new killer scope in town - a true game changer from R&S - RTB2002 & RTB2004

[Xyphro]

@taotech: You can PM me. That thread relates to R&S device.

[Xyphro]

It seems that TI part was at least used also in HMO scope variants:

https://youtu.be/3THvWEKGOwE?si=uFzqoWlHOyP6YzRa

Visible at ~4:00

Though from the video it is not visible which % of full scale range they dimensioned the frontend to.

[tautech]

@taotech: You can PM me. That thread relates to R&S device.

I've taken this here:
https://www.eevblog.com/forum/testgear/siglent-technical-support-join-in-eevblog/msg6047897/#msg6047897

[Hydron]

No heatsink on the PGA, but it was a bit of a pain to get off due to the thermal vias to the board ground plane. Pre heating would certainly have helped a lot. No concerns about calibration after swapping - this is all handled by the self alignment process of the scope. I'm very interested in whether running full scale low will fix it again!

Edit: looked at the datasheet, while it's from 2008 it has had a revision a little over a year ago. Still says 1V input max in abs max notes though. They have also removed the "Revision History" section I have in an older copy of the datasheet Seems like there's only ONE mention of 400mV (0.8 Vpp) input limit in either datasheet version, and it's mentioned only in passing in the applications section - totally reasonable to assume that 1V is OK when reading this steaming pile!

[edavid]

The current LMH6518 datasheet does have this note:

"When the LMH6518 output is held at saturation conditions for long time periods the part can develop a permanent output offset
voltage. To manage this output offset condition the device attenuation must be set properly to avoid long periods of output saturation."

This is a little different from what was described before, so I guess they did investigate.

https://www.ti.com/lit/ds/symlink/lmh6518.pdf

[nctnico]

Well, this looks like an issue which can be prevented through software then.

[rf-messkopf]

Actually I would consider a Magnovia, but only if I could try it risk free for a month or so at home. From playing at a booth for a few minutes does not give you the same experience as using it really in lab and the investment is for hobby money to big for risks.

Batronix offers free demo units and additionally a 30-day right of return, even for business customers: https://www.batronix.com/versand/verkauf/Batronix.html#02.

[Hydron]

The current LMH6518 datasheet does have this note:

"When the LMH6518 output is held at saturation conditions for long time periods the part can develop a permanent output offset
voltage. To manage this output offset condition the device attenuation must be set properly to avoid long periods of output saturation."

This is a little different from what was described before, so I guess they did investigate.

https://www.ti.com/lit/ds/symlink/lmh6518.pdf

Thanks, I only had time to skim it and missed that. Still a total BS move not making it more clear though! I.e. what is "long periods" in this context???
I concentrated on the input stuff as that's what it sounded like from the prior reports was the issue.

[Xyphro]

Those kind of aging related things should typically end up in abs. max ratings or recommended operating conditions and not as a fluffy pretty undefined sidenote in the flow text sections.

[Hydron]

Those kind of aging related things should typically end up in abs. max ratings or recommended operating conditions and not as a fluffy pretty undefined sidenote in the flow text sections.

It should ALSO be listed in a separate errata doc given that it's not in the orignal datasheet, at least that way there's a chance that you'd catch it in an existing product.

[Darkover]

Well, this looks like an issue which can be prevented through software then.

Of course, that is the minimum, because without a software that prevent from it, you will alway have a bad feeling when measuring with DC level.

Perhaps we need a kind of egg timer that measure the DC over 400mV time and switch on a screensaver.

Oh...and it is time dependend, so I guess the main reason is heat. I have a aircon in my workshop, so never above 25degree. It is probably bader if you living and measuring in a country with 35 degree room temperature.

I am still waiting for the answer from Batroniks, perhaps they have to discuss a little with R&S , so I can not open my RTB now, but I become curious how it looks inside, perhaps there is a way to optimize the cooling of the LMH.

TI wrote me they need a day to investigate it....

Olaf

[Darkover]

Those kind of aging related things should typically end up in abs. max ratings or recommended operating conditions and not as a fluffy pretty undefined sidenote in the flow text sections.

Of course! But I think the salespeople would feel bad of it.  With the assumption that this PGA is in many brands of scopes they can not feel happy to say it to there customer.
Perhaps it is also important to have a very good and massive ground layer to remove the heat. Yesterday I have seen somewhere the layout of TIs demoboard. That surprised me by having small ground for a device with 1W.

