Author Topic: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS  (Read 2179 times)

0 Members and 1 Guest are viewing this topic.

Offline luky315Topic starter

  • Regular Contributor
  • *
  • Posts: 244
  • Country: at
I have a quite long LVDS chain with equalizers and isolators (ADN4622). Unfortunately there are sporadic bit errors and I need to look at the signal at different stages to optimize the equalizer settings and etc.
What would be the most cost efficient solution for this problem? Maybee a SJL GigaWave sampling scope and a frequency generator (EraSynth?)
I have a budget of 3000€, including probes.
And a related question: Are there any reasonably priced BERTs (Bit Error Rate Tester) out there? I don't need a lot of extra functions, "just" a device that transmitts data, receives it and then calculates the error rate.
 

Offline tszaboo

  • Super Contributor
  • ***
  • Posts: 7948
  • Country: nl
  • Current job: ATEX product design
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #1 on: July 17, 2024, 03:35:41 pm »
I would start with an Eye diagram at different stages. So a sampling oscilloscope, and DIY Z0 probe (or a pair of good probes that come with it) is probably the cheapest way to evaluate that.
 

Offline luky315Topic starter

  • Regular Contributor
  • *
  • Posts: 244
  • Country: at
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #2 on: July 17, 2024, 04:23:14 pm »
Which "low cost" sampling scope is able to take Eye Diagram measurements? I know that the measurement is not that simple (or better it is not that simple to get reliabble measurements, its simple to make something that looks like an exe diagram) with realtime scopes.
 

Offline tszaboo

  • Super Contributor
  • ***
  • Posts: 7948
  • Country: nl
  • Current job: ATEX product design
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #3 on: July 17, 2024, 04:38:29 pm »
Which "low cost" sampling scope is able to take Eye Diagram measurements? I know that the measurement is not that simple (or better it is not that simple to get reliabble measurements, its simple to make something that looks like an exe diagram) with realtime scopes.
Mostly used ones. For this budget, I would look for the rental market.
 

Online uer166

  • Super Contributor
  • ***
  • Posts: 1000
  • Country: us
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #4 on: July 17, 2024, 07:03:03 pm »
Which "low cost" sampling scope is able to take Eye Diagram measurements? I know that the measurement is not that simple (or better it is not that simple to get reliabble measurements, its simple to make something that looks like an exe diagram) with realtime scopes.

There's some new-age USB sampling scopes that aren't actually thAt expensive brand new, e.g. this might be (barely) enough bandwidth: https://www.sjl-instruments.com/

edit: just noticed you were aware of these. I think that's not a bad option but have not used one of these myself (just seen reviews).
« Last Edit: July 17, 2024, 07:10:57 pm by uer166 »
 

Offline luky315Topic starter

  • Regular Contributor
  • *
  • Posts: 244
  • Country: at
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #5 on: July 18, 2024, 06:58:19 am »
I have also never seen the "cheap" sampling scopes for taking eye diagrams, has someone experience wirh those?
 

Offline voltsandjolts

  • Supporter
  • ****
  • Posts: 2452
  • Country: gb
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #6 on: July 18, 2024, 08:55:43 am »
Cross-ref:
https://www.eevblog.com/forum/testgear/pocket-sized-6-ghz-1-tss-et-scope/
(maybe someone there has feedback, I didn't read whole thread)
 

Offline luky315Topic starter

  • Regular Contributor
  • *
  • Posts: 244
  • Country: at
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #7 on: July 19, 2024, 03:02:29 pm »
Thanks, the GigaWave looks like a nice device, I'm ordering one as soon as they are up and running again!
 

Online joeqsmith

  • Super Contributor
  • ***
  • Posts: 11938
  • Country: us
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #8 on: July 20, 2024, 06:17:14 pm »
I have also never seen the "cheap" sampling scopes for taking eye diagrams, has someone experience wirh those?

If there is something specific you would like me to show with the 6400, feel free to ask. 

Their latest software is pretty decent.  You can download it and run it without the scope in demo mode if you just wanted to try it out. 

