Broke my TDS3000B

[coromonadalix]

 

[Jarek]

I did just like SICCO.
My TDS3012B now works.
I used ADG361 IC's in version C not D. I bought it on Ebay once.
Maybe the original IC's work without this modification?

[ragge]

Hello,

Just got my hands on a TDS3054B, but think it is a bit noisier than I would have expected, see below for examples with all ports unconnected, with plain sampling mode and with envelope mode, and varying sample speeds.

Is this the expected noise level?
If not, does anyone have any good pointer for where to start looking for problems?

Thanks for any hints!

[sicco]

unlikely a hardware problem.

Is it as bad when selecting GND instead of AC/DC? If no then you’re looking at thermal noise in the 1Mohm input resistor.
Try to limit the input bandwidth from 500 MHz to a lower settings: does that improve things?

You did run SPC and that passed ok?

Look at the spectrum (option Math, FFT): is it white noise, pink, or some distinct peaks? You may be sitting in a noisy environment, nearby radio transmitters, cell phone towers, WiFi APs, overhead power lines, VFDs and so on…

[voltsandjolts]

vid envelope mode

Is this the expected noise level?

Erm, envelope mode?

[ragge]

unlikely a hardware problem.

Is it as bad when selecting GND instead of AC/DC? If no then you’re looking at thermal noise in the 1Mohm input resistor.
Try to limit the input bandwidth from 500 MHz to a lower settings: does that improve things?

You did run SPC and that passed ok?

Look at the spectrum (option Math, FFT): is it white noise, pink, or some distinct peaks? You may be sitting in a noisy environment, nearby radio transmitters, cell phone towers, WiFi APs, overhead power lines, VFDs and so on…

Hello sicco!

AC/DC/GND, or 1M/50 ohm, doesn't matter really, it is more or less the same in all modes.

With Envelope acquisition mode, and measuring Pk-Pk, on channel 1, 1.00 mV/div, DC-coupling, it after some 15 seconds gets:
Full Bandwidth: 1.60 mV
Bandwidth 150 MHz: 1.40 mV
Bandwidth 20 MHz: 860 µV

Yes, the SPC has passed 6 times or so. At one occasion it didn't and said it needed a repair, but when just running it again everything was OK.

I would call the FFT pink (see attached picture), with the strongest peaks on channel 2 for some reason, at worst about -100 dB, peaks at 220 MHz, 300 MHz, 525 MHz, 600 MHz, 680 MHz, 750 MHz, 825 MHz, 900 MHz, 1.14 GHz.
Actually, the peaks get attenuated when the channel is set to GND instead of DC or AC, more attenuation at higher frequencies (see second picture).

It could be a noisy environment, but I haven't seen this problems on other scopes on the same place (TDS6xx, TDS7xx, Rigol DS1104Z, ...).

Thanks for your help and advice, much appreciated!

[ragge]

Erm, envelope mode?

Sorry, that was not clear: Acquisition Mode = Envelope, and with infinite persistence (see pictures).

[sicco]

I checked my TDS3054, TDS3054b and TDS3034c, and all behave pretty much the same: ~1 y-scale DIV lights up with infinite envelope, 1 mV/DIV full BW, GND.
When I look at the spectrum I think I see genuine radio transmitters. If I hold a 868 MHz LoRa radio nearby, then I get to see a 868 MHz peak...

So I'd say your scope behaves pretty much as good/bad as a same model brand new one would.

Digging a bit deeper, with input coupling at GND, 1 mV/DIV, Averaging at 512, FFT, 2.5 Gs/s one starts to see this series of peaks: 75 MHz, 151 MHz, 225 MHz, 300 MHz, 375 MHz, 451 MHz. The TDS3000 system clock comes from a crystal that is 75.75 MHz. So that's what we're seeing as the noise here I think - besides the plain noise floor that shows ~20 dB lower. These peaks do stay with coupling at GND, whereas the external radio transmitter peaks do disappear.

These spectra look like this on my TDS3054b, c and -no suffix. What's interesting though is that the models do behave differently w.r.t. which 'square wave' overtone gets a high peak vs which is lower.

PS you do have the metal shield decently covering the inputs / attenuator section? As in that metal cap is not missing. And the -b and -c model extra flat metal piece bridges to the mains input GND is screwed in tightly?

[todopapa]

Hi,perdrix-san
Thank you for this schematics regarding in Main board.
I had a white screen TDS3014B, it sometimes went to blank screen when starting up.
But this time it had completely dead.
In the blog that TDS3014 adventure, the 75.7575MHz clock is suspicious.
Therefore I broke the scope and check the clock, I found 3.3V line is not appear with the clock module.
the display is white so 15V power line is alive.
Then, I digged into main board power line, U900 3.3V LDO IC pin throughhole solder was not adequate.
When I touch it,  it sounds starting up clicking relay.
I just did additional solder on the LDO pins.
I guess that LDO is not fixed on the PCB, therefore it seems that vibration caused cracks in the solder, causing this malfunction.
Thank you for your this informative post.
Atatched photo is after re-soldering the LDO.

[todopapa]

Hello, there,
Unfortunately the scope had been in white display again in a few days.
Therefore I've checked inside the TDS3014b again, this time Y700 48.000MHz near the U900 3.3V LDO is OK, but Y500 75.7575MHz Vdd was just 1.2V.
Something wrong happened with Y500 Vdd.
I have checked Y500 Vdd to U900 3.3V output resistance, this time it was 0.9MΩ.
When I fixed the scope previously, it was 12.1Ω. It was pretty high but at that time I guess it was caused by EMI filter or like.
I've checked nearby IC U560 MT58L128LIBF Vdd from LDO 3.3V output, that is about 0Ω. and  Y500 75.7575MHz Vdd(pin8) and enable(pin1) was ０Ω. Ummm, PCB pattern is disconnected ? or soldering problem?
Anyway, I decided to connect directly from LDO 3.3V output (C905 tantalum cap +).
Before soldering, I've checked current to Vdd to  LDO 3.3V with multimater, that was 14.8mA and the scope relay sounded working!

The real cause is PWB pattern maybe disconnected inside or soldering matter.
But I'm very grad to survive this scope  from death  twice again!
Best regards,