Perhaps it is better for you to buy Rigol because of there infamous noisy fan. 

Olaf

 

[Xyphro]

They certainly will.

In case you end up replacing the frontend ICs yourself after the warantee stuff is cleared out: Can you first solder wires at the PGAs output and measure the voltage at the differential output voltage in your suspected case that triggered the drift?

Would be really curious to see if that one is "in saturation".

About fixing it in SW: 2 approached are possible to improve but not fully fix it:
- improved offset compensation in digital domain if analog one is in saturation already. That one has side effects on complicating trigger, dynamic range, etc...
- or a statement in scopes manual that per Vertical setting another max voltage rating applies. SW can also detect such overload conditions and signal it like on the "expensive big boy" R&S scopes where they do it for other reasons (e.g. affecting frequency domain content).

[salomonander]

Is there a way to have some kind of averaging the counter function? Im trying to tune an old synth and the counter is just a little too nervous. Thanks!

[salomonander]

I went totally nuts today trying to tune this synth with the counter. All of a sudden nothing stayed in tune and the frequency shifted around. Took me hours to find out that the counter gets instable when the waveform is scaled to a certain size. Im so confused. When the waveform fills only three squares from the middle line, the counter stays pretty steady. Now when i use the vertical scale knob and zoom in, the counter gets unsteady and does occasional shifts. Im wondering if the unit has an error or if i do not understand something fundamental. Any ideas?

[KungFuJosh]

Check your trigger level, vertical resolution usage, etc. Share some screenshots too so we can see what you see.

[salomonander]

Thanks, i think i just solved it. User error obviously. Stupid me had the probe on 10:1 and the synth has low output. So i picked up interference that made the counter go nuts i guess. Id still love to find an averaging feature for the counter….

[KungFuJosh]

I don't have that scope, so I would need to RTFM to see if it can. I assume it can. My Siglent & Magnova scopes can. If not directly from the main counter, then check the measurement functions.

[RBBVNL9]

Thanks, i think i just solved it. User error obviously. Stupid me

No worries, happens to us all 

[salomonander]

Thanks, still learning just wondering: would a differential probe help suppress interference in general?

[2N3055]

Thanks, still learning just wondering: would a differential probe help suppress interference in general?

Do you use standard passive probe?
If you do, do you use grounding wire with crocodile clip?
That piece of wire acts as antenna.
If you have it, try using grounding spring...

[salomonander]

Yes, i use the standard probes that came with the unit. And also yes, i use the crocodile clip. Thanks for the tip, i was not aware of this. Ill try to get a spring.

[Darkover]

I've never had this sort of problem with any other probes or cables for any device or equipment.  Has anyone else ever experienced this, silicon leaking?  While the probes seem to function fine, not too crazy about the situation.  Can anyone recommend a different set of probes that would work well with this scope. 

No, I have never seen this beavior myself or read about it. I checked it after I read your words. And may probes are still in there original bags and not used more than 1-2times, because I did not like the original probes. I think they are to huge for most of my work. I am using PML211 from PMK or PML711. I searched for them, but can not find the PML211 on PMKs homepage anymore. Only probes with 500Mhz, probably more expensive. 

https://www.eevblog.com/forum/index.php?action=post;quote=5452322;topic=84856.3250

The reason why I like this probes so much is that the tip is springy. That made the use ten times easier when using the ground spring.

Olaf

[Darkover]

Some new information.

I tried to "repair" the offset problem bei setting GND to the top with 100mV/div and a voltage of -800mV for one hour, the offset did not change, so I tried another hour with -2V. It did not change anything.
So it is not possible to repair it in this way.

I also tried Setup/Selfalignment, but this gives an error because the offset is to huge. (see pic)

So I think the only solution is to replace the PGA. They are ordered and I will write about it in 2-3weeks.

Olaf

[Hydron]

Olaf - try FW 2.000 or earlier while you wait for the parts (see my post here: https://www.eevblog.com/forum/testgear/new-killer-scope-a-true-game-changer-from-rs-rtb2002-rtb2004/msg3246780/?topicseen#msg3246780). Ping me if you need a copy.

If you're out of warranty then yes just fix yourself - R&S repair cost is a fucking joke, more than I paid for the unit and to top it off, scaled by what BW licenses you have installed (what the hell does that have to do with repair costs??).