azonenberg changed the topic of #scopehal to: libscopehal, libscopeprotocols, and glscopeclient development and testing | https://github.com/azonenberg/scopehal-apps | Logs: https://freenode.irclog.whitequark.org/scopehal
juli966 has quit [Quit: Nettalk6 - www.ntalk.de]
maartenBE_ has joined #scopehal
maartenBE has quit [Ping timeout: 256 seconds]
electronic_eel has quit [Ping timeout: 265 seconds]
electronic_eel has joined #scopehal
_whitelogger has joined #scopehal
Degi_ has joined #scopehal
Degi has quit [Ping timeout: 240 seconds]
Degi_ is now known as Degi
_whitelogger has joined #scopehal
_whitelogger has joined #scopehal
sorear has quit [Ping timeout: 264 seconds]
sorear has joined #scopehal
<azonenberg> Gonna be assembling the AKL-PD1 prototype in a few minutes
<azonenberg> as well as replacing the AKL-PT2 clips with the new ones
<azonenberg> also i think i finally resolved the last of the EQ issues around the PT2 prod run
_whitelogger has joined #scopehal
<azonenberg> So it looks like the relationship between the project and Pico is starting to develop. They're going to be sending me a 6248E as a dev platform so I can add driver support
<azonenberg> They're also interested in offering a 5000 series dev unit to somebody interested in adding support for that. Obviously some level of commitment to being a driver maintainer in exchange is expected
bvernoux has joined #scopehal
<bvernoux> hello
<bvernoux> azonenberg, woo very nice the AKL-PD1 congratulations
<bvernoux> I'm impatient to see the results ;)
<azonenberg> I'm planning to use the ADL5580 for the next generation of the AKL-AD1 (the matching amplifier for the PD1 tip)
<azonenberg> current one is ADL5565 based and has... issues
<azonenberg> bvernoux: also not sure if you were around when i mentioned this
<azonenberg> but the CEO of Pico reached out to me about developing a relationship between them and our project
<azonenberg> Since it has a lot of advanced capabilities their software lacks
<azonenberg> (01:03:12) azonenberg: So it looks like the relationship between the project and Pico is starting to develop. They're going to be sending me a 6248E as a dev platform so I can add driver support
<azonenberg> (01:03:56) azonenberg: They're also interested in offering a 5000 series dev unit to somebody interested in adding support for that. Obviously some level of commitment to being a driver maintainer in exchange is expected
<bvernoux> I was not aware of the ADL5580 i have seen results of ADL5565 which are disappointing for a diff amp advertised as 6GHz which is in fact 3GHz
<bvernoux> very strange that AD do that marketing bullshit on ADL5565 ...
<azonenberg> The s-parameters for the 5580 look solid
<bvernoux> ha great
<azonenberg> This is S31 which is forward gain from VIN_N to VOUT_N
<bvernoux> ha yes very nice
<bvernoux> it shall work fine up to 10GHz
<bvernoux> let say very clean up to 5GHz+ here
<azonenberg> Yeah it's advertised as 10 GHz and is very flat to 5
<bvernoux> as the scale is not easy to read and check -3dB
<azonenberg> well this is their S2P, there's a lot more resolution in the raw data
<azonenberg> S4P*
<azonenberg> I'm just plotting it as an overview
<bvernoux> it is pretty expensive 41USD
<bvernoux> for 1kunit ;)
<azonenberg> We're talking about a multi GHz differential probe here
<azonenberg> the stuff from lecroy, tek, etc is four digit price tags
<bvernoux> yes ;)
<azonenberg> so using an amp that costs a hundred bucks or so in low volume isn't horrible
<bvernoux> very impressive it is very flat even up to 6GHz
<azonenberg> I can always make a cheaper one later on. Right now i'm trying to source the part though
<azonenberg> i see lots of distributors with devkits
<azonenberg> havent found the chip itself for sale yet
<bvernoux> yes great
<bvernoux> Do you think you can buy 1 unit
<bvernoux> ?
<bvernoux> as maybe they ask for big batch to sell only 5580 for big companies ?
<azonenberg> I'm sure ADI will give me a sample if i tell them i'm designing a differential scope probe
<bvernoux> if they do not have some agreement with Lecroy... ;)
<bvernoux> sometimes buisness is very bad with exclusivity ...
