Function Generator protectionbox using EL2009's

[Verpies]

Looks well.

Keep in mind that this is an analog voltage source, so if you connect a LED to its output and set it to output 40Hz sine wave voltage, then the brightness of the LED will not be sine shaped.  This is because the brightness is proportional to current ...and the current is not sine shaped (as shown here).

To convert the EL2009 to an analog current source you'd have to do Diag.2.

[Verpies]

Also, keep in mind that the EL2009 box does not provide any galvanic isolation and will not protect you from evil loads like the Kacher or TC.
Making a DC+HF analog buffer with galvanic isolation would put hair on your chest!  But if LF and DC operation is not needed, then the isolation can be achieved by a HF 1:1 signal isolation transformer between the FG and the EL2009 box.

For such evil loads, the galvanically isolated digital switch (rectangular pulses only) from Diag.3 can be used.  Its major disadvantage is the cost of these four DC-DC converters, when you buy them as premade commercial devices.

[Itsu]

Thanks, i want to stay away from Kacher / TC like devices as much as possible anyway.

Not sure what evil load Peters crackling coil is, therfor i will use his specially designed box for it for the time being.

I did found my Tek scope inoperative after some measurements on his crackling coil the other day and needed to do a full calibration to get it back.
So i will use my Owon scope to do measurments.

Itsu

[Itsu]

For Lost_Bro and others,

Bill Of Material:


1x  TEKO 373 metal HF 106x50x26 CF373.
2x  BNC-Adapter 90° 50 Ohm, Amphenol B5072A1-ND3G-50
2x  Connector BNC male plug bulkhead handle solder panel mount straight
2x  Connector BNC female plug bulkhead handle solder panel mount straight
2x  EL2009CT
2x  isolation pads to isolate the EL2009's from the casing
2x  tantalum capacitor 10uF 35V
5x  ceramic capacitor 100nF
6x  feedthrough capacitor 2nF
2x  50 Ohm inductionfree resistor and/or 100 Ohm multiturn potmeter (for intial adjustment)
4x  ferrite beads
4x  M3 screws and nuts to mount EL2009 and ground lugs.
2x  ground lugs.

The above list is for a 2 channel FG, so if you want only 1 cannel protected, the list can almost be halved.
The number of feedtrought and 100nF decouple capacitors can vary depending on how clean the input voltages are.
I use a separate bulky +/- 24V 7A powersupply set to +/-16v to feed the EL2009's, but variations are of course possible
The EL2009 can withstand a max curent of 1A, so that is minimal needed.
I did not use the mentioned (and drawn) Schottky diodes (from supply voltages to output) as they caused severe oscillations.
Perhaps the used  SB240-E3/54 diodes junction capacitance of 170pF is to much.

Regards Itsu

[poynt99]

Over all Itsu, are you happy with this build and its performance?

[Itsu]

Hi Poynt99,

i see you found the thread,   i am happy with it, allthough i do not have used it often, its nice to have when needed.
Only drawback is the fact that i have to hook it up to my +/- 24V PS (set to +/ - 16V), so there i still need to set up a dedicated +/- 16V PS.

Itsu

[poynt99]

Ok, thanks itsu.

It seems you ran into a few problems with it and wanted to make sure they were all resolved and that it is performing well. I'm playing with the idea of using this or another method to accomplish Pin measurements, all the while protecting the FG as a bonus.

[Verpies]

According to the datasheet, the EL2009 has approximately 2MΩ input impedance and approximately 1Ω output impedance.
This makes it excel at driving low impedance loads despite it not offering any voltage gain.

You had terminated its input with resistor/pot so its combined input impedance approaches 50Ω in order for it to match the 50Ω output impedance of the FG.  You had a hard job with this termination because you did not have the VNA then ...but I digress.

I was wondering whether it would be possible to take advantage of the EL2009's high input impedance and terminate it to be 1MΩ, so a scope probe can be connected to its input ?  Without it becoming unstable/self-oscillating...

This would be useful in situations where it is unacceptable for your SA's 50Ω input impedance to load the DUT so much.
Kind'a poor-man's active HiZ "FET" probe, which has no problems driving the 50Ω input impedance of the SA with signals from high impedance sources (e.g. resonant parallel LC tanks) ...and without loading them down.