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

[theatrus]

Just a heads up for those who want a few more probe accessories than the limited selection supplied with the RT-ZP03 these scopes come with.

The LeCroy "PKIT3-5MM-101" is completely compatible with the R&S probe (I assume they use the same OEM), and is available very cheaply from RS UK:
(note that the spring hook grabber thing is a slightly different design but is still compatible)

This is a nice little kit, especially for spares of the most vulnerable parts, the BNC adaptor, low inductance ground spring and the RED (not pink) channel marker cable clip!

Seems that the US (i.e. non-RS) price is much higher, but I think they got the better end of the deal with $2k scopes.

I didn't realize the GBP and USD deviated so much, since that part number runs $45 on the US distributors 

[Robaroni]

More scopes on eBay if anyone is intertested:

http://www.ebay.com/itm/Rohde-Schwarz-RTB2004-300MHz-4-16-Channel-Oscilloscope-All-Opt-R-S-RTB2K-COM4-/152536506153?hash=item2383e25f29:g:NawAAOSwX61ZC0bH

http://www.ebay.com/itm/New-Rohde-Schwarz-RTB2004-300MHz-4-16-Channel-MSO-Full-Options-R-S-RTB2K-COM4-/232323512044?hash=item36178f76ec:g:14UAAOSw~y9ZCjUM

[TK]

More scopes on eBay if anyone is intertested:

http://www.ebay.com/itm/Rohde-Schwarz-RTB2004-300MHz-4-16-Channel-Oscilloscope-All-Opt-R-S-RTB2K-COM4-/152536506153?hash=item2383e25f29:g:NawAAOSwX61ZC0bH

http://www.ebay.com/itm/New-Rohde-Schwarz-RTB2004-300MHz-4-16-Channel-MSO-Full-Options-R-S-RTB2K-COM4-/232323512044?hash=item36178f76ec:g:14UAAOSw~y9ZCjUM

First one already sold...

[Robaroni]

More scopes on eBay if anyone is intertested:

http://www.ebay.com/itm/Rohde-Schwarz-RTB2004-300MHz-4-16-Channel-Oscilloscope-All-Opt-R-S-RTB2K-COM4-/152536506153?hash=item2383e25f29:g:NawAAOSwX61ZC0bH

http://www.ebay.com/itm/New-Rohde-Schwarz-RTB2004-300MHz-4-16-Channel-MSO-Full-Options-R-S-RTB2K-COM4-/232323512044?hash=item36178f76ec:g:14UAAOSw~y9ZCjUM

First one already sold...

Wow, that was quick!

[Hydron]

I've just had a play with the screen protector that Mike linked too - I can confirm that it works well, especially if you make sure everything is extremely clean (especially dust free) and put it on behind the front cover (looks like it is original). That said, I've taken it off again, as I like the crispness of the shiny screen and don't have it in an position where reflections are an issue.

While I had the cover off I also had a look at the encoders, and purchased a similar ALPS EC11J series part to investigate removing the detents, which are still one of the most irritating things about this unit. Unfortunately a stock part is not available without detents, and it also seems that R&S is using a semi-custom model (ALPS don't show the combination of shaft length and style on their product list). Having pulled the similar encoder apart, I'm pretty confident that the detents can be removed from the original R&S parts without destroying them, with the following process:

- buy a spare encoder, remove the outer metal "shell" carefully by un-bending the tabs underneath. Disassemble this one to get an idea of what's inside before working on the real thing (have a go at removing the detent spring too)

- carefully cut the shell of the R&S part on the pcb, as shown in the picture, to separate it into 6 individually soldered down parts (side cutters, possibly a small file could be useful here)

- unsolder each of these parts individually, paying special care to the 2 wide pieces (these have the tabs bent under the black plastic bit)

- pull the top assembly off (black plastic part, switch and contacts remain on the PCB), and disassemble the encoder to remove the detent spring (this is the only part which needs to be forcibly removed, everything else is just keyed to fit together and is held by the metal shell)

- re-assemble, and solder the "shell" from the spare encoder (with the bottom tabs removed) onto the PCB to replace the original one which was cut for easy removal. The removed tabs shouldn't matter as the black part that they retained is still held down by solder.

I'm not going to do this mod to my scope in the near term as I think it'd void the warranty pretty thoroughly, but I really don't like the detents so once I'm confident that the unit is working well I'll be seriously considering it. It's a real shame that R&S didn't ask ALPS to do a small customisation to the encoder to make all this unnecessary, but in their defence, the well-executed trace movements on the touch screen do make up for a lot of the annoyance.

[Robaroni]

I've just had a play with the screen protector that Mike linked too - I can confirm that it works well, especially if you make sure everything is extremely clean (especially dust free) and put it on behind the front cover (looks like it is original). That said, I've taken it off again, as I like the crispness of the shiny screen and don't have it in an position where reflections are an issue.

While I had the cover off I also had a look at the encoders, and purchased a similar ALPS EC11J series part to investigate removing the detents, which are still one of the most irritating things about this unit. Unfortunately a stock part is not available without detents, and it also seems that R&S is using a semi-custom model (ALPS don't show the combination of shaft length and style on their product list). Having pulled the similar encoder apart, I'm pretty confident that the detents can be removed from the original R&S parts without destroying them, with the following process:

- buy a spare encoder, remove the outer metal "shell" carefully by un-bending the tabs underneath. Disassemble this one to get an idea of what's inside before working on the real thing (have a go at removing the detent spring too)

- carefully cut the shell of the R&S part on the pcb, as shown in the picture, to separate it into 6 individually soldered down parts (side cutters, possibly a small file could be useful here)

- unsolder each of these parts individually, paying special care to the 2 wide pieces (these have the tabs bent under the black plastic bit)

- pull the top assembly off (black plastic part, switch and contacts remain on the PCB), and disassemble the encoder to remove the detent spring (this is the only part which needs to be forcibly removed, everything else is just keyed to fit together and is held by the metal shell)

- re-assemble, and solder the "shell" from the spare encoder (with the bottom tabs removed) onto the PCB to replace the original one which was cut for easy removal. The removed tabs shouldn't matter as the black part that they retained is still held down by solder.

I'm not going to do this mod to my scope in the near term as I think it'd void the warranty pretty thoroughly, but I really don't like the detents so once I'm confident that the unit is working well I'll be seriously considering it. It's a real shame that R&S didn't ask ALPS to do a small customisation to the encoder to make all this unnecessary, but in their defence, the well-executed trace movements on the touch screen do make up for a lot of the annoyance.

Hydron,
Are you sure that if you remove the detents the scope will work properly? I set up uC algorithms to run from encoders and there are several ways to divide encoder steps. R&S may have used detents to coordinate with changes so as to get positive results per each detent. What happens if you stop in the middle between detents? Flicker? I've seen and had it happen. Sometimes you can stop in between detents and you'll get a delay because the uC is expecting a quadrature change.
Rob

[mikeselectricstuff]

The other issue is how smoothly a non-detented encoder turns. If it's too stiff, it's little better than the detented version. A detented encoder doesn't need to have smooth motion.

Compare to the ones on the Keysight, which turn extremely freely, so can be operated with a very light touch ( though the knobs could possibly do with being a little bigger).

The other day I was alternating between the KS and R&S for various reasons and was really starting to hate the detents, in particular on the horizontal position knob.

[Hydron]

Both very good points, and are the reason why I got another encoder from the same family to play with before seriously considering messing with the R&S ones.

- The test encoder I pulled the detents from is now very nice, smooth and free to turn (can feel tiny variations during rotation, but WAY better than before the mod)

- I also worried about what Rob mentioned - will it go crazy if it stops just between sectors or bounces more than is expected by the controller.
Unfortunately this is harder to test - at the minimum it would require getting the existing encoder in a position where it's not pulling the sense lines to VCC/GND (depending on how it's wired up), and then attaching the test encoder in parallel.
If this isn't possible, then the upper half of the (detent-free) test encoder should be able to be substituted temporarily for the R&S part after removing the retaining shell, allowing for operation to be checked before modifying the un-obtainium R&S part.

[mikeselectricstuff]

There ought to be a position on the original encoder where both contacts are open ( maybe even the detented position), so holding it that position would allow another to be connected across it to test. 

[Hydron]

There ought to be a position on the original encoder where both contacts are open ( maybe even the detented position), so holding it that position would allow another to be connected across it to test.

Yeah that's what I was hoping/planning. Annoyingly I only realised it as I was typing my previously reply, not while the case was off the scope.
ALPS gives some info about the detent positions in the datasheet (snippet attached) - looks like one signal is guaranteed to be stable at the detent stop, but the other is not.

[piranha32]

I've got my scope today, and I haven't had much time to play with it, but I'd like to share my first impressions:
1. Detents: much more stiff than I expected, but after an hour of so playing with the scope I don't hate them yet. Maybe they're not as annoying as I expected after reading the thread?
2. Responsiveness: not top notch, but acceptable.
3. "Bang" sensitivity: almost as bad as seen in Dave's video, but I don't think it will be a problem. Light touches with a stick did not result in huge response. Using the touchscreen without banging on it did not produce noticeable response. I just need to be more careful when using at the highest sensitivity.
4. Remote operation: perfect. Although I managed to lock the scope after playing with usb/ethernet options in setup. Also, web server not always starts after the cable is unplugged for a short time. I had to restart the scope to be able to use the remote interface again.
5: Triggering: this is the biggest issue I've seen so far: is it possible to synchronize trigger with sweep of the generator? I haven't seen such option anywhere, and it would add a lot to the functionality.
6. Open ports: quick scan revealed that the scope responds on TCP ports 80, 111,  1024, 1025, and 5025. 80 is HTTP,  5025 SCPI. 111 should be rpcbind. Hm, does it mean that the scope export volumes over NFS?

[piranha32]

I've got my scope today, and I haven't had much time to play with it, but I'd like to share my first impressions:
1. Detents: much more stiff than I expected, but after an hour of so playing with the scope I don't hate them yet. Maybe they're not as annoying as I expected after reading the thread?
2. Responsiveness: not top notch, but acceptable.
3. "Bang" sensitivity: almost as bad as seen in Dave's video, but I don't think it will be a problem. Light touches with a stick did not result in huge response. Using the touchscreen without banging on it did not produce noticeable response. I just need to be more careful when using at the highest sensitivity.
4. Remote operation: perfect. Although I managed to lock the scope after playing with usb/ethernet options in setup. Also, web server not always starts after the cable is unplugged for a short time. I had to restart the scope to be able to use the remote interface again.
5: Triggering: this is the biggest issue I've seen so far: is it possible to synchronize trigger with sweep of the generator? I haven't seen such option anywhere, and it would add a lot to the functionality.
6. Open ports: quick scan revealed that the scope responds on TCP ports 80, 111,  1024, 1025, and 5025. 80 is HTTP,  5025 SCPI. 111 should be rpcbind. Hm, does it mean that the scope export volumes over NFS?

port 111 is also open on UDP, what could be a strong indication of running rpcbind, but rpcinfo can not connect. showmount does not show anything either. Bummer

EDIT: The scope also runs Bonjour. Here are services that I found in the advertisements:
http: 80/tcp
lxi: 80/tcp
scpi-raw: 1024/tcp
vxi: 5025/tcp

Ports 1025, 111, and one of UDP port (53686) still remain a mystery.

EDIT2: RPC responds, but seems to support only version 2 of the protocol. And does not list any services, not even portmapper.

[mikeselectricstuff]

5: Triggering: this is the biggest issue I've seen so far: is it possible to synchronize trigger with sweep of the generator? I haven't seen such option anywhere, and it would add a lot to the functionality.

Not currently possible - I did report this a while ago so hopefully it may be added.

[Robaroni]

There ought to be a position on the original encoder where both contacts are open ( maybe even the detented position), so holding it that position would allow another to be connected across it to test.

Yeah that's what I was hoping/planning. Annoyingly I only realised it as I was typing my previously reply, not while the case was off the scope.
ALPS gives some info about the detent positions in the datasheet (snippet attached) - looks like one signal is guaranteed to be stable at the detent stop, but the other is not.

I don't understand why test instruments use mechanical encoders instead of optical or magnetic. I've had encoders go bad in equipment. When my designs use encoders that get spun often I use optical, I only use mechanical when they are only occasional used, say, to adjust frequency in a precision oscillator every couple of months.

Most of these mechanical encoders get 15k cycles, I see the Alps used in the R&S gets 100k which is a lot better but it's also mechanical and that always seems to lower MTBF that I've seen.

Here's an interesting thing I've found, it looks like the detents, the little ball and indentation makes for shorter life, so maybe you're onto something removing them.

The data sheet for the Bourns EM14 optical encoder which I've used in several designs with good success shows 1M life cycle non-detent but look at the MTBF for detents!

Rob




 

[Hydron]

The detents aren't done by a little ball, it's a copper-coloured spring (I assume phosphor bronze) with 2 little ball shaped protrusions that pushes up against a plastic ring with little holes in it. There is grease, but it doesn't fill me with confidence regarding lifetime (though as you say, the ALPS parts have some good lifetime specs). Pic of the parts (sans spring, which I already lost!) attached. Note that the switch is a 1.5mm travel version rather than 0.5, but shouldn't make much difference.

Regarding whether detent removal is even going to work at all, I did a quick test with the hacked up encoder in parallel with one of the real ones (after finding a spot where both A and B switches were open):

It's not perfect, the feel is slightly uneven while turning (I assume because of the contacts moving between riding on plastic/metal), and the velocity sensing is pretty basic so can be a little jerky when running near the velocity threshold, but this is present anyway.

[Navarro]

I've just had a play with the screen protector that Mike linked too - I can confirm that it works well, especially if you make sure everything is extremely clean (especially dust free) and put it on behind the front cover (looks like it is original). That said, I've taken it off again, as I like the crispness of the shiny screen and don't have it in an position where reflections are an issue.

While I had the cover off I also had a look at the encoders, and purchased a similar ALPS EC11J series part to investigate removing the detents, which are still one of the most irritating things about this unit. Unfortunately a stock part is not available without detents, and it also seems that R&S is using a semi-custom model (ALPS don't show the combination of shaft length and style on their product list). Having pulled the similar encoder apart, I'm pretty confident that the detents can be removed from the original R&S parts without destroying them, with the following process:

- buy a spare encoder, remove the outer metal "shell" carefully by un-bending the tabs underneath. Disassemble this one to get an idea of what's inside before working on the real thing (have a go at removing the detent spring too)

- carefully cut the shell of the R&S part on the pcb, as shown in the picture, to separate it into 6 individually soldered down parts (side cutters, possibly a small file could be useful here)

- unsolder each of these parts individually, paying special care to the 2 wide pieces (these have the tabs bent under the black plastic bit)

- pull the top assembly off (black plastic part, switch and contacts remain on the PCB), and disassemble the encoder to remove the detent spring (this is the only part which needs to be forcibly removed, everything else is just keyed to fit together and is held by the metal shell)

- re-assemble, and solder the "shell" from the spare encoder (with the bottom tabs removed) onto the PCB to replace the original one which was cut for easy removal. The removed tabs shouldn't matter as the black part that they retained is still held down by solder.

I'm not going to do this mod to my scope in the near term as I think it'd void the warranty pretty thoroughly, but I really don't like the detents so once I'm confident that the unit is working well I'll be seriously considering it. It's a real shame that R&S didn't ask ALPS to do a small customisation to the encoder to make all this unnecessary, but in their defence, the well-executed trace movements on the touch screen do make up for a lot of the annoyance.

I did this on a Hameg Oscilloscope and did a video about the procedure.

[JPortici]

Just a heads up for those who want a few more probe accessories than the limited selection supplied with the RT-ZP03 these scopes come with.

The LeCroy "PKIT3-5MM-101" is completely compatible with the R&S probe (I assume they use the same OEM)

just so you know, the probes seem also to be the very same supplied with picoscopes

[Hydron]

I did this on a Hameg Oscilloscope and did a video about the procedure.

Thanks for the video - very useful to see the results of your similar procedure.
Can I ask you why you changed the "Scale" encoders? And which encoders did you replace them with? Thanks.

On another topic, it seems that the P0 probe comp output has a fairly fast risetime (<1.1ns) - useful for a quick check of the scope/probe system performance.

[JoHr]

It's not perfect, the feel is slightly uneven while turning (I assume because of the contacts moving between riding on plastic/metal), and the velocity sensing is pretty basic so can be a little jerky when running near the velocity threshold, but this is present anyway.

How is the click function running without detents? Any problems when clicking and unwanted rotaion?

[Hydron]

How is the click function running without detents? Any problems when clicking and unwanted rotaion?

Couldn't test that sorry - the test encoder I got had a longer stroke switch (1.5 vs 0.5mm), and I didn't connect up the switch anyway (just the 3 quadrature wires).
I don't anticipate it to be a big issue though (as it is on the rigol multifunction knob) - the click function centres the trace and worst case you'll only be a tiny amount off it with encoder movement after pressing the switch. Also the scope is "sticky" around the centre point, at least for horizontal and vertical movements.

[korlatos]

I just received my scope from Test Force in Ottawa and it works well.

Has anyone got a wireless mouse to work and if so which brand and model?

It works fine with a wired mouse, but I'd like to go wireless.

Thanks,

Stuart

The scope works fine with a Logitech V220 (M-RBS136) wireless mouse. I also tried three different USB 2.0  hubs (one Targus and two generic ones) and they all worked. I use the hub so I can attach both a wireless mouse and a USB flash drive.

[korlatos]

I measured the 300 MHz RTB2004 rise time as 940ps using a Tektronix 284 pulse generator (<= 70ps rise time) and a 50 Ohm Tektronix 011-0099-00 termination. I also measured the bandwidth and found that the -3dB point is very close to 350 MHz. The scope is still usable for relative measurements at 500 MHz, and it can still display a 1GHz signal, although very attenuated. The bandwidth was measured with a Marconi 2022A signal generator and the same Tektronix 50 Ohm termination; the Marconi output was set to 6 dBm (447 mV).

[agdr]

... it can still display a 1GHz signal...

Very good post!  I'm just curious if anyone knows what would cause just the bottom of the 1GHz wave to appear modulated like that.  I know this is 3x the rated frequency of the scope, just curious why the degradation wouldn't be symmetric vertically.  Shouldn't be aliasing if the wave is purely sinusoildal, given the 2.5Gsps sample rate.  The 500MHz wave looks symmetrical vertically.

[mikeselectricstuff]

... it can still display a 1GHz signal...

Very good post!  I'm just curious if anyone knows what would cause just the bottom of the 1GHz wave to appear modulated like that.  I know this is 3x the rated frequency of the scope, just curious why the degradation wouldn't be symmetric vertically.  Shouldn't be aliasing if the wave is purely sinusoildal, given the 2.5Gsps sample rate.  The 500MHz wave looks symmetrical vertically.

Standing wave on the cable perhaps?

[bktemp]

Could it be a result of the 2.5GS/s samplingrate and a not perfect interpolation filter? Every other period of the 1GHz waveform showing the artefact and the .5 from 2.5GS/s look suspicious to me.
I had a similar effect when using an audio DAC with internal oversampling filter very close to 0.5x samplerate (outputting >45kHz sinewave for a DAC running at 100kHz).