Offline luky315Topic starter

  • Regular Contributor
  • *
  • Posts: 244
  • Country: at
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #9 on: July 21, 2024, 07:21:53 am »
Hello Joeqsmith,
I (and I think a lot of other people here...) would be very much interested in your experience with the GigaWave devices. I am a bit concerned that the inputs are only +-1V, so nearly every signal that I have in my line of work  (CML, LVDS, (P)ECL)....) has to be attenuated to look at it
 

Online joeqsmith

  • Super Contributor
  • ***
  • Posts: 11938
  • Country: us
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #10 on: July 21, 2024, 01:32:26 pm »
I would think the signals would be attenuated by your choice of probes.  If you needed to look at the trigger's edge, you have a 6dB loss from the splitter as well.  You could check out Lasmux's thread on their active probe as well.  I did show a few decent 10x resistive probes I made. 

There are also two reviews.  The Gigawave thread contains a lot of information that goes beyond the review.  The problem with the reviews is the software wasn't mature while I was working on it.  Some of the problems I ran into were improved after the release.   It made for a choppy video where I repeated a few experiments. 

Have a look and let me know if there is something specific you would like to see.

Offline luky315Topic starter

  • Regular Contributor
  • *
  • Posts: 244
  • Country: at
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #11 on: July 22, 2024, 02:40:55 pm »
I have one very basic question: Is the Gigawave the right tool for looking into high speed digital signals? I'm not only thinking about LVDS, but also USB 3.x, Gigabit Ethernet etc. I have often situations where I need to tune transmission lines, very long or non-ideal cables and linear redrivers and each time I have to go back to the HF lab at my old university to use the nice, but absolutely non portable and expensive 40GSPS LeCroy WaveRunner.
 

Online joeqsmith

  • Super Contributor
  • ***
  • Posts: 11938
  • Country: us
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #12 on: July 22, 2024, 03:35:14 pm »
Hardly a basic question.  "Right tool" is difficult to answer without you providing all of the constraints.   It's certainly well made, low cost, easy to use, fits in your shirt pocket, open interface allowing you to write custom software for it.  In the case of the prototype they provided, it exceeded every specification they claimed, which I demonstrated as part of my review.

It is a sampling scope which imposes certain limitations compared with your LeCroy.   The next limitation is the trigger.   Then comes the slow update rate (which is going to be signal dependent).  I would say last, you have a product that works differently than a typical oscilloscope.  The company has made a LOT of improvements to the manual which documents much of these details in terms that even I can understand.   I strongly suggest you take time reading that before making a choice.  You could also ask the company for help on deciding if it would be a good fit.  They are very honest and are not trying to sell products to people that can't utilize them.   

As for Ethernet,  I tried experimenting with PAM and I would say for PAM4, it could be useful.  PAM8, things fall apart.  Those experiments are posted in the long thread. 
 
Not sure what you are looking to spend.   Regards to "cheapest", the Gigawave takes it.  But it also needs to fit your requirements or that money is wasted.  That's your job to figure out as no one is going to understand your applications like you.   I mentioned PicoScope a few times in that thread as they also offer a low cost, high BW sampling scope.  Keeping in mind, low cost is 10+. 

Not very helpful but again, if there is something specific you would like to see, let me know.   


Offline SJL-Instruments

  • Regular Contributor
  • *
  • Posts: 217
  • Country: us
    • SJL Instruments
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #13 on: August 01, 2024, 08:18:00 pm »
Just found this thread - some responses below, and happy to answer any other questions you might have!

All EU orders ship for free with duties paid by us (DDP). The price shown is VAT-inclusive and will be the total cost to you.

The 6 GHz bandwidth of the GigaWave will give accurate results up to ~3 Gbaud. Your 2.5 Gbps LVDS is fine by this metric.
If your 3000€ budget is strict, the GigaWave is the only option with sufficient bandwidth, outside of the used market.

