HP 11693A limiter broken...?

[ytterligare]

Hello, a couple of years ago I got this limiter from eBay, to protect my HP 8563A.
Now, fun fact, I thought I lost it ever since as I wasn't able to find it in my room until today ( I'm a bit messy, I know...).




So I decided it was worth a check on it to see whether it could really "level" the amplitude of the signal at the SA ingress or not.

Now, from Keysight specs :

- Frequency range 0.1 to 12.4 GHz, useable to 18 GHz
- Guards against input signals over 1 milliwatt up to 1 watt average power and 10 watts peak power

In my noob mind, it was reasonable to think it could level to 0 dBm the signal > 0 dBm.

I'm sad to report I probably got a lemon, as I could not even test its performance due to the fact that the insertion loss is too big ( a.k.a. I think it is broken ) as per the following table :

WITHOUT Limiter, HP 8672A Signal Gen Ampl. = 0dBm -> good cable -> peak amplitude read on the 8563A :

100 Mhz -> -0,17 dBm
300 Mhz -> -0.67 dBm
1 Ghz -> -0.67 dBm
2.1 Ghz -> -1.17 dBm
4 Ghz -> -1.17 dBm
8 Ghz -> -1.33 dBm
12 Ghz -> -1.50 dBm
18 Ghz -> -3.83 dBm


While, WITH 11693A Limiter :

100 Mhz -> -39.67 dBm
300 Mhz -> -31.17 dBm
1 Ghz -> -20.33 dBm
2.1 Ghz -> -18.50 dBm
4 Ghz -> -9.5 dBm
8 Ghz -> -6.17 dBm
12 Ghz -> -4.67 dBm
18 Ghz -> -6.17 dBm

Anyone had the chance to crack it open a specimen before and was successful in repairing it, or at least any idea what to look for ? In my opinion, with frequencies > 10 Ghz, it seems to be working well or generically speaking, it seems to perform better the higher in frequency I go...

Tnx
Andy

P.S. : I should have known something was fishy, as the eBay listing picture was showing a different limiter ( different s/n ) 

[MaxFrister]

I always assumed that it could limit (i.e., shut off) and never thought it could "level" to 0dB.

What happens if you try a bit lower power, say -10dBm. 

I have a limiter (not sure the number off the top of my head) but I can check this evening.

[ytterligare]

I always assumed that it could limit (i.e., shut off) and never thought it could "level" to 0dB.

Well it might be because I'm not native English language speaker but in this case I assumed they meant the same, limiting the output at a certain level ( in this case, for the HP limiter ->  +0 dBm ).

Turned out that maybe the "leveling" is not that precise : take a look at the graph for my other limiter : you'll notice by the lower left graph that once the power in reaches +13 dBm it tends to be leveled to that power ( maybe a couple of dB more ) until maximum acceptable input power is reached, which is +33 dBm .



These data/graph are taken from Mini Circuit's site, and I don't have any for the HP limiter, but I guess inputting +0 dBm shouldn't be out of the specs of the limiter....

A.

[MaxFrister]

I think the HP part would have a similar kind of graph, but shifted down.  The minicircuits says it limits to 12dBm whereas the HP limits to 1mw (0dBm).  Since you are testing right at 0dBm I think you are seeing a reasonable result. 

I do have the same part.  If you'd like I can run some (non-destructive) tests if it would help.

[ytterligare]

I do have the same part.  If you'd like I can run some (non-destructive) tests if it would help.

Max, sorry I missed your kind offer !! Please do run a test for me on your limiter, if it's not too much a trouble...

In the meanwhile, normalizing to take into account cables and adapters, I have run a transmission test through the VNA, only 100 Mhz to 3 Ghz, 0 dBm...not much of a difference...and to me this looks like a piece of junk... 





While the MiniCircuits, on other hand.... :



Ciao
A.

EDIT : corrected and unified Scale/Div