Author Topic: DSO-X 3024A Power Supply defect  (Read 20128 times)

0 Members and 2 Guests are viewing this topic.

Offline georges80

  • Frequent Contributor
  • **
  • Posts: 912
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #50 on: February 16, 2021, 07:07:59 pm »
It's mostly self evident once you have the scope opened up and remove the dead PS. There's a few wires/connections.

Mains input, Front panel low voltage on/off control,  +12/0V main DC power and sense wires. If you are competent enough to use the scope then you can handle the re-wiring :)

Most of the work is getting the new supply to fit nicely (with the re-wiring) into the rather cramped space. Take your time.

cheers,
george.
 

Offline gman76

  • Contributor
  • Posts: 32
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #51 on: February 22, 2021, 11:37:34 pm »
Hi georges80, I'm buying the RPSG-160-12.  I took the original supply out and looked at the CCH125 pinouts. Everything looks obvious except the pin called REMOTE on the 1x6 header.  There are 3 wires coming from the main board into the CCH125, -S +S and REMOTE.

On the RPSG-160, I see two possibilities: PG and PS/ON.  I think its the PS/ON that I need to connect the REMOTE wire to, correct?  The pin called PG looks like an output.  If this is the case, then I'll need to remove the 3 wires from the orig connector and place the two sense wires in a new housing and the REMOTE wire into another housing since the RPSG has these split into two connectors.

 
The following users thanked this post: rernexy

Online 2N3055

  • Super Contributor
  • ***
  • Posts: 6448
  • Country: hr
Re: DSO-X 3024A Power Supply defect
« Reply #52 on: February 22, 2021, 11:54:09 pm »
Hi georges80, I'm buying the RPSG-160-12.  I took the original supply out and looked at the CCH125 pinouts. Everything looks obvious except the pin called REMOTE on the 1x6 header.  There are 3 wires coming from the main board into the CCH125, -S +S and REMOTE.

On the RPSG-160, I see two possibilities: PG and PS/ON.  I think its the PS/ON that I need to connect the REMOTE wire to, correct?  The pin called PG looks like an output.  If this is the case, then I'll need to remove the 3 wires from the orig connector and place the two sense wires in a new housing and the REMOTE wire into another housing since the RPSG has these split into two connectors.

PG is power good output. Remote is ON/OFF..
 

Offline rolkinas

  • Contributor
  • Posts: 44
  • Country: lt
Re: DSO-X 3024A Power Supply defect
« Reply #53 on: February 25, 2021, 03:08:01 pm »
Regarding original PSU.
I experimented with higher value resistors in place of cooking 4pcs 22K. AC voltage controlled by variac. Resistor values and corresponding psu start stop voltages below:

220k - start 80Vac , stop 75Vac
310k - start 98Vac , stop 75Vac
440k - start 135Vac , stop 75Vac
660k - start 190Vac , stop 75Vac
750k - start 217Vac , stop 75Vac

NOTICE: Tested ohmic values are total resistance for all 4 resistors (because all 4 all are in series). So 220K means 4pcs of aproximately 56k

When psu starts, voltage drop across resistors becomes near 380V because PFC starts as well.
turn off voltage is determined by PFC

Even with 220k resistor stayed at about 70 degree C what is much more usable {hint to 110Vac grid fellas :) }

« Last Edit: February 12, 2023, 11:34:55 pm by rolkinas »
 
The following users thanked this post: martinr33, Reprobyte, rernexy

Offline gman76

  • Contributor
  • Posts: 32
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #54 on: March 01, 2021, 08:19:44 pm »
Received the Meanwell RSPG160 supply, expanded holes to get the screw holes lined up, re-arranged wiring on connector.  Plugged it into 120VAC, not sure if it would go bang, but quietly powered up.  Scope is now back up and running.

Worked fine at 12.0V., but adjusted Vout up to 12.6V. 

The only issue now is that the front power button doesn't work.  I probed the REMOTE signal (PS/ON) and it's always low.  Pushing the power switch doesn't change the DMM reading, just slightly.  Maybe need a GND return?  Or is this the wrong signal?

