MSO5000 Bode Plot Capability: Is it Good Enough? [Many tips about a 1st scope]

[jasonRF]

But any 5000-series pico has 8 independent ADCs (I think that's why they can get to 16 bit resolution.. By stacking ADCs..)

Dave did a teardown of a 5000B scope a bunch of years ago (I think it was a 5443B), so I know that it has a single ADC chip that draws something like 1/2 Watt.  It is a very specialized chip, and they basically designed their scopes around it.  I suspect they use that same chip in the current 5000D series, unless it is obsolete.     

Regarding FRA4Picoscope, it works fine with my 2204a, the cheapest Picoscope in the current lineup.  For years that was my only scope and I used that app many times.  There are examples from at least two different 2000-series (including my 2204a) in the thread
https://www.eevblog.com/forum/testgear/picoscope-2000/

jason

Thanks for the information!

I was making reference to this forum post: https://www.picotech.com/support/topic14311.html
Particularly, the third post from above. Says that the 2204A has something that hinders the FRA app. Maybe it was solved in the successive software releases?

About the 5000 series, they advertise them as having eight (!) ADCs. Maybe it's just for the lastest 5000 series models?

One question about that little 10 MHz scope. How often did you hit the bandwidth limit in your work? For example, you were able to accurately observe transients up to which frequency?

Thanks!

That post about lack of support for the 2204a is from 2014.  There have been many improvements to FRA4Picoscope since then!

Regarding bandwidth, my particular unit has a -3dB bandwidth around 25 MHz give-or-take.  It also falls off relatively slowly, depending on which vertical scale setting you are in.  If I forget to use a decoupling cap on an op amp and it is oscillating at 20 MHz it is no problem to see it.   Likewise, when I have built discrete audio amplifiers that had transient induced instabilities (eg “fuzz” showing up after the falling edge during square-wave testing).   Bumping up the compensation capacitor to the next one in my drawer made the fuzz go away.  But if I had an emitter-follower with a parasitic oscillation at 150 MHz I would be blind to it.   Not that I think I have run into that, but increased bandwidth was the primary reason why I upgraded this year.   My review in the thread I linked has more info on my thoughts about the 2204a. 

Jason

[balnazzar]

My review in the thread I linked has more info on my thoughts about the 2204a. 

Jason

Reading right now, thanks!

[balnazzar]

Just as reference for people who comes reading this thread, here is an RLC circuit captured with the AD2, while the AWG makes a sweep 50-20K, with the FFT below. How cute

https://youtu.be/DvlKVv5lGa0

I still have to figure out why the non-peak frequencies do exhibit that pulsation.. It's interesting.. 

[nctnico]

Bottom line: chinese scopes tend to be noisy (and in the case of the MSO5K, not only acoustically...). That's what I don't like about chinese products in general. They tend to deliver stuff that works, but they are always kind of half-finished. How much more would cost to be a bit more attentive to the acoustical noise? Almost nothing. A better fan, PWM control, maybe make the scope 1 cm thicker and put a 120mm fan on the back, exhausting from the side vents..

At work my office door opens into an electronics lab packed with high-end gear made by non-Chinese companies (Tek, Agilent, R&S, Anritsu, Keysight, etc.) and it is loud enough that I need to keep my door shut in order to hear myself think.  While I know I am extra sensitive to the noise, it leads me to suspect that the nationality of the company may not have a lot to do with this.

From what I have seen over the years is that older equipment tends not to be designed with acoustic comfort in mind. In some cases the amount of airflow is overkill for normal, room temperature operating conditions. I have a couple of HP6012A (1kW lab supplies) and I have modified these with thermal controlled fans (at low RPM by default). Even under high load, the fans don't spin any faster at room temperature. The thermal solution is clearly designed to survive high temperatures and a lot of dust on the heatsinks. I also have a recent Keysight lab PSU. This one is whisper quiet and the fan only becomes audible under a considerable load. Personally I don't like to sit in an office with a lot of noise coming from fans. Whether you notice it or not, it does affect your thinking because your brains are busy filtering the noise away.

Replacing fans in existing equipment can be tricky. The replacement needs to provide the same airflow AND pressure. Noctua fans have a good name but aren't always the most optimal solution. There are many other fans out there which offer better airpressure at lower RPM by using more blades and/or by using a thicker frame.

I spend quite a bit of time making a Lecroy Wavepro 7k series quieter. I found a couple of rather exotic (and obsolete) fans that offer the same amount of airflow / pressure but make about 6dB less noise (which is significant). Using the service menu I could monitor the temperatures of various parts and compare the fan performance and thus make sure that the new fans offer at least the same amount of cooling.

But in the end low noise starts with a well thought out thermal design that includes things like directing the airflow. I have a Dell workstation which is whisper quiet due to the airducts that direct the airflow to the places where it is actually needed. The Lecroy Wavepro 7k OTOH has no airflow design at all so it needs a crapload of air and a lot of it goes nowhere.

[Anthocyanina]

One thing that may be of interest.

The AD2 also has the 'no zoom out' issue, exactly as the Siglents and many other scopes.

Isn't the zoom out thing with siglent that it will just capture however many samples are on the screen and not let you zoom out if you stop the acquisition?.

If that's it, then the AD2 doesn't do it. It lets you zoom out, but what limits it is the sample memory. As long as you're sampling at 100MSPS and have configured it for 16384 points for the scope, you'll be able to get 163microseconds of record, even if you stopped the acquisition at the fastest timebase. When you are sampling at a slower rate, it's because now it needs to reduce the sample rate to fill the sample memory, and at some point, it will be just enough to fill the screen.

[tautech]

Isn't the zoom out thing with Siglent that it will just capture however many samples are on the screen and not let you zoom out if you stop the acquisition?.

Only affect some models, certainly not all and is a mimic of the LeCroy way of managing captures.

[Martin72]

Replacing fans in existing equipment can be tricky.

It IS tricky when you know what´s behind the fan thing - Most didn´t and simply exchange the fan.

The replacement needs to provide the same airflow AND pressure..

Yepp, thermal design is the keyword although I have doubts in some cases, manufacturers are taking care of this.
For example the MSO5000...
As known I´d give mine away for modifying it by rigol itself(Display, Fan).
It came back and was hearable more silent - They simply took a resistor in series to the fan, no fan exchange, no regulation implemented.
But the scope was getting significant warmer so I´ve asked them, if this was proofed before and they answered yes.....yes it´s no problem.
Aha...

Noctua fans have a good name but aren't always the most optimal solution. There are many other fans out there which offer better airpressure at lower RPM by using more blades and/or by using a thicker frame

Signing this.

But in the end low noise starts with a well thought out thermal design that includes things like directing the airflow.

Yepp, most manufacturers don´t follow this and letting the flow going anywhere.

[balnazzar]

Personally I don't like to sit in an office with a lot of noise coming from fans. Whether you notice it or not, it does affect your thinking because your brains are busy filtering the noise away.

But in the end low noise starts with a well thought out thermal design that includes things like directing the airflow. I have a Dell workstation which is whisper quiet due to the airducts that direct the airflow to the places where it is actually needed. The Lecroy Wavepro 7k OTOH has no airflow design at all so it needs a crapload of air and a lot of it goes nowhere.

As you may imagine, I agree about the overload on the brain.

And I learned how important the airflow is by months of trial and error in my journey into building a silent computer. Also, fan grills and filters do impact upon noise and air pressure and airflow up to some extent.

[balnazzar]

If that's it, then the AD2 doesn't do it. It lets you zoom out, but what limits it is the sample memory. As long as you're sampling at 100MSPS and have configured it for 16384 points for the scope, you'll be able to get 163microseconds of record, even if you stopped the acquisition at the fastest timebase. When you are sampling at a slower rate, it's because now it needs to reduce the sample rate to fill the sample memory, and at some point, it will be just enough to fill the screen.

Thanks man, I didn't know that. I'm still delving into the manual, there are a lot of things to learn for such a small and apparently simple device.

Let me ask you: is there any way to get a log sweep out of the AWG? Thanks!

[jasonRF]

Interesting thing about Hexamer's FRA fro Picoscope is that full source is available. And if you look into source, signal generating part is separated in module and can be replaced with code that drives some other generator... In  fact there are preparations in code for it..

That's very good news. If it's witten in Python, I have some programming experience with it.
Thanks!

It is not written in Python.  Sorry!

[baldurn]

From what I have seen over the years is that older equipment tends not to be designed with acoustic comfort in mind.

Or sadly just older designs even though the instruments are sold as new. Or even new designs. I own multiple new instruments from Siglent and some of them are silent with clearly temp controlled fans. Others clearly have no fan controller and just run at full blast at all times. It may be several design teams and the guy that cared only works on one of the teams :-)

An instrument that runs the fan at full blast will have thermal limits that allow one to install a controller to regulate that down in a normal office. The thing will be designed to run in a desert country, and if you do not happen to be in such a location, it will have more airflow going than needed.

People suggest replacing the fan but my first choice would be to install a fan controller. The existing fan might be just fine when not at max rpm.