<bvernoux> but it is clearly a must have if the hardware is like on simulation and very flat like that up to 6GHz+
<bvernoux> which can be maybe usable up to 10GHz for extreme case ;)
<bvernoux> so it is a big bonus
<bvernoux> in datasheet they provide 1dB flatness up to 6GHz very nice
<azonenberg> So ADI can give me two free samples of it
<azonenberg> I can buy it from them in reels of 1500
<azonenberg> I assume it will hit distributors soon, it's a brand new part
<bvernoux> it is programmable through SPI so it will need a small MCU ...
<azonenberg> Yeah, I know. I'll probably throw a stm32f031 on the prototype
<azonenberg> I'm also going to add ferrites and local LDOs to the new board
<azonenberg> It will make it more complicated, but eliminate noise from the long bench supply leads
<azonenberg> And add an EMI can
<bvernoux> yes great
<bvernoux> an ultra low noise LDO from TI ;)
<azonenberg> I was thinking the LTC part we used on the BLONDEL frontend for the prototype
<bvernoux> they have not added details on Single Ended output anyway
<azonenberg> LT304 and LT3093
<azonenberg> LT3042 and *
<bvernoux> they speak only about diff in / diff out or single ended In / diff out
<azonenberg> Yes, I know. literally nobody makes differential to single ended converters that have GHz of bandwidth
<azonenberg> because the only reason to build one, vs using diff output, is if you're making a scope probe that has to output a single ended signal
<azonenberg> My plan is to terminate OUTN and feed OUTP to the scope
<azonenberg> The S-parameters suggest that should work OK
<bvernoux> yes
<azonenberg> however there is a slight mismatch in the cross path
<azonenberg> S31 and S42 are very well matched, and S41 and S32 are very well matched
<azonenberg> So if you have a differential receiver the overall match is very good
<azonenberg> But S31/32 and 41/42 are not as well matched as I'd like
<azonenberg> unfortunately every differential amplifier i looked at had this issue and i'm not sure if there is a solution
<bvernoux> you need a negative voltage supplot of -1.8v too
<azonenberg> short of having twinax output to the scope and a true differential input port on the scope or something
<bvernoux> and positive voltage supply +5v
<azonenberg> That's not a problem, i'm assuming bipolar supplies for all of my amps
<azonenberg> the ADL5565 on the current probe is configured with +/- 2.5V symmetric supplies
<azonenberg> asymmetric is a bit odd but i can make it work
<bvernoux> the challenge is to find something ultra compact and very very high PSRR like the LT3042...
<bvernoux> for both
<azonenberg> Yeah. Right now i am not optimizing for size
<azonenberg> I'll worry about that later
<azonenberg> I might also try the ADL4960-1
<bvernoux> this part ADL5580 is very new
<bvernoux> datasheet is Rev0 from 2020
<azonenberg> It's only 5 GHz bandwidth but has higher gain (variable to 18 dB) and is much less expensive
<azonenberg> and is less new and probably readily available at distirbutors
<azonenberg> ADA4960-1*
<bvernoux> ha yes could be a good alternative with good performance
<bvernoux> especially if the ADL5580 which is very new is too hard to buy (or to have sample)
<azonenberg> I can get two samples. they're in my cart at AD now
<azonenberg> i just don't know if i can get more than two
<bvernoux> ADL4960-1 seems to requires just 5V power supply
<azonenberg> I plan to run it split +/- 2.5V
<azonenberg> such that the nominal vcc/2 output common mode is at the scope's ground potential
<bvernoux> yes
<bvernoux> it seems clearly designed for scope probes use case too ;)
<azonenberg> I have two samples of both on the way
<azonenberg> Digikey has the ADA4960-1 for $18.48 in stock
<azonenberg> actually no, $14.60
<azonenberg> only difference is how many parts on the reel
<azonenberg> but digikey cuts it down to size, they're the same chip - not sure why they offer both versions in cut tape lol
<azonenberg> That one uses a single resistor for gain setting
<azonenberg> So no MCU, much easier to do
<bvernoux> 14.60USD available per unit is great
<bvernoux> especially if it reach 5GHz+ like in datasheet
<bvernoux> will be a very nice diff probe not too expensive
juli966 has joined #scopehal
bvernoux has quit [Quit: Leaving]
<azonenberg> So i'm thinking what i might do is respin both the probe and tip
<azonenberg> Tentatively the new amplifier will be ADA4960 based and the new probe will use tighter coupling and be a little bit skinnier