EDIT:  It was indeed the ground connection of the 1x4 header (CN901).  I made a wire with a crimp on each end.  Front power button works great.
« Last Edit: March 08, 2021, 11:02:04 pm by gman76 »
 
The following users thanked this post: salvagedcircuitry, Reprobyte, rernexy

Offline tapelu

  • Newbie
  • Posts: 2
  • Country: de
Re: DSO-X 3024A Power Supply defect
« Reply #55 on: March 15, 2021, 10:09:13 am »
I'm liking the fact that the Murata unit is higher efficiency and most importantly does not have a perpendicular daughter board, like the meanwells. This should make the murata unit easier to repair, but then again, none of these units should need any attention for a long time.

The Murata PQC250-12 works just fine with three DSOX3014As that I've repaired - the oldest repair has had close to 1 year of constant use with no issues so far.

Like with other units, the wiring harness needs to be modified to make it work but it does offer remote sense as well as remote on/off (possibly with inverted logic, but the button will still work.)

I must also admit that I'm biased - I work at the Murata power supply design center where the PQC250-12 was designed and qualified.

Hi, how do I have to modify the wiring harness so that the power supply starts with the power switch? If I connect pin 1 to pin 8, the oscilloscope starts directly, but I want the power switch to be able to be used?

:-DD
 

Offline Madmanguruman

  • Contributor
  • Posts: 10
  • Country: ca
Re: DSO-X 3024A Power Supply defect
« Reply #56 on: March 16, 2021, 12:50:44 am »
I'll share my reverse-engineering below.

The pinout of the Molex header (on the DSO-X side) is as follows:



1: +12V
2: +12V
3: +12V
4: +12V return
5: +12V return
6: +12V return
7: Return sense
8: +12V sense
9: Remote on/off

For the PQC250-12:

- Run 3 wires from the DSO-X Molex header pins 1-2-3 to +12V (any of J2 pins 1-6)
- Run 2 wires from the DSO-X Molex header pins 4-5 to +12V return (any of J2 pins 7-12)
- Run 1 wire from the DSO-X Molex header pin 6 to J3 pin 8 - this is the return for the remote on/off signal
- Run 1 wire from the DSO-X Molex header pin 7 to J3 pin 6 - this is the low-side remote sense
- Run 1 wire from the DSO-X Molex header pin 8 to J3 pin 5 - this is the high-side remote sense
- Run 1 wire from the DSO-X Molex header pin 9 to J3 pin 4 - this is the remote on/off signal
- The AC cable assembly will plug in directly without modifications.



Because the PQC250-12 is open-frame I usually put a thin insulating sheet behind the wiring harnesses / in front of the power supply.



Three DSOs with PQC250-12 are deployed in my lab, and all have passed their annual traceable calibration tests (done by an external accredited laboratory).

(Depending on the remote enable logic, you may find that the pushbutton is reversed: in=off, out=on - this can be changed via a few resistor pops/no-pops but why bother?)
 
The following users thanked this post: EB5AGV, martinr33, salvagedcircuitry, rernexy

Offline tapelu

  • Newbie
  • Posts: 2
  • Country: de
Re: DSO-X 3024A Power Supply defect
« Reply #57 on: March 16, 2021, 06:02:32 am »
 :-+
Perfect ! Many, many thanks for your effort !
 

Offline ketch

  • Newbie
  • Posts: 1
  • Country: pl
Re: DSO-X 3024A Power Supply defect
« Reply #58 on: April 21, 2021, 08:26:21 am »
Hello. I need to know the value of the resistor from the photo.
 

Offline martinr33

  • Frequent Contributor
  • **
  • Posts: 363
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #59 on: December 06, 2021, 02:32:28 am »
I used madmanguruman's recommendation of a Murata power supply.

The only difference is, I used the existing sense lines from the oscilloscope board for the sense lines back to the small 8-pin connector. The reversed power switch is not bothersome.

I did try to troubleshoot the existing supply, but it was less cooperative than  I liked. The PSU has a lot more parts than  the example schematic for the 3842A (presumably for PFC) so there could be other failures.

My notes on that journey below.

Apply some 63/37 tin-lead solder to the pads before you try to remove parts. This technique is particularly effective on the metal frame. The existing lead free solder has a high melting point and is difficult to rework otherwise, the tin-lead trick makes it much easier to melt. Make sure that the solder flows for this trick to work. Alternatively, hunt down some indium based solder. 

1) It seems that units running on 220 - 250V might be more vulnerable to failure. The main cap is rated 400V, and can see 350V in some countries. Couple that with the local heat, and life could be short. Also, the 7A fuse is overrated for 250V, meaning more energy can let loose in a failure. Finally, some 250V countries (UK, for example) have very low impedance and can provide a lot of destructive current in a half cycle. 

2) Capacitor ripple current should be lower at 250V, but I think the higher voltage offsets any advantage of the lower current. A higher voltage electrolytic cap always improves life. 

3) The failure appears to cascade like this:
 - Main capacitor 100uF 400V cooks down to almost no capacitance
 - Switching FET shorts out under voltage/current stress caused by high resistance C1 

 - High current flows to ground through 0R11 2W current sense resistor on top side of board, which connects the drain of the switching FET to ground. Note that the fuse could blow here, limiting the damage to the fuse, the transistor, and the cap. There's no thermistor on the TI sample schematic, but this part is also at risk in the Agilent PSU. , - 0.11 ohm resistor blows open circuit. The drain of the main switching FET now rises to the voltage across the cap, 350V in 250V regions and 170V in 110V regions. The gate, now part of a shorted blob, follows. 

 - With the FET drain path to ground open,  the 350V now flows from the FET gate through the 2R2 gate drive resistor and into the output stage (pin 6)  of the switching PSU controller 3842A. Therefore, the 3842A does not survive if the 0.11 Ohm resistor blows. The 2R2 resistor also blows. 

- there is also a 2K2 resistor that connects to the sense input on the 3842A. This resistor can also blow, if the shorted internal 3842A drive path goes open circuit. The high voltage will flow to ground through the sense circuitry.

 - There's a final path to ground through the optocoupler phototransistor, but these transistors usually have a quite high current rating and seem to survive. 
 
The following users thanked this post: rernexy

Offline asp

  • Regular Contributor
  • *
  • Posts: 63
  • Country: ro
Re: DSO-X 3024A Power Supply defect
« Reply #60 on: January 05, 2022, 12:45:05 pm »
Hi

   Does anyone know the value for the resistor in the picture below? Mine has a hole in it and it's open. The resistor in question is located on the daughter board near the UC3842 PWM controller of the PSU

Thanks

Update: The resistor is a 499Ohm, and if that is gone the comparator might be gone as well
« Last Edit: January 09, 2022, 11:42:15 am by asp »
 

Offline NY2KW

  • Regular Contributor
  • *
  • Posts: 148
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #61 on: July 03, 2022, 07:20:28 pm »
Thanks to all for the detailed repair advice.  My 3054A suddenly died but on inspection I saw no obvious blown parts, but the PCB near the 4 22k resistors was badly discolored, no obvious cracks.  The fuse was still good but the pair of IF840's were both reading as dead shorts in both directions.  I had a PQC250-12 from another project so quickly made the new harness's and used a Teflon sheet as an insulator between the harness and the new SMPS.  Turned on and works perfectly.   The power button is reversed as the PQC250 datasheet says that its J2 Pin 4 (PS-ON) which goes to the scope's Molex Pin 9  has to be pulled low to power on while the original CCH125 series its J6 Pin 6 to the scopes Molex Pin 9 has to be pulled High to power on.   I am not a purist but I need the PQC250 for my original project so instead of buying another Murata, it seems the button logic on the Mean Well RPSG-160-12 should work same as original, so that's where I'm going next.  Too bad this means making a new harness.

Jerry NY2KW   
 

Offline jtmp

  • Newbie
  • Posts: 1
  • Country: nz
Re: DSO-X 3024A Power Supply defect
« Reply #62 on: September 23, 2022, 04:29:38 am »
I have had the same blown power supply on a DSO-X 2012A, and replaced with a Mean Well RPS-400-12 following the advice above.

However now I am getting a ~120mVp-p ripple @~4kHz on the scope trace, even with the probe shorted to ground.

Have I done something wrong here, or are these power supplies much noisier than the original? I don't remember what I used to get but surely it wasn't this bad?!