[nctnico]

A fan controller is an  option but the problem is where to locate the temperature sensor. There may not even be a single place because temperature can depend on the type of load/use of test equipment.

[2N3055]

A fan controller is an  option but the problem is where to locate the temperature sensor. There may not even be a single place because temperature can depend on the type of load/use of test equipment.

LOL, almost verbatim what I wanted to write.... I agree.

Also sometimes it is not even about evacuating air from the case but mixing it inside to even out hotspots...

[baldurn]

A fan controller is an  option but the problem is where to locate the temperature sensor. There may not even be a single place because temperature can depend on the type of load/use of test equipment.

If all else fails you can use ambient temperature. My Siglent SDM 3055 appears to run the fan at full blast always. It is specified as "Full accuracy from 0 °C to 50 °C, 80% RH and 40 °C, non condensing". Currently the temperature in my lab is 19 degrees celsius - who thinks the thing really needs to run at full blast as designed for 50 degrees?

Just don't turn it all the way down. 50% removes most of the noise.

Of course they also specify aging depending on temperature, and so allowing higher temperature will increase aging. That would be a tradeoff you need to decide on. Maybe also the reason they run the fan at max is so they can get slightly better numbers in the datasheet regarding aging.

[Martin72]

A fan controller is an  option but the problem is where to locate the temperature sensor. There may not even be a single place because temperature can depend on the type of load/use of test equipment.

Would be interesting to know where the tempsensor is placed on a scope which got fancontrol.
Can´t imagine for example that the adc´s got a internal sensor like cpus got, makes me curious to look where my HD got the sensor placed.
I´ve got a feeling that most regulator circuits are measuring simply the ambient temp on the board.

Currently the temperature in my lab is 19 degrees celsius - who thinks the thing really needs to run at full blast as designed for 50 degrees?

The specified ambient temperature based on the "full blast" fan, when you reduce the fanspeed to the half it will not mean that the max ambient temp will also reduced to the half, to 25°.

[rf-loop]

Can´t imagine for example that the adc´s got a internal sensor like cpus got, makes me curious to look where my HD got the sensor placed.
I´ve got a feeling that most regulator circuits are measuring simply the ambient temp on the board.

I do not mean how fan temp control is done here but SDS2kX HD ADC have temperature sensor diode.
ADC data sheet page 66.: 8.1.5.1 Temperature Sensor

...has an on-die temperature diode connected to pins Tdiode+/- which may be used to monitor
the die temperature.
There is, in datasheet, also more text about using this.




Disclaimer/note: I do not claim or say here anything that Siglent used this method as shown here or for this purpose!

But ADC includes this feature.

Using, for example, the LM95235/95213/95214 circuit, and this particular diode inside ADC  as sensor, can do many kinds of temperature monitoring and control accordingly.

[balnazzar]

Would be interesting to know where the tempsensor is placed on a scope which got fancontrol.
Can´t imagine for example that the adc´s got a internal sensor like cpus got, makes me curious to look where my HD got the sensor placed.
I´ve got a feeling that most regulator circuits are measuring simply the ambient temp on the board.

It shoud be sufficient to measure the ambient temp over the board, or even the general ambient temp. Indeed, the manufacturer should know how the single components thermally behave as function of ambient temperature.

[nctnico]

A fan controller is an  option but the problem is where to locate the temperature sensor. There may not even be a single place because temperature can depend on the type of load/use of test equipment.

Would be interesting to know where the tempsensor is placed on a scope which got fancontrol.
Can´t imagine for example that the adc´s got a internal sensor like cpus got, makes me curious to look where my HD got the sensor placed.
I´ve got a feeling that most regulator circuits are measuring simply the ambient temp on the board.

Currently the temperature in my lab is 19 degrees celsius - who thinks the thing really needs to run at full blast as designed for 50 degrees?

The specified ambient temperature based on the "full blast" fan, when you reduce the fanspeed to the half it will not mean that the max ambient temp will also reduced to the half, to 25°.

If you look at heatsink datasheets, you'll see that the relation between airflow and thermal resistance is not linear at all. And then you need to account for the buildup of dust that increases the thermal resistance. All in all it is not very straightforward to simply reduce the airflow.

[Martin72]

I know this, see my post you've quoted.

[Fungus]

I just found this...

https://www.eevblog.com/forum/testgear/3d-printable-case-for-analog-discovery-2/

I put it in a separate thread to make it easier to find via search.

[balnazzar]

I put it in a separate thread to make it easier to find via search.

I think that will make a lot of people happy. I think that digilent should have put something similar for sale...!