Products > Test Equipment

New bench scope - Fnirsi 1014D, 7", 1GSa/s

<< < (43/67) > >>

donwulff:
Honestly, I'm talking about FNIRSI 1014D not because it's the absolute best (it obviously isn't), but because this is the thread for it and I already own it. The argument for price falls flat with that, although I'm considering 1013D because it already has working open source firmware. I can probably get it working on 1014D if I'm going to have the time to play with the software at all.

I'm taking notes for "unbreakable" hobbyist scope though, but I can't help but think there has to be a thread for that. DSO1511G has same AD9288 ADC as FNIRSI, LCMXO2 is 1280 LUT CPLD that should support 8k sample depth and has lot less power and potential. More importantly, it's single channel, which already disqualifies it for most uses, and more expensive per channel, and the review says it struggles at 21.5MHz. I paid 172 EUR for the FNIRSI (I keep forgetting AliExpress cheats and always lists price of cheaper option, you can get it cheaper but maybe not down to 150 EUR unlike the 1013D if you don't need signal generator, tactile sensation and look professional instead of causing a security alarm for unauthorized tablet;)


Also speaking of ADC, in https://www.eevblog.com/forum/testgear/fnirsi-1013d-100mhz-tablet-oscilloscope/msg4425640/#msg4425640 @marauder reported exchanging ADC buffer opamp. I'm assuming those are under the shield boxes on the PCB, so wondering how hard they'd be to replace, especially since my scope seems to have slightly different amplitudes on the channels too. Indeed was just looking at available components and thinking TI OPA355/OPA356 could be a better performing replacement, except the datasheet says "with sampling speeds up to 1 MSPS", but I'm sure it depends on other components. It sounds like @marauder used these without problems though. It's not going to get around any limitations of the AD9288 ADC, but on paper does seem to have lot better specs (360V/uS, 5.8 nV/Hz, 200 MHz G10@R1k vs 180V/uS, 8 nV/Hz, 10 MHz G10@R1k! - shouldn't that mess with the response?). So that would be interesting experiment to reduce noise and maybe recover most of the ADC resolution. I'm also wondering whether it would make sense to adjust the gain to be more sensitive with lower noise floor, if you can change the software to match the gain (At true 50mV resolution you should have 12V range which seems fine for most purposes, even 25mV would work for digital logic but probably more issues). If you're going to change that many components, you run risk of introducing more noise... And the AD9288 datasheet shows it could be pushed at least to 110MSPS per channel, for increase in noise if the FPGA can be pushed that high. Of course I'm not sure anybody has tested how far the silicon goes, especially if it's not genuine. Just playing with some ideas to get the most out of existing hardware.

Edit: The 10MHz is probably wrong, the specs for both of them have some differences in assumptions/units and it's hard to compare, especially as I'm not really analog circuits guy either. The figures in "Small-Signal Bandwidth" for AD9288 do look bad though. I'm not thrilled to be making component substitutions, but swapping the OpAmp wouldn't be out of question (before looking under the tin cans), especially due to the gain(?) difference between the channels.

pcprogrammer:
Ah yes, the defective opamp in marauder his scope.

The analog circuit is indeed underneath the tin cans, and could do with more improvements then just swapping the opamp. At the least the change evi made to get rid of the ground lift problem. https://www.eevblog.com/forum/testgear/fnirsi-1013d-100mhz-tablet-oscilloscope/msg4536116/#msg4536116

It would be better to make a dedicated dual supply for the front end and another option could be to bypass the front end altogether and make a new one, but then it grows and grows into what one thinks up next.  >:D

The one thing I like about these two scopes is the "simplicity" with which new firmware can be wiped up, but that is only now after many and many months of work to get to the root of the system. Even for the 1014D it could be done relatively quick. All that needs to be figured out first is the FPGA command to load the function generator with its waveform, the controlling of the frequency, which most likely is done with the clock ic and the commands send from the user interface processor to make the settings.

Basing it on the new 1013D firmware can be done, but needs tweaking of the screen layout and if it needs to have the same look and feel as the original 1014D the graphics engine needs some polishing.

There are some differences in the FPGA commands. The screen brightness works with only a byte of data, so the range needs to be determined. Also saw some difference in the sampling control, but not that big of a deal.

About raising the sample rate, probably not worth the trouble cranking it up with 10%, with the risk of damaging the ADC's. I guess the FPGA can handle it, but needs a proper design to get it done.

Rasz:

--- Quote from: donwulff on December 30, 2022, 09:19:03 am ---Honestly, I'm talking about FNIRSI 1014D not because it's the absolute best (it obviously isn't), but because this is the thread for it and I already own it.
--- End quote ---

Im sorry for your loss. Being in this thread should make you realize you got scammed. Best course of action is selling this piece of crap and moving on with your life.

donwulff:
I'm skimming through the old threads, just a lot to read lol. Nothing compared to actually reverse-engineering the stuff though! Anyway I edited the earlier post slightly, I'm not really sure on the specs of the Gain-Bandwidth Product/Small-Signal Bandwidth of those opamps, as the datasheet make different assumptions and not all written out etc.

I'm not too thrilled to be replacing components, either, but it's worth noting you probably can't build a lot better scope without quite different and more expensive architecture (I'm curious how the low-price 200MHz scopes do and if they're anywhere real that performance in reality), which suggests to me small improvements for DYI and learning could be worth it. Software, if it gets re-used, will at least see better return on the time used. On the other hand, if you're anyway learning & doing hobbyist projects, can't put a price on everything. Still do need to find time, though.

pcprogrammer:

--- Quote from: Rasz on December 30, 2022, 11:50:45 am ---
--- Quote from: donwulff on December 30, 2022, 09:19:03 am ---Honestly, I'm talking about FNIRSI 1014D not because it's the absolute best (it obviously isn't), but because this is the thread for it and I already own it.
--- End quote ---

Im sorry for your loss. Being in this thread should make you realize you got scammed. Best course of action is selling this piece of crap and moving on with your life.

--- End quote ---

That is bullshit. It is a nice piece of kit to play with from a development perspective. It has a 7 inch display, a reasonable fast processor and a nice FPGA. The second processor to handle the user input gives another bit for learning. Inter processor communication and scanning of rotary dials and buttons. With the schematics available it is a nice entry point to gain knowledge on programming both hardware and firmware.

The other scopes you mentioned have smaller screens and one would need to start from zero and reverse engineer the firmware to be able to make new firmware for it. Most of that work has been done for the 1014D.

I would say if you have nothing positive to add, just stay away from posting in this thread.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod