Siglent SDS1000(X) / SDS2000(X) or LeCroy WaveAce (Blackfin) hacks?

[tautech]

here is my HW mod on 2104x, this changed the 3db limit from around 250Mhz to 350Mhz

That's sounds about right for a 2304X -3dB point.
A screenshot I've posted before on the forum plus an additional few for reference.
HP 500 MHz sig gen used for all these screenshots.
Factory SDS2304X, 100mV signal:




200 and 300 MHz shots below = 1V p-p signal
It can be seen in the 500 MHz screenshot (100mV p-p) is past the stability frequency of the internal counter.

[markus_jlrb]

@tautech

could you please provide a sweep picture, if your sig gen is
capable of this feature to see the flatness of the BW mod
when the caps are removed.

Thanks
Markus

[tautech]

@tautech

could you please provide a sweep picture, if your sig gen is
capable of this feature to see the flatness of the BW mod
when the caps are removed.

Thanks
Markus

Sorry I can't as I don't have an AWG that can sweep to high frequencies.
ATM 30 MHz is the best I have.

You have SSA don't you ? Maybe that will do.

[markus_jlrb]

The caps are mounted between the pos and neg output
(OPHI/OPLO) of the AD8370 var gain amp.
Both lines are have round  50 ohm impedance according
to DS (100Ohm diff).

Have made a LT-Spice simulation to check the influence
 of the cap in parallel to the output (Pin 8/9)

[markus_jlrb]

The LT-Spice file is attached too.
To play arround, remove the .txt suffix.

The 0.95(V(n002)-V(n004) 5% recuction
is due to avoide overlapping of the red
curve.

bouth sig gens have 1V amplitude so that the
difference is 2V and the corresponding voltage
at load (ADC) should be 1V without attenuation
and influence of the cap.

As I do not know the ADC input impadance, I took
the same as the gain amp has in the hope the disign
is matching the impedances 1:1.

The C value is sweeped from 1pF to 10pF to visualize
the cap influence - BW reduction.

Markus

PS.: The simulation was made for 200MHz as I would
        like to compare with my DSO.

[markus_jlrb]

FYI,

I can confirm the mod described by tautech extend the BW above 300MHz.
For my model SDD2204X only one cap is mounted nearby the AD8370 ic
in front of pin 8 and 9. This one cap I had desoldered today.
The 3dB (voltage) BW is located at 330MHz.
10dB BW reach the 70cm ham band at 440MHz. (50mV/div) of a 1Vpp
signal.

I have attached some pictures measured with my DSO.

Markus

PS.: The 500MHz signal is passing a 3dB splitter to a second scope
to compare the sigs. Means that the 316mV from the SMY01 has to be
devided by 0.707.

[tautech]

FYI,

I can confirm the mod described by tautech extend the BW above 300MHz.

FYI
So there is no confusion, the waveforms I have posted are from factory SDS2304X, unmodified in any way.
The signal amplitude/s in screenshots are either 1V or 100mV p-p so roll off can easily be seen.
The screenshots were not taken in one session.

[markus_jlrb]

Sorry tautech,

It was the post from KeBeNe and his mod.
But good to know that the dso BW is comparable
with the factory tuned.

Markus

[tv84]

The 3dB (voltage) BW is located at 330MHz.
10dB BW reach the 70cm ham band at 440MHz. (50mV/div) of a 1Vpp
signal.

Cool!

And, if I'm successfull in my investigations, soon you'll all be able to reclassify the scopes to the SDS2304X category (to match it's newer HW capabilities).

I think it's just a magic byte that needs to be changed...

[KeBeNe]

Here is my measurement with 100mV, -3db point and 500Mhz.


@ tv84, that would be great and then all the options, have all bought up on PAS, I would like to have them

[tv84]

All the "investigations" talked about in this thread should be applicable to the Lecroy models:

WA102
WaveAce 1001  40MHz
WaveAce 1002  60MHz
WaveAce 1012 100MHz
WA204
WaveAce 2014 100MHz
WaveAce 2024 200MHz
WaveAce 2034 300MHz
WaveAce 2004  70MHz
WA202
WaveAce 2012 100MHz
WaveAce 2022 200MHz
WaveAce 2032 300MHz
WaveAce 2002  70MHz

[markus_jlrb]

@KeBeNe

>Here is my measurement with 100mV, -3db point and 500Mhz.

Should be much more then -3dB from 300MHz to 500MHz.
(or did you other additional mods besides the cap removal?)

My scope (modified) have around 17dB att. from 300MHz to 500MHz.
The -3dB point is located around 325MHz.

The gen. provides 316mV/eff. Sig was splitted by a resistive  power
devider ( att > -6dB), so at scope input the amplidude @500MHz is:
316mV * sqrt(2) / 2 = 223mV (amplitude) and the scope shows 30mV
(amplitude) ==> 17,4dB attenuation. (a faktor of 7.4 concerning voltage
from sig to scope displayed)

Hopefully not making a mistake in my calculation ;-)

Markus

[tv84]

Markus,

Show us a freq. sweep of the new config.

[markus_jlrb]

 As requested,

the bw picture from 10MHz to 500MHz.
Zero line shifted down to get a better resolution.

Markus

PS.: Looking for marker hints. Perhaps a second channel
is uesfull to do this to mark the -3dB point.

[KeBeNe]

here a sweep from 1 to 800Mhz from my scope

[markus_jlrb]

@KeBeNe,

looks very ipressive, -3.6dB at 460MHz.
But I wonder about the increased voltage
level at low frequencies.
You have terminated the dso input with 50
Ohm properly?
Sorry that I ask. It's a big difference if I compare
with my scope. Did you make other mods?

Btw what is your smallest time base value.
1ns or less?

Markus

[KeBeNe]

So, the SMT06 is no longer in ALC mode at 1V, the levels are no longer true. I did not pay attention.

We should agree on measurement levels and settings on the SDS (setup file from the SDS).

For the measurements, I activate the internal 50ohm resistor.
1ns is my smallest time.

Other modifications than the 4 capacitors removed I have not done, maybe there is still a calibration file, that would be an explanation

[markus_jlrb]

@KeBeNe

I have modified your pic with some markers.
Now It's look for my more realistic.

Thanks for your effort!

Markus

[markus_jlrb]

In addition to this BW picture, it would be of
interest if someone could measure the input
impedance of the scope channels when
terminated (internally) with 50Ohm up to
150% of the BW - let say to 500MHz.
The theoretical BW according to Nyquist
with 2Gs/sec is 1000MHz. Taking the
drop of a aliasing filter 600 to 700 MHz
in BW should be possible.
The AD83700 is specified -3dB@750MHz.
I compare this with the Lecroy models
with 2Gs compatible with the SDS230xX
that are  specified up to 400++MHz.

Markus

 

[markus_jlrb]

FYI,

I have made the effort and check the reflection on ch1 to ch4
om my SDS2204X model. Looks not very good according to my
opinion. 10dB reflection from 50 Ohm terminated port/channel
means (>3% voltage err or 10% power error) error in maesurement
- right?

I had check the spawn from 0 to 1000MHz.
At the point of 100mV/div coming from the top (10,5,2,1,0.5,0.2V)
the reflection change due to internal switching of components.
All pictures show 50 Ohm terminated channels beside one that
I measure with 50 Ohm disabled (1Meg).

So I present screenshots of my measuring device as well as swr
bridge and cabling and the dso screenshots.

Waiting for your comments.

Markus

Sorry pic. #7 should be labeled with CH3 and not CH1.

 

[markus_jlrb]

To compare with a proper input channel impedance,
I check a  old Hameg 350 MHz DSO.

See the difference in the attached pictures compared
with the Siglent scope.
But also notice the decrease of impedance match,
when switching to 50mV and lower amplification.

Would be interesting to figure out how the SDS2KX input
impedance could be  improved.

Markus

[KeBeNe]

Here are the measurements of my 2104x


part 1:

[KeBeNe]

part 2:

[tv84]

Since the 300 MHz HW patch has proven to be successfull (with light differences between different PCB boards) on a SDS2104X and a SDS2204X, the only thing missing was my contribution.

Attached is the latest FW version (SDS2000x_1.2.2.2R10) patched to fully behave as a 300 MHz scope (the official Siglent SDS2304X/SDS2302X).  It's a 1-byte patch.

The .ADS file is zipped and splitted in 3 parts (.ZIP + .Z01 + .Z02). Place them in the same dir and extract.

The CRC32 of the .ADS must be 0x5AD47123.

The file has been tested in a SDS2204X and should work on all scopes that accept the official Siglent v1.2.2.2R10 FW (with CRC32 = 0xFBD42874). It's a simple runtime patch. If someone wants to test a permanent patch pelase contact me.

Edit: New file released Sep 25th.

[KeBeNe]

Hello,

how do I unpack the files correctly, get errors with 7zip