor is needed in the data transmission line to the oscilloscope. These two methods work just fine compared to the 50 ohm to Vtt termination with almost no performance degradation. However it still depends on the (functioning of the) circuit, so the designer must select best termination method for himself...

cheers,
rfmw