Cheers

« Last Edit: October 02, 2022, 10:57:00 pm by jtmp »
 

Offline EE-digger

  • Frequent Contributor
  • **
  • Posts: 333
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #63 on: January 17, 2024, 08:01:24 pm »
Old thread but still highly relevant.

I finally had time to open my (10/2016) build DSOX3104T and was surprised to find the same troubled (??) supply, same revision, but with the GE branding.  So my question for all.  Did these supplies happen to come to this thread to die?    :-BROKE

Perhaps units that smoked, exploded or otherwise died had enough run time and heat to bring out the problems.
With the nature of the problems, I would expect a high failure rate unless improvements were made without revision change.

This problem, NAND corruption AND NAND LIFE IN GENERAL have me worried.  If Daniel drops by, I would love to hear the Keysight take on these issues.
« Last Edit: January 17, 2024, 08:49:02 pm by EE-digger »
 

Offline electr_peter

  • Supporter
  • ****
  • Posts: 1300
  • Country: lt
Re: DSO-X 3024A Power Supply defect
« Reply #64 on: January 17, 2024, 09:22:26 pm »
PSU seems to be fragile, at least in older DSOX2/3 models. Scope is newer fully off (ON/OFF switch is actually soft power-on) and 6W heat in standby has to go somewhere. One strategy to prolong PSU life is to disconnected mains externally when not in use.

NAND issues seems to have been mitigated (at least partially) via FW updates.
 

Offline EE-digger

  • Frequent Contributor
  • **
  • Posts: 333
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #65 on: January 17, 2024, 11:29:39 pm »
The NAND issues have been masked by redundancy and error checking.  The NAND chip has a life of 10 years on paper (Micron). 

It would be nice if someone could identify the SMPS used in scopes of recent builds.  It only takes 5 minutes to peek at it.  If it's the same model, that's disappointing.  If it's changed, an ID might help with replacements.

« Last Edit: January 20, 2024, 05:20:12 pm by EE-digger »
 

Offline EE-digger

  • Frequent Contributor
  • **
  • Posts: 333
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #66 on: January 21, 2024, 01:42:40 am »
In the good old days (and now), HP and others would keep power on for critical areas (oscillators, converters, etc.).  Some of those are now in the 40 year old range and still trucking.  Enter the SMPS ...

This SMPS power supply failure issue, especially when explosion of one or more components is involved, is disturbing enough to take affirmative action.  For someone who keeps his cars ~15 years or more, the expensive little oscilloscopes is a close second.

For our off shore friends running on 230v ac mains, the picture is not so good.  For 120v, the resistors should pose no problem.  From calculations, dissipation in each of the 22k resistors and approximate temperature rise for the mounted part is:

120vac - 0.08 watts - estimated temperature rise 15 to 20 degrees
230vac - 0.3 watts - estimated temperature rise 45 to 55 degrees

Add this to the temperature rise of the cap itself from ripple dissipation and you have a bad picture for 230v.  Credit to poster rolkinas in reply #53 of this thread for his bleed resistor measurements.

The killer is that if your scope is plugged in (most labs) this heat rise is constant.

After finding several excellent quality supply replacements, I decided to replace only the cap with an 8000 hour part and change the resistors to 56k each.  Combined, this should allow the scopes to last a LONG time.

added - the above assessment was incorrect.  The main capacitor has 390 volts across it on a 120v ac line.
« Last Edit: February 01, 2024, 02:14:32 am by EE-digger »
 

Offline CapLeaker

  • Contributor
  • Posts: 38
  • Country: ca
 

Offline EE-digger

  • Frequent Contributor
  • **
  • Posts: 333
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #68 on: January 21, 2024, 05:34:52 pm »
Thanks.  It points further at that cap.  If an earlier poster is correct, that cap is rated for 3000 hours.  That's only 4 months.

Under "normal" use, this would stretch into years but with the resistors under it, all bets are off.  You're in Canada though so the heat rise in that group wouldn't have added a whole lot but still could have been another 20 degrees C.

I have two DSOX, one coming in from Keysight repair shortly.  I'm going to open and look at both with an IR camera.  I'll lose the forced air but this supply should easily handle the moderate load (from what I see on bench supplies).  One scope is 2016 and the other is 2013 vintage.

They are usually not supplied with ac so may appear more like new-ish.
 

Offline CapLeaker

  • Contributor
  • Posts: 38
  • Country: ca
Re: DSO-X 3024A Power Supply defect
« Reply #69 on: January 21, 2024, 06:25:20 pm »
I do know that Agilent didn’t make the power supply at least not for my 2012. It was a third party power supply. Look at my post at Badcaps, I had uploaded the flyer from Cherokee. At least they are easy to fix. If I remember right, the biggest problem of my repair was to find a capacitor that actually fit.
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6877
  • Country: ca
Re: DSO-X 3024A Power Supply defect
« Reply #70 on: January 21, 2024, 06:37:49 pm »
The NAND issues have been masked by redundancy and error checking.  The NAND chip has a life of 10 years on paper (Micron). 
10 years for data retention, that is when you write once and leave it for 10 years. For erase/write cycles there should be a separate specs. For the Winbond NAND used in the DSOX 1200 (Linux- based) it is 100 000 cycles. So I'd say 10 years or 100 000 write cycles whatever comes first, for that particular IC.
« Last Edit: January 21, 2024, 06:57:57 pm by Bud »
Facebook-free life and Rigol-free shack.
 

Offline Bud

  • Super Contributor
  • ***
  • Posts: 6877
  • Country: ca
Re: DSO-X 3024A Power Supply defect
« Reply #71 on: January 21, 2024, 06:57:07 pm »
I would love to hear the Keysight take on these issues.

In the 1200x (Linux-based, black color) series scopes, including the EDUX 1052x model, beside the vanila ECC correction at NAND level, which does not seem helping much, they use UBIFS filesystem which has wear leveling feature. What it does is instead storing a file in a contiguous block of sectors, it sprinkles it all over the free pages across entire NAND partition. It also updates the erase counter for each page every time erase/wrire happens and monitors the counter against a set threshold. When a page erase counter exceeds the threshold, the page is marked bad and the data on it is moved to a new NAND page. This is general UBIFS feature, not specific to Keysight. They just decided to use UBIFS.

While UBFS handles wear leveling, this may not help with random NAND cells corruption. I'd think plain corruption may still be happening, whatever a reason might be. On 1200x scopes Keysight has a separate Recovery kernel, which seems to be kicking in when normal boot cannot be performed.
« Last Edit: January 21, 2024, 07:01:42 pm by Bud »
Facebook-free life and Rigol-free shack.
 

Offline EE-digger

  • Frequent Contributor
  • **
  • Posts: 333
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #72 on: January 27, 2024, 12:57:17 am »
Posted some new findings on these supply failures here:

https://www.eevblog.com/forum/testgear/stunning-3000a-2000a-3000t-power-supply-discovery-must-read-if-you-have-one/msg5300998/#msg5300998

On 120v mains, a new Murata vs the Scope factory supply is an eye opener.  My pristine little 3000T was running at 397 volts when supply was removed and no load.  Murata was at 168v.

I may be missing the doubling+ mechanism for this but the voltage is there, the cap is very hot (will measure and image in the next days).

For the most part, please reply in the other thread, I guess.
« Last Edit: January 27, 2024, 01:49:33 am by EE-digger »
 

Offline EE-digger

  • Frequent Contributor
  • **
  • Posts: 333
  • Country: us
Re: DSO-X 3024A Power Supply defect
« Reply #73 on: January 29, 2024, 05:33:58 pm »
Has anyone using the Murata PQC250 supply noticed a fair amount of audible noise from it, with or without load?  I believe its coming from one inductor mounted on "stilts" toward the lower right corner (viewed with ac inlet on the left).

Murata support says they've never heard noise from this series.  If true, it may explain why I was able to buy these  :)  I've used a number of medical SMPS and never had a noise issue.
 

Offline peterbitter

  • Newbie
  • Posts: 1
  • Country: sk
Re: DSO-X 3024A Power Supply defect
« Reply #74 on: February 16, 2024, 09:42:47 pm »
Can I ask you to share please the possible wiring of Meanwell power supply (Meanwell RPSG-160-12) with the Keysight
Many thanks
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf