I took some measurement on Extron distribution amplifier

[tkamiya]

I just posted this information to time-nuts mailing list but I thought I'd share it here as well.

Many of us have a lab standard and use it to sync our timing related equipment, be it frequency counter or RF generators, etc.  One of the very popular option is inexpensive video amplifier widely available on eBay.  Whenever this topic comes up, someone always mention port isolation and distortion not up to par.  I haven't seen any test results though.

Since I have a few of these and use them often, I thought I'd do some measurement.  Please note, the device is completely stock.  I did not modify anything including input and output impedance.  They are 75 ohms.  Everything else including cables were 50 ohms.

My specimen is one of my Extron MDV3.  Ebay special I paid $10 for it.  1 port in, 3 port out. Completely stock.  This is a very little device, obviously.  I'd like to do a similar test with larger ones as well but unfortunately, I don't have one handy.

Instruments used were:

TruePosition based GPSDO producing 10MHz at 11.9dBm.  This is my test signal source.
HP8644A producing 15MHz at 10dBm.  This is my "worst case" interference source.
HP8591EM spectrum analyzer
HP437B power meter + 482A sensor

I was mainly concerned with distortion and port isolation in reverse direction and port-to-port.

Here is how I did this.  Please know that other than looking at harmonics, I don't have a real good way to measure distortion.  So if someone wants to replicate this and have proper equipment, that would be appreciated.

1)  I connected 10MHz to OUTPUT port of Extron MDV3. 
No signal were detectable on input port either by power meter or SpecAn.

2)  I connected 10MHz to OUTPUT port. 
No signal were detectable on other output ports either by power meter or SpecAn

3)  I connected 10MHz to INPUT port and took a power measurement at output port of Extron.
11.2dBm

4)  I connected 10MHz to INPUT port and looked at output port by SpecAn.  Keeping eyes on CRT
Then I connected 15MHz to OUTPUT port.  This, is a simulation of a worse case scenario.
ZERO change on CRT

5)  I then connected HP8644A to the input port and hooked up SpecAn to one of the output.  Then slowly increased the output level.
Up to 12.3dBm input, it was pretty clean.  Above this, 2nd harmonic started to show up.  At 14dBm, all harmonics were -40dB below the primary signal.  This is probably the limit I'd go.  I went up to 20dBm.  Pretty messy....  it looked like a comb generator!

My concision is this....
With this, I can conclude this particular device has a "pretty good" isolation between output port and output to input port, as well as sufficient protection for reverse leakage.  For standard level timing signal, it's pretty clean.  Obviously, my result is specific to my device.  Some future/past design revision, other models, other makes, etc has to be re-measured.

I left test setup as-is.  If anyone want me to test something else, I'd be happy to.  If you want to add  your own result for this or other device, please do.

I am now using this little amp as built-in buffer/mini-distribution in DIY projects.  One of such is my new GPS steered Rb.  I am happy with the results.

[4thDoctorWhoFan]

All I know is that I have been using 2 of the larger Extron distribution boxes for a while now and have had no problem.  They work great.

I replaced all the 75 ohm resistors to 50 ohms and built into the same box the 10 Mhz standard.  I have an HP OCXO in one box and a GPSDO in another box.  I really like these boxes.  They work great.  There is plenty of room inside to add the standards.  The pictures below show the GPSDO installed inside with a power supply to power it.

P1010142 (2)

P1010140 (2)

P1010139 (2)

P1010141 (2)

[don.r]

I think quite a few of us now have Extron boxes thanks to Gerry Sweeney. Mine was moded to 50 ohms by adding a 150 ohm resistor in parallel. I haven't found blanking plates yet so each row remains a separate circuit which is fine because I only have 2 devices that use 10MHz input so why I have 18 outputs is beyond me. 

[tkamiya]

Very nice....

I briefly considered both methods to make it 50ohm box all around.  But I decided against it.  VSWR of 1.5 is pretty good and it's a sine wave, so there is only fundamental frequencies to be concerned.  What I have been doing is taking my 1 input 3 output box and putting it in a larger case with an oscillator.  My GPSDO + Rb setup is near completion.  It has this dist box inside it, box and all....

It's hard to believe how cheap these things are.  Just bought 4 more.  10 dollars each with shipping included. 

I see you also have a GPSDO in yours.  What did you do with PPS output?  My unit is almost done.  Last thing to do is to make a pps distribution board, drive an LED from it, and feed 2 output to outside.  I usually make a small board with 74AC04 but I'm getting tired of doing this same circuit over-and-over.