I am a bit concerned that the inputs are only +-1V, so nearly every signal that I have in my line of work  (CML, LVDS, (P)ECL)....) has to be attenuated to look at it
I would think the signals would be attenuated by your choice of probes.  If you needed to look at the trigger's edge, you have a 6dB loss from the splitter as well.  You could check out Lasmux's thread on their active probe as well.
Yes, usually your probe will provide more than enough attenuation.
Keep in mind that every scope with a selection of input ranges is simply swapping out amplifiers/attenuators into a fixed-range frontend. The ±1V rating is an inconvenience but does not limit the range of applications - you just need some SMA inline attenuators on hand.
More of an issue is having too much attenuation - a 350 mVPP swing into a 20x probe will result in an 18 mVPP eye diagram. Relative to the 1 mV RMS noise floor of the instrument, you may run into dynamic range issues. If you can tolerate the loading, a 1:5 or 1:10 Z0 probe will work better.

Due to the principle of operation (single comparator), the fidelity of the instrument is limited to ~1e-4 BER within the open eye. For tuning equalizers/isolators, you will be able to optimize the overall shape of the eye. This may be enough to fix your SI problem if your error tails are normally distributed, but you won't be able to catch long tails in the distribution. (Of course, the only way to be sure is to tweak your system and see if the problem goes away.)

I have one very basic question: Is the Gigawave the right tool for looking into high speed digital signals? I'm not only thinking about LVDS, but also USB 3.x, Gigabit Ethernet etc. I have often situations where I need to tune transmission lines, very long or non-ideal cables and linear redrivers and each time I have to go back to the HF lab at my old university to use the nice, but absolutely non portable and expensive 40GSPS LeCroy WaveRunner.
Given that you have access to a 40 GS/s WaveRunner, the GigaWave could be a convenient tool to "catch" most of your day-to-day needs. In your situation, it doesn't need to do everything - it just needs to do enough to cut down on your trips to the HF lab. USB 3.x @ 5 Gbaud is a bit too fast for accurate results on the GigaWave, but Gigabit Ethernet should work fine.
SJL Instruments | San Jose, CA, USA
Pocket-Sized 6 GHz 1 TS/s ET Sampling Oscilloscopes
https://sjl-instruments.com
 

Offline KE5FX

  • Super Contributor
  • ***
  • Posts: 2013
  • Country: us
    • KE5FX.COM
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #14 on: August 01, 2024, 11:19:13 pm »
Just found this thread - some responses below, and happy to answer any other questions you might have!

All EU orders ship for free with duties paid by us (DDP). The price shown is VAT-inclusive and will be the total cost to you.

That's very generous!  I once made the mistake of shipping a TimePod to the UK without checking the 'Bill Recipient' option for duties and taxes on the FedEx airbill.  As I recall, it cost me about 15% of the sales price, or almost $1000 in total.   

That was over 10 years ago, but I can't imagine that the rates have gone down since then.
 

Offline SJL-Instruments

  • Regular Contributor
  • *
  • Posts: 217
  • Country: us
    • SJL Instruments
Re: "Cheapest" Equipment for measuring + optimizing 2.5 Gbps LVDS
« Reply #15 on: August 01, 2024, 11:35:36 pm »
That's very generous!  I once made the mistake of shipping a TimePod to the UK without checking the 'Bill Recipient' option for duties and taxes on the FedEx airbill.  As I recall, it cost me about 15% of the sales price, or almost $1000 in total.   

That was over 10 years ago, but I can't imagine that the rates have gone down since then.

Thanks - we want the process to be as smooth as possible. Purchase the product for the advertised amount, and have it show up at your door, zero surprises involved.  :)

By the way, in situations like the OP's, we have sometimes handed out free evaluation units for up to 30 days. At the end of 30 days, you can either return it or purchase the unit.
This works well within the USA, but we have never done this in the EU. If you (or anyone else) has some insight onto how much of a hassle this is in terms of EU customs, we may be open to doing this for luky315 as well.

(Slight word of warning: nobody has ever decided to return the unit. You may as well buy it.  ;D)
« Last Edit: August 01, 2024, 11:39:17 pm by SJL-Instruments »
SJL Instruments | San Jose, CA, USA
Pocket-Sized 6 GHz 1 TS/s ET Sampling Oscilloscopes
https://sjl-instruments.com
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf