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

[richtoy]

Yesterday I asked in Munich for a new firmware release date for RTB2004 and today I got a fast response from the customer support center:
It was planned to give a new release in 2. quarter, but due to pandemic it is shifted now to autumn.

Best regards

[Kean]

I cannot establish a network connection to IP port 5025. With Telnet (Windows 7) I cannot establish a connection. With a self-written program I can create a network connection, but no data is received when sending e.g. *IDN?
The LXI connection over the network works without problems. Changing the port does not help, even rebooting the oscilloscope does not help.

Can anyone receive data from port 5025?

Works for me.  I can send *IDN? and get a response.
I used PuTTY in Raw mode.  It doesn't work in telnet mode, possibly due to problems with telnet protocol negotiation.

Edit to add: Also keep in mind that queries and responses are CR terminated, no LF - so your receiving program needs to handle that.

[Kean]

I have used my network setting with manual IP address again. Only gateway and DNS I set to 127.0.0.1.
After a reboot the connection works, for whatever reason. The termination character is LF (\x0A), is received and sent by the RTB2000.

Glad you got it working.  Sorry, you are right - it is LF not CR.

My RTB is using DHCP.  I just tapped on the network icon at top right of the screen to get the current IP address allocation to use with PuTTY.
I don't believe I've done anything special to make it work.  It is possibly the first time I've tried port 5025 with it, and it worked first go.
I often use Python and vxi11 library or USB serial to control other gear like my HMC8042, HMC8012, 34970A, TH7321, etc, but not yet with the RTB.

[Hydron]

I've just had a very odd problem today with my RTB2k, and was wondering if anyone else had seen it or had any suggestions.

While using the scope today I noticed an unexpected offset on one channel that appeared after fiddling with probe attenuation factors (I was setting a custom Amps/V ratio). It seemed to behave bizzarely, not scaling as expected with attenuation factor which suggested a software rather than hardware fault. Seeing this I did a "Secure erase" to try and reset everything (should probably have just done a "Factory Defaults" in the save/load setup menu, but I forgot where this was at the time).
After the secure erase, EVERY channel was displaying an offset, and I noticed that the "Zero Adjust" setting was set to -39mV. When I attempted to adjust this it would flick to +39mV and then change no further. No other value can be entered either by SCPI, onscreen or with the rotary control; every attempt returns a high or low out-of-range error. When probe attenuation ratios are changed this zero offset also changes, but seems to scale oddly as well - e.g. going from 1:1 to 10:1 increased it to 3.9V rather than 390mV. Self alignment failed with a "Info: Error zeroAdjLimit -0.039000V < 0.03V" message in the log file (there is another numeric onscreen error code I can grab if needed).

With no other obvious option I tried some older firmware versions - anything up to and including 2.000 works fine and even allows the zero adjust value to be changed as intended (self alignment works too). Unfortunately the same problem returns when upgrading back beyond 2.000 (other than the Zero Adjust value actually increasing to 44mV and staying there). The fact that previous firmware works (and I think the probe cal routine shows normal voltages too, though I didn't test that thoroughly) points pretty strongly to a software bug.

Rich, anyone, any ideas on a fix? Am I missing something obvious? Not being able to use recent FW is a massive issue!
I have no idea of warranty state, it was a US launch promo unit and I now have it in the UK, but I hope such an clear software issue is still covered by R&S, whether in a firmware fix (obviously the easiest) or otherwise.

Attached screenshot shows the problem with open inputs and a couple different attenuation ratios (and 2 channels flipped to the other offset polarity).

[Joel_l]

My scope is on 2.202 and I tried to get it to do something similar and could not. While playing I noticed that I had about 4mV of offset on channel 1 at 10mV/div. The zero adjust worked as expected to dial it out.

Seems your scope is in some odd state.

[Hydron]

OK so I did some further digging and it looks like this is a combination of hardware and software faults.

Looking at the self alignment log and comparing it to an older one that I was lucky to have kept, it seems that channel 3 does indeed have an offset issue - the DCVGA calibration section log shows a greatly increased offset voltage required to zero the VGA output when using it's "Low Gain" (10dB) pre-amp setting (plus 8dB output amp = 18dB gain shown in log file). The (new) required offset reaches about 5x that required for any other channel. There is no substantial change from the previous (good) log for V/div settings when the High Gain (30dB) mode is active.

The VGA is a LMH6518, with the datasheet spending quite a few graphs on the offset vs gain mode (giving 3 example units rather than just a "typical" graph), so it seems that it's something that needs to be carefully accounted for, hence I guess all the calibration steps used.

The VGA output offset going out of spec on one range certainly explains the first symptoms I saw (including the odd scaling of the offset and being present on some V/div settings but not others) and also accounts for the alignment log. What it doesn't account for is the massive difference between firmware versions.

2.000 and earlier firmware is quite happy to calibrate the increased offset out (it's still there if you do an alignment on the older fimware and check the log) without any obvious problem - the channel in question seems to work fine when I compared it to another one.

Later firmware on the other hand refuses to complete the alignment (with no differences in the log up until the zeroAdjLimit error) then does the bizzare fixed Zero Adjust thing on every channel, making it basically unusable. I think that I did attempt an alignment before the secure erase so I'm not completely sure what triggered the change from the single channel problem into all 4 playing up.

I'll have a go at seeing if I can this fixed under warranty, though it is 3.5 years old so I may need to rely on R&S's goodwill. If unsuccessful I at least have a probable angle on a hardware fix. That said, the firmware clearly shouldn't behave as it does when a calibration value is out of spec (but still in a usable range) - it should at least give a warning or something rather than leave the instrument unusable!

[norks]

I had a similar issue with my RTB2004 early this year and luckily it was still within warranty but getting dangerously close to expiration. This is what I sent to R&S describing the issue:

"Yesterday when I turned it on channel 2 had a permanent DC offset about ~1-2 divisions in height with nothing connected to the input. As I changed the vertical sensitivity the size of the offset remained fixed at that same ~1-2 divisions, it didn't shrink with larger V/div settings. I checked that it wasn't because of offset or position settings, the offset remained after restoring factory defaults. Strangely it would disappear at 2 specific vertical sensitivities. The vertical sensitivity range is divided into a high and low range separated by an audible relay click when you transition between them at 150 mV/div (1:1 probe ratio). At the highest sensitivity of those two ranges the offset would suddenly be gone. i.e. at the 1 and 2 mV/div settings and then again after the relay click at the 200 mV/div setting. In all other V/div settings the offset was there.

I was sure there was a hardware problem by then but I ran the self-alignment routine because I wanted to exhaust all potential fixes and I thought you would want to see the results. So I ran the self-alignment (after full warm-up) and then a 77 mV offset appeared on ALL channels. See photo "RTB2000 permanent offset.jpg". This time the offset did grow and shrink with changing vertical sensitivity, through the entire range. 77 mV appears to be the maximum of the "zero offset" setting per channel. If I try to adjust the number myself the setting instantly changes to -77 mV and doesn't respond to any further changes. Restoring factory defaults returns it to +77 mV. I power cycled the scope, including shutting of the hard switch on the back for several seconds and then tried running the self-alignment again and it failed with the message:

"Channel: 0
Alignment not possible.
Description: 201-59.0:6"

I tried self alignment again and it failed with the same message."

R&S support was very responsive but they wouldn't handle the warranty claim directly, I had to go through the vendor. After all was said and done they replaced it with a new unit.

Of note, the new unit had a 2020 cal sticker on it (none on my original) and the hardware revision numbers were the same as my first one.

[Hydron]

Thanks for the reply - sounds like the exact same issue as I have, even down to certain portions of each relay range working (It looks like the high-gain pre-amp setting is used for 1-5mV/div and then 200mV/div, with low-gain used everywhere else).

I didn't spend as long diagnosing the initial offset as you did - I was in the middle of an urgent debug task and wanted it fixed ASAP so I reached for self alignment/factory reset/secure erase before investigating fully (wish I'd played around a bit more, or even dumped the flash as described in a different topic). Note that I didn't have anything dangerous connected to the input of the problem channel - just a AM503 current probe "amplifier" which can only output a few hundred millivolts max.

As for warranty, I might be a bit stuck there, as noted I'm just _outside_ 3 years, and returning to the vendor would be an extremely slow and expensive process given they're in a different country.
I'll still see if I can get some local support, but it might be a job for the hot air gun and some crossed fingers (this also assumes the stupid all-channel-offset firmware bug will be resolved once the VGA is in-spec). Not blaming R&S for the VGA semi-failing (it's marketed for the exact use seen here and used by many manufacturers) but the way the self alignment dealt with it is crazy.

Edit: Are the black bits on the knobs a DIY addition btw? Don't match mine which are all grey.

[Rich@RohdeScopesUSA]

Thanks for the reply - sounds like the exact same issue as I have, even down to certain portions of each relay range working (It looks like the high-gain pre-amp setting is used for 1-5mV/div and then 200mV/div, with low-gain used everywhere else).

I didn't spend as long diagnosing the initial offset as you did - I was in the middle of an urgent debug task and wanted it fixed ASAP so I reached for self alignment/factory reset/secure erase before investigating fully (wish I'd played around a bit more, or even dumped the flash as described in a different topic). Note that I didn't have anything dangerous connected to the input of the problem channel - just a AM503 current probe "amplifier" which can only output a few hundred millivolts max.

As for warranty, I might be a bit stuck there, as noted I'm just _outside_ 3 years, and returning to the vendor would be an extremely slow and expensive process given they're in a different country.
I'll still see if I can get some local support, but it might be a job for the hot air gun and some crossed fingers (this also assumes the stupid all-channel-offset firmware bug will be resolved once the VGA is in-spec). Not blaming R&S for the VGA semi-failing (it's marketed for the exact use seen here and used by many manufacturers) but the way the self alignment dealt with it is crazy.

Edit: Are the black bits on the knobs a DIY addition btw? Don't match mine which are all grey.

Hi Hydron - I talked with the R&D team this morning and it sounds like there is for sure a hardware issue (there were also some changes around SW in newer revisions, but nothing that would cause what you're seeing).  I'll PM you to see if I can help on the warranty side.

-Rich

[norks]

I don't think I ever probed anything dangerous with the channel that failed either, the offset was just there on boot-up one day. It's very interesting that older firmware can bypass the issue and restore full control, I figured it was an inescapable hardware fault. I never thought to look into the self alignment log, I wish I had done that. It always helps to have another data point.

I'd also expect fixing the bad channel to fix the all-offset problem. My guess is the self-alignment routine never expected to see an offset it couldn't get rid of and so as an unintended consequence it puts an offset on all the channels to get them to match each other.

Yes, I added the black tape to the knobs. The tape has a rough grippy texture and so I don't have to apply so much finger pressure when turning the knobs. The smooth surface of the knobs combined with the turning resistance and detents gets a little fatiguing after a while. I liked the difference enough that I reapplied it to the replacement scope.

[norks]

Greetings All,

Just thought I'd share my recent service experience with Rohde & Schwarz when my RTB2004 suddenly died earlier this month. [...]
They arranged for prepaid FedEx shipping to their Maryland shop and total turnaround time was about 2 weeks. [...]

I just saw this. That's great you had such a trouble-free experience, I wish I had the same. My experience had some of the typical signs of big companies and the small-time customers that get a lower level of attention. When I had a warranty claim R&S support was very good responding to my issue. After they concluded it should be sent in and handed me off to R&S Service they told me this:

"I was able to pull this up and it looks like the unit was purchased through test equity.
You would need to go through them to have the unit sent in for service, as we are unable to set up accounts for a single person.
I hope this helps and I apologize for any inconvenience this may have caused."

My vendor was actually Tequipment so I dealt with them. RMA was approved and I had to ship the scope to Maryland on my dime. Deciding how to do that was stressful because if the scope never makes it there then I just lose my scope, end of story. As small as that chance is it's a scary possibility. So do I insure for the launch price? The current $3700 sale price? What if it arrives physically damaged by the shipper and there's a drawn out process that ends with R&S denying the warranty and the scope is back to regular price by then? I'm looking at shipping choices ranging from about $80 to $180. That's real money to someone like me. I end up opting for $3000 insurance and crossed my fingers.

I don't hear anything for a while. The R&S RMA status page shows service complete. R&S Service won't tell me any information other than it was already shipped (because Tequipment is their customer on file, not me). I find out from Tequipment that it was shipped to the address on the original order 3 years ago rather than my updated account address, and tracking showed it was delivered over a week ago. Did I mention it was stressful? Fortunately I still had a way to contact the people that live there now, and they were honest people, and it was only an hour away. In total it was about 5 1/2 weeks beginning to end. For their part Tequipment apologized and said they would review what happened, they were supposed to confirm my shipping address in writing. No, it didn't make me feel any better about it. Everyone makes mistakes, and for the agent that handled the RMA I'm sure it was a very easy mistake to make on autopilot, but man what a nightmare it could have been for me.

And now seeing that perhaps I didn't have to take on the risk and cost of shipping myself if I had dealt with the right people pours some salt in the wound. Or maybe it's a policy difference between the US and Canadian markets. I'm not blaming R&S for what happened, they didn't make any clear mistakes. But their choice to be a step removed had a contribution and made room for more human error in the process. And if they had been willing to deal with me directly there's greater chance the old address could have been caught before the new scope shipped.

[Robaroni]

I don't think I ever probed anything dangerous with the channel that failed either, the offset was just there on boot-up one day. It's very interesting that older firmware can bypass the issue and restore full control, I figured it was an inescapable hardware fault. I never thought to look into the self alignment log, I wish I had done that. It always helps to have another data point.

I'd also expect fixing the bad channel to fix the all-offset problem. My guess is the self-alignment routine never expected to see an offset it couldn't get rid of and so as an unintended consequence it puts an offset on all the channels to get them to match each other.

Yes, I added the black tape to the knobs. The tape has a rough grippy texture and so I don't have to apply so much finger pressure when turning the knobs. The smooth surface of the knobs combined with the turning resistance and detents gets a little fatiguing after a while. I liked the difference enough that I reapplied it to the replacement scope.

You know what? I had the same kind of response. I had a probe that went intermittent, R&S said no problem you don't have to send it back we'll take care of it. Never happened! Called back a couple of months later and got someone else, said who told you that? Never gave me an RMA, nothing. OK, it's just a probe but if I had a major problem I would have been more aggressive. Got better treatment from Tek, see who gets my next scope order (better probes too - a lot better!).

[Harjit]

A couple of feature requests that are very helpful when working with capacitors and inductors - esp. on power converters, motors, etc.:

1) Automatic time constant calculation
a) Select the channel.
b) Use horizontal and vertical cursors and it shows you the time constant (and voltage) assume the cursors are set at Vmin, Vmax, Tmin and Tmax.
c) Bonus show cross hair on the trace.

2) Calculate slope
a) Select the channel
b) Use horizontal and vertical cursors and it shows you the slope assume the cursors are set at Vmin, Vmax, Tmin and Tmax.

[glowman]

I don't think I ever probed anything dangerous with the channel that failed either, the offset was just there on boot-up one day. It's very interesting that older firmware can bypass the issue and restore full control, I figured it was an inescapable hardware fault. I never thought to look into the self alignment log, I wish I had done that. It always helps to have another data point.

I'd also expect fixing the bad channel to fix the all-offset problem. My guess is the self-alignment routine never expected to see an offset it couldn't get rid of and so as an unintended consequence it puts an offset on all the channels to get them to match each other.

Yes, I added the black tape to the knobs. The tape has a rough grippy texture and so I don't have to apply so much finger pressure when turning the knobs. The smooth surface of the knobs combined with the turning resistance and detents gets a little fatiguing after a while. I liked the difference enough that I reapplied it to the replacement scope.

You know what? I had the same kind of response. I had a probe that went intermittent, R&S said no problem you don't have to send it back we'll take care of it. Never happened! Called back a couple of months later and got someone else, said who told you that? Never gave me an RMA, nothing. OK, it's just a probe but if I had a major problem I would have been more aggressive. Got better treatment from Tek, see who gets my next scope order (better probes too - a lot better!).

I had the opposite of what you have experienced on R&S.  I also had a small problem on 1 of the probes.  It was just a loose connection of the x1,x10 switch.  While moving the switch from x10 to x1 prior to changing the probe settings, I noticed multiple jitter of the waveform i'm probing, indicating it has a loose connection.  I told R&S about it.  Without any hassle, he sent me a RT-ZP10 as a replacement since the standard probe was not available yet.  I chose to just settle for it since it is a 500mhz probe costing $440! (standard probe costs only $105). Asking if I have to send back my probe to them, they said to just keep it.  I even told him it was practically fixed after I sprayed it with contact cleaner.  Again, he told me to just keep the new one!  Wow, that's customer service at its finest.

I am from Philippines and bought my RTB2004 from R&S Philippines.  I was initially planning to ask my friend to buy it from USA and send it to me.  I'm glad I did not choose that option and chose to buy it locally.  Price and service had been excellent.  Even with the Bode Plot option which I'm entitled to get for free (from getting PK1 option) at that time, they were the one bugging me to give them my details so they can process it already.  It should have been me bugging them.

By the way, RT-ZP10 is a high quality probe with spring loaded tip.  It has "Made in Germany" printed along the cable sheath.

[1design]

The probes are a genuine issue, I had to fix 3 out of 4 probes, 2 of those have never been used.

Issues:

• Probe 1 had an intermittent cable fault 5 cm from the probe
• Probe 2 had an intermittent issue at the cable/probe connection being assembled incorrectly
• Probe 3 had an intermittent cable fault 10 cm from the scope connection

Now I have 4 probes, all with different cable lengths...but hey, at least they work now.

[Hydron]

Huh, certainly sounds like you got some bad luck (or a bad batch). Haven't had an issue with any of mine other than one of the spring clips (witch's hats) being a bit sticky at first.

The RTB2k probes are made by "Hong Kong Texas" (http://www.hktexas.hk/en/index.html), a few other manufacturers use them too for their low-mid range probe offerings. I suspect the higher end R&S ones are from PMK (https://www.pmk.de/en/products).

As for my offset issue, am waiting on R&S to see what they can do, but am almost completely certain it's the LMH6518, which thankfully is a $10 part and should be relatively easy to replace - if I end up needing to do so I'll post pics and results here.

[Robaroni]

Huh, certainly sounds like you got some bad luck (or a bad batch). Haven't had an issue with any of mine other than one of the spring clips (witch's hats) being a bit sticky at first.

The RTB2k probes are made by "Hong Kong Texas" (http://www.hktexas.hk/en/index.html), a few other manufacturers use them too for their low-mid range probe offerings. I suspect the higher end R&S ones are from PMK (https://www.pmk.de/en/products).

As for my offset issue, am waiting on R&S to see what they can do, but am almost completely certain it's the LMH6518, which thankfully is a $10 part and should be relatively easy to replace - if I end up needing to do so I'll post pics and results here.

They pale in comparison to Tek probes, even on the lower priced Tek scopes. My Tek scopes had a better overall build too.

[Weston]

I am working on FPGAs where I can put down a dedicated debug connector and I was tired of all the flying leads (and had some noise pickup from that) so I made an adapter PCB to break out the RTB2004 LA pod to a 0.05" connector.

Figured I might as well share it here in case anyone else would find it useful: https://github.com/westonb/utility-pcbs/tree/main/LA_Adapter

[bayjelly]

There's a new firmware update, 2.300.

[mikeselectricstuff]

There's a new firmware update, 2.300.

Maximum  channel  probe attenuation user factor expands  to 10M  for unit 'V'.

There has to be a story behind why that got added....

[Hydron]

Thanks for the heads up. Added option to beep on trigger (or a few other things) gets a thumbs up from me. Not a major update like some previous ones though.

[maginnovision]

Well I attempted the update and I'm moving into hour two of nothing happening.

[ppeterl]

Ooops scary shit, let us know how it turns out! I'll hold my update for a bit then...

--/Peter

[Micke]

Yes, upps... 
This week I have upgraded two RTB2004 (my private and the one at work) from 02.202 to 02.300, went very quick and without problems.
The FW updater looks to be safe, it seems to inspect and validate FW image (CRC check?) before doing actual update.
Hope you get it sorted!

[maginnovision]

  Yea, every other update went off without a hitch. 

Ok, I had a minute so I turned it off with the soft power button and turned it back on. Booted right into the firmware update page. Pressed update and it very quickly updated.  At least it worked even if it decided to try and die on me. It's my favorite scope here so I'd be pretty disappointed.