Rigol MSO5000 Power Supply

[Protegimus]

Gents, please don't respond to the threadcrappers any further here. They have an agenda and that has nothing to do with improving the MSO5000 PSU.

@2N3055 directly - your hysterical statements and the over complication that you seek to use in order to support your claims is laughable. I didn't come here telling people what to do, nor am I ruining my credibility by making personal insults and being passive aggressive.
I'm glad Rigol have continued to make improvements to the firmware, increases the scope of what the tool can be used for. If you want to take measurements that help identify what's been done; great, go do them, but do it in your own thread. I deliberately didn't publish measurements using the 500uV range, although I do have screenshots.

It's a simple low pass filter and yes it's possible to improve the performance of a low pass filter. As pointed out Rigol already bypass the 470uF capacitors, we are simply doing the same. Using SMD could offer improved results, nothing to stop you using SMD if you prefer.

[2N3055]

So what is your agenda ??

[beatman]

2n3055 have you ever made test's? I mean measure components or acoustical test temperature test or sound analysis? i do all the time for years and I see only positive results when is something better soldered on circuit. we don't push here nobody to do anything only we tell our experiences the guy open the thread ask for something and get answers.one example here expensive pro power amplifier with main smps filter capacitors Daewoo brand one of them leaked.for me is unacceptable.replaced with kemet.the amp now it sounds fantastic on subwoofers.low tier caps  is not for stressing things like smps.is a mater of time to goes bad

[TurboTom]

Maybe, after the re-capping job, the scope's fan sounds better... 

[Protegimus]

With the foolery now out of the way let's engage on some of these ideas. Seems looking at ways to improve the PSU wasn't nonsense at all.

Did you by any chance performed frequency domain analysis so we can see what frequencies got attenuated ??

No I was previously thinking in simplistic terms of DC output rather than targetting filtering.
I may do this when I next open the unit to see where it's at.

...  Probably few well placed SMD capacitors placed on mainboard PCB locally would be even more gain...

Agreed. One step at a time though - it makes sense to address the PSU first as the root, improving output quality there benefits everything down the line.

... Also why did you place TH film capacitors across the elcos instead of ceramic SMD capacitors closer to output, or closer to high current paths to minimize current loop length.

Rigol already decouple all supply rails with ceramic SMD capacitors on the output C in the CLC filter:
They were not fitted on the input side close to the rectifiers.

J5
+5V - C16
-7V -  C15
+7V - C12

J4
+9V - C13
+7V - C14

+5V rail on J5 appears to be the high current rail as it has two pins on that connector.
Appreciate the suggestion of using SMD instead of thin film. That may offer some further improvement.

... Did you try to put more filtering stages between PSU and mainboard?

No. I like the simple approach.

Maybe, after the re-capping job, the scope's fan sounds better... 

You should try staying quiet

[Sighound36]

All I'm going to say on this matter is Crapzone caps would come no where near my own designs. Cost is the first and for most use of this brand imho. If you are purchases 100K or Millions of these the cost per unit must be rather appealing. I'm back to my hum drum life now

[BRZ.tech]

Hi.
I liked @Protegimus' suggestion of adding polyester capacitors to the primary and secondary SMPS of the MSO5000.
The effect of drastically reducing the NOISE FLOOR was excellent, from 202.93uVrms to 67.715uVrms, on the 1mV scale, remember that this scale has the BW locked at 20MHz, see the “B” on the 1.00mV side.

The NOISE FLOOR reduction was 66.63%. This, in terms of NOISE FLOOR, elevates the MSO5000 to a higher level of quality.

One detail I noticed is that in the Excel BOM, the 0.1uF/200V and 0.01/200V capacitors, in parallel with C1, have insufficient insulation, the voltage there can reach Vp>340V. I suggest, in those positions, 630V isolation.

I also verified that a 10nF/J/400V polyester capacitor was added, in parallel with CS5, which is the +5VDC voltage output.
My suggestion is to add a 10nF/J/400V capacitor in parallel with the CS1 capacitor... You can keep the 10nF/J/400V capacitor, in parallel with CS5, without any problems. (see the photo).
Maybe the NOISE FLOOR will decrease a little more.

In my opinion, the topology of this MSO5000 SMPS is Fly-back. With the IC-U1 controlling the PWM and the TR POWER MOSFET, which does not have silkscreen on the board.

I also verified that the soldering of PTH components are not excellent. The suggestion is to carry out a welding review of these components, with a soldering station at 480ºC, and good quality solder.
Request: if anyone has already reverse engineered this MSO5000 SMPS, please post the circuit here in this topic. Identifying the name of main components such as IC-U1 and POWER MOSFET.

As soon as possible I intend to do this update on my MSO5074.

[Protegimus]

Thank you for reviewing the updates and posting the recommendation for increasing the component rating.
Good to see the improvement was repeated.

[thm_w]

Good to see the improvement was repeated.

They are just quoting your numbers, they haven't actually done anything yet.

[Protegimus]

Good spot, jury is still out then. 

[Protegimus]

...
I also verified that a 10nF/J/400V polyester capacitor was added, in parallel with CS5, which is the +5VDC voltage output.
My suggestion is to add a 10nF/J/400V capacitor in parallel with the CS1 capacitor... You can keep the 10nF/J/400V capacitor, in parallel with CS5, without any problems. (see the photo).

Quick note to say that CS2, CS3, CS4 and CS5 are all on the +5V rail. Taking the suggestion from earlier in the thread I've ordered KEMET C0805C104J1REC7210 Multilayer Ceramic Capacitors MLCC - 100V 0.1uF X7R 0805 5% X7R SMD capacitors to bypass these just as Rigol have done with CS9, CS12, CS15 and CS21 but not yet had the scope open again to fit them.

CS1 is actually between the anode/cathode on one side of DS1 the JH SRF2060C Schottky barrier rectifier diode (DS2 & DS6 also have the same 26R1 resistor and cap, to damp switching noise?).
Attached image has the additional components marked up.

[BRZ.tech]

Hi,
Dear @Protegimus
You are right about the CS1 capacitor. See in my post #31 the marking I made in the position in the image I sent. Note that the CS2 (electrolytic) capacitor was without the silkscreen, and you added it to your message, in message #35. Correct.
Therefore, my suggestion is to place the 10nF/J/400V capacitor in parallel with CS2 (electrolytic), on the solder side. This will maintain the logic of your brilliant article.
Here in my country I managed to buy TDK 10nF/J/400V and also 630V polyester capacitors. But PANASONIC brand electrolytic capacitors are not available here, and to import them, the legal taxes and shipping are astronomical and unfeasible. I won't change them for now.
I also intend to place an extra 12VDC cooler, to blow the heat generated from inside the case to the outside, over the SMPS heatsink, to help with the internal ventilation of my MSO5000. I will put between the Voltages (+7V and -7V), which will give a total between them of 14VDC. And insert there (3x  1N4007 diodes) in series with (+7V), to give a drop of 3 x 0.7 = 2.1VDC. To maintain the Cooler voltage specification at 12VDC.

In my country, temperature and humidity variations are very high, and on days with temperatures above 30 degrees Celsius, the inside of my MSO5000 heats up too much.
I have already placed the new cooler outside the case, with an external 12VDC source, and the results are good, the internal temperature drops a lot. Without the new fan, sometimes the CH2, CH2, CH4 channel inputs are activated randomly, without the external cooler.
I hope to soon complete the entire procedure and list the results here.
I haven't opened my MSO5000 yet. It is now out of Factory Warranty. I have already defined what to do to install the polyester capacitors, and how to attach the new Cooler.
I am very grateful for your commitment.