<Degi>
What if we could bias the input of the scope
<Degi>
Then you could directly sense photodiode output
<Degi>
Oh neat, photodiode conducts like 5 µA when I shine my phone light on it... Hmm... According to my quick calculation, it should output like 5-50 kA when the thingie shines on it... I think itll need some 60 dB of damping
bvernoux has quit [Quit: Leaving]
<electronic_eel>
Degi: if you want to sense signals from photodiodes, you want to really bias the diode, like -15v or more. and then you want a nice transimpedance amplifier afterwards, something like two stages or similar
<electronic_eel>
don't think that works for a regular scope input
<electronic_eel>
but what I would like to have is a specialized active photodiode probe
<electronic_eel>
but I think that circuit can be improved in several parts
<electronic_eel>
...areas is the better fitting word than parts
<Degi>
Hehe I've build a photodiode input from a photodiode and a capacitor and a 9 V battery lol
<Degi>
Gotta test it later at 50 ohms
<electronic_eel>
the question is what you want out of it
<Degi>
Hm maybe stick an SFP to an input
<Degi>
Well the signals I wanna measure with this one are very bright anyways
<electronic_eel>
I'd like to have something to monitor pwm and update rates from displays
<electronic_eel>
but and SFP sounds like you want something really fast
<Degi>
Yes I only wanted like 70 MHz BW or so since thats what my scope currently can do
<electronic_eel>
didn't you just buy a new scope?
<Degi>
Yes that one currently has 70 MHz BW but can be hacked to 350
<electronic_eel>
why didn't you hack it yet?
<Degi>
Idk was kinda too lazy, also how does it relate to warranty? Does it affect that any?
<electronic_eel>
don't know, I think they mostly care for the opening-prevention sticker, the hologram-like one
<Degi>
Hm okay
<Degi>
Idk, should get around to it soon-ish
<Degi>
It can measure 100 MHz signals tho heh
<electronic_eel>
yeah, of course, they don't have a brick-wall filter in there
<electronic_eel>
it is just the level that is attenuated a bit
<Degi>
Yes looks more sine-y
<electronic_eel>
but even if you just want to measure 70 MHz from a photodiode, I don't think you will get far with just a photodiode and a battery on a scope input
<electronic_eel>
the diode outputs a current, you have to convert it into a voltage to measure it on a scope
<Degi>
Well that does the 50 ohm resistor
<Degi>
It outputs like 5 microamps when I point my phone light at it, that's quite a bunch
<Degi>
Like my phone light is maybe 20-30 dBm and I have like 110 dBm available
<electronic_eel>
hmm, don't know, usually you want a proper transimpedance amp with a lot of gain there
<Degi>
Yes, but not in this case. This case needs attenuation before the photodiode
<electronic_eel>
huh, why that? don't photodiodes output just a few nA or µA?
<Degi>
Well it does 5 µA with my phone lamp
<Degi>
Which has to be like idk 20-30 dBm maybe?
<Degi>
Like the light I wanna measure is up to 110 dBm and would probably melt the photodiode if it was undamped
<electronic_eel>
how does dBm work with optic power? I'm always a bit puzzled by all the units used in opto systems...
<Degi>
Well dB milliwatts, so 10^(110/10)*1 mW
<Degi>
Idk if thats often used in opto systems, I think its used for data transmission?
<electronic_eel>
yeah, but I've never seens leds or other lights having dBm datasheet figures
<electronic_eel>
probably it isn't used that often for leds or lighting because they spread over an area and some wavelenght, both isn't that much of an issue
<Degi>
Hm yes usually used for confined light
<electronic_eel>
and for comms it makes it easier to calc if you work in dB
<electronic_eel>
...isn't that much of an issue >for a laser<
<Degi>
Hmm idk
<Degi>
Like if you have a filter with 30 dB absorption and you stack two, you have 60 dB absorption
<electronic_eel>
yes
<Degi>
Anyways using linear extrapolation, the laser would cause like 50 kA to flow through the photodiode (besides vaporizing its surface), so it needs to be somehow damped that only some mA flow through it
<electronic_eel>
hmm, which kind of laser do you have that outputs 110 dBm? thinking about it, this sounds hellishly dangerous
<electronic_eel>
and I don't think something like this is built into regular SFP slots...
<electronic_eel>
...modules I mean
<Degi>
Idk you probably shouldn't look at it or be in the same room tbh
<Degi>
Well it is a Q switched Nd:YAG with 20 ns pulse duration of 2 J energy
<electronic_eel>
I just looked at the specs for some random 1000Base-BX module and they specced 3 dBm power out
<electronic_eel>
which sounds more reasonable to me
<Degi>
Hm yes, that'd need an amplifier
<electronic_eel>
now if you have a pulsed laser, what do you want to measure? the pulse shape? the power amplitude? ...?
<Degi>
The pulse shape
<Degi>
I think I can measure the amplitude with a small calorimeter
<electronic_eel>
so if you use just a resistor to convert to voltage, you are faster than any opamp, given that you have enough current
<Degi>
yes
<electronic_eel>
that is a rare thing for photodiodes
<Degi>
Idk if that actually works or if there is some kind of saturation effect or how receptive it even is at 1064 nm
<electronic_eel>
most photodiodes are naturally sensitive in a very wide band, the peak is often around 900nm. the manufacturers then offer some that have filters built in for the users that don't want that wide sensitivity range
<azonenberg>
electronic_eel: its basically the same as sgmii minus the 10/100 mode
<azonenberg>
8b10b encoded raw ethernet framing at 1.25 Gbps
<azonenberg>
with some control characters to denote start/end of packet, idle, etc
<electronic_eel>
so they just send out pulses? they don't have some frequency they modulate it on top of?
<Degi>
Hm doesnt 8b10b do that? It sets a lower frequency limit
<Degi>
Well they do modulate it on top of a frequency of some hundred terahertz heh
<electronic_eel>
I thought lasers don't like to be switched on and off, so it is better to use some carrier and modulate your signal on top
<Degi>
Oh theyre dc biased too
<azonenberg>
electronic_eel: no ethernet over fiber is always baseband modulation
<azonenberg>
i think technically it's binary ASK so there's a low and high level pulse
<azonenberg>
instead of off and on
<azonenberg>
actually ethernet over copper is too. that's what the "base" in 1000base-sx etc means
<electronic_eel>
ah, ok, that is how they get around switchin on and off
<azonenberg>
Yeah but if you set a threshold midway between them it's effectively OOK
<azonenberg>
for all intents and purposes
<electronic_eel>
why do you want to watch sx? isn't it easier to convert it to copper and watch the GMII or similar?
<electronic_eel>
or even the pins on the SFP module?
<azonenberg>
no i just think it would be a cool project to look at the optical power directly
<azonenberg>
see how fiber length effects it etc
<azonenberg>
i have no practical application
<azonenberg>
its more if i am building an optical input probe
<azonenberg>
i want it fast enough to decode ethernet
<electronic_eel>
hmm, makes it more difficult and expensive, and probably not suited well for visible light stuff
<electronic_eel>
I would want one for visible light for example, to see display update rates, pwm of background lighting in displays, flickering of leds and similar stuff
<azonenberg>
ok yeah that would also be useful
<azonenberg>
low MHz bandwidth max
<azonenberg>
maybe with an integrating sphere or similar attached
<azonenberg>
i was thinking about something with an 850nm simplex LC input lol
<electronic_eel>
ah, not too low MHz, maybe 200 or 300 MHz, so you can really see the pwm and fast updates