Repairing LeCroy WavePro 7300A power supply

[ollopa]

A couple years ago the power went out, and when it came back on I could no longer boot my WP7KA.  The motherboard (an upgraded Advantech) tested good with a regular PC power supply so I figured it was time to pull the scope's power supply and get to work...   Oh boy, just getting it out is quite the job and then... This thing is an over-engineered beast!

I decided I will keep a thread here to document my progress so that it might help someone else facing a similar issue and also to collect suggestions if I get stuck.  I'll add some pictures later, but for starters here are my symptoms:

As soon as power is connected I hear a continuous "CLICK CLICK CLICK CLICK..." at a rate of about 2x per second.  There is no response to the power button.  Standby LEDs on the motherboard are lit but dim and fade away with every click.

There is a large harness, tightly wrapped with zip ties, that connects to the power supply.  It spans the entire width of the scope and has seven dual-row Molex connectors.  After removing the power supply I had to wrangle this stiff harness outside of the scope so I could connect the power supply board and be able to probe and troubleshoot.  There is also a large daughter-board attached to the top cover of the power supply which I'm fairly certain is a mains 50Hz/60Hz filter.  It is not necessary for troubleshooting and I did not want a giant board carrying mains voltage to be flopping around next to an even larger board also carrying mains voltage, so I just bypassed it with jumper wires / alligator clips.

The clicking sound comes from a 12V actuated relay (AJW2211F) on the power supply, located next to the mains input.  It is rated for 250VAC/10A and I think it is responsible for switching power to the main power supply, so perhaps there is a separately powered supervisor that is shutting it down shortly after attempting to turn it on.

I messed around with the ATX connector to the motherboard for a while.  What I eventually discovered is that I can get the scope to turn on if I follow these steps:

• Disconnect AC power cable
• Disconnect ATX cable to the motherboard
• Connect AC power cable (relay clicks once and stays on)
• Connect ATX cable to the motherboard

Next I did an experiment where I removed only the +5V_SB pin from the ATX connector and confirmed that the scope power supply powers up with this rail disconnected, then I replaced the +5V_SB

Then I tried swapping in a very old ATX motherboard from my junk pile -- it didn't even have RAM in it, but the scope powered on and attempted to boot the old motherboard.

So this is interesting -- it looks like my Advantech MB puts a bit more of a load on the +5V standby rail and it is causing the power supply to reset in an endless loop.  If I let the +5VSB rail come up before putting a load on it then everything is fine.  The Advantech MB still boots with a regular PC power supply and it booted fine for years with the LeCroy power supply, so I suspect something in the LeCroy supply is aging.

I'll focus on the +5VSB rail and also investigate what is driving the relay.  I suppose it could also be related to the +12V control signal for the relay -- if that supply is sagging when the +5VSB starts up then it could be causing an uncommanded shutdown.

[tautech]

A couple years ago the power went out, and when it came back on I could no longer boot my WP7KA.  The motherboard (an upgraded Advantech) tested good with a regular PC power supply so I figured it was time to pull the scope's power supply and get to work...   Oh boy, just getting it out is quite the job and then... This thing is an over-engineered beast!

I decided I will keep a thread here to document my progress so that it might help someone else facing a similar issue and also to collect suggestions if I get stuck.  I'll add some pictures later, but for starters here are my symptoms:

As soon as power is connected I hear a continuous "CLICK CLICK CLICK CLICK..." at a rate of about 2x per second.  There is no response to the power button.  Standby LEDs on the motherboard are lit but dim and fade away with every click.

There is a large harness, tightly wrapped with zip ties, that connects to the power supply.  It spans the entire width of the scope and has seven dual-row Molex connectors.  After removing the power supply I had to wrangle this stiff harness outside of the scope so I could connect the power supply board and be able to probe and troubleshoot.  There is also a large daughter-board attached to the top cover of the power supply which I'm fairly certain is a mains 50Hz/60Hz filter.  It is not necessary for troubleshooting and I did not want a giant board carrying mains voltage to be flopping around next to an even larger board also carrying mains voltage, so I just bypassed it with jumper wires / alligator clips.

The clicking sound comes from a 12V actuated relay (AJW2211F) on the power supply, located next to the mains input.  It is rated for 250VAC/10A and I think it is responsible for switching power to the main power supply, so perhaps there is a separately powered supervisor that is shutting it down shortly after attempting to turn it on.

I messed around with the ATX connector to the motherboard for a while.  What I eventually discovered is that I can get the scope to turn on if I follow these steps:

• Disconnect AC power cable
• Disconnect ATX cable to the motherboard
• Connect AC power cable (relay clicks once and stays on)
• Connect ATX cable to the motherboard

Next I did an experiment where I removed only the +5V_SB pin from the ATX connector and confirmed that the scope power supply powers up with this rail disconnected, then I replaced the +5V_SB

Then I tried swapping in a very old ATX motherboard from my junk pile -- it didn't even have RAM in it, but the scope powered on and attempted to boot the old motherboard.

So this is interesting -- it looks like my Advantech MB puts a bit more of a load on the +5V standby rail and it is causing the power supply to reset in an endless loop.  If I let the +5VSB rail come up before putting a load on it then everything is fine.  The Advantech MB still boots with a regular PC power supply and it booted fine for years with the LeCroy power supply, so I suspect something in the LeCroy supply is aging.

I'll focus on the +5VSB rail and also investigate what is driving the relay.  I suppose it could also be related to the +12V control signal for the relay -- if that supply is sagging when the +5VSB starts up then it could be causing an uncommanded shutdown.

Or the mainboard caps are very sick and overloading the PSU.
Had very similar symptoms with a friends PC which if you kept hitting the power button it would eventually boot.

Power it up using your tricks and let your fingers do some walking around the mainboard caps and very likely you'll find a HOT one so do this in a way you can remove your finger darn quick !
Or let a thermal camera do the walk for you.

[hpw]

A couple years ago the power went out, and when it came back on I could no longer boot my WP7KA.  The motherboard (an upgraded Advantech) tested good with a regular PC power supply so I figured it was time to pull the scope's power supply and get to work...

Interesting as an upgraded Advantech mainboard. Normally as an microATX & AGP Intel D865GLC as seen on WM/SDA models.

Do you still run on AGP?

May you provide the exact Advantech brand.

Usually, the electrolyte cap's, as not high quality, to pot gets expanded and mainboard does at likes.
So recap as on tube amp's or exchange of the desktop board is may simpler.

In addition, the front board as with the encoders gets heavy dirty as the time encoder to replace as with time setting jumps.

Also interesting would be, how your front coding's is? As attached pictures.. As T00...T03

Hp

[ollopa]

Or the mainboard caps are very sick and overloading the PSU.

Thanks tautech, I have recapped a few MB's before and you're right that I should get some hard data to rule that out, especially since it powered up with a different MB.

I measured the current on this rail and the MB only draws 200mA for a few seconds, then it decreases to 183mA.  After I turn the MB on it goes up to around 250mA.  This is well within the capabilities of the +5VSB output and the power supply has no problem maintaining all the voltage rails, just so long as I take the load off the +5VSB and let all the filter caps charge before trying to turn it on.

Interesting as an upgraded Advantech mainboard. Normally as an microATX & AGP Intel D865GLC as seen on WM/SDA models.
Do you still run on AGP?
May you provide the exact Advantech brand.

That's all covered in the upgrade thread:  https://www.eevblog.com/forum/testgear/upgrading-mainboard-in-lecroy-dda-3000-(aka-wavepro-7300a)/msg1626073/?topicseen#msg1626073

I don't think anybody retains the AGP card these days because LeCroy is using it in a special Intel passthrough mode which then forces you to have to purchase an Intel board produced in a limited range of years, and they all seem to have driver issues with Windows 7.  I have upgraded the LCD and chose a MB with LVDS outputs, then I had to change the BIOS so I could configure the necessary LVDS settings and I had to make a custom LVDS cable.  I am using the original LCD backlight driver but I needed to invert the control signal between the MB and the LeCroy PCB so I made a really simple inverter with a transistor and a resistor.

Most people seem to go an easier route -- choose a board with multiple DVI or HDMI outputs and use a DVI/HDMI to LVDS adapter to driver the display.  I did not want to have an adapter and a cable hanging out the back of the scope and at the time there weren't many MB options with internal video connectors, so I went the really hard way.  I don't recommend people try to replicate my approach unless you like a challenge and can handle the frustration of troubleshooting your own LVDS cable and custom BIOS.  I also had some clearance issues with the CPU heat sink and fan and I wound up purchasing a special low-profile cooler intended for 2U servers where the vertical clearance is limited.  Clearance is not a problem, however if you remove the rear DVD drive and tray, but now I have the option to keep it.

[nctnico]

This could be anything. Perhaps try using a DC load on the 5V standby and see if the original power supply comes up as it should with various load levels. It could be the standby power supply has developed a fault. It can be anything though; from something simple like a dried electrolytic to a resistor gone bad. Likely the standby power supply is build around a standard switching chip and they used the application note circuit. This would provide some values for the various components to check. Even better would be having the schematics but I don't know if these are available.

[DaJMasta]

I feel like I've seen/heard this symptom before, but I don't remember if there was a fix for it.

I know with earlier models (WP960), while a very different architecture, they had some voltage monitoring logic on the acquisition board which had an external reset line for the power supply - does the acquisition board being connected to it play any role?

They really seem to have gone out of their way to make the power supply serviceable, sandwiching it in between layers and with all the stiff looming.

[ollopa]

Well thanks everyone for the help.  I'm still not sure if the power supply is sick but I know now that the motherboard is.  After a lot of repeated tests I can see that the motherboard sometimes draws only around 200mA of current in standby but often consumes anywhere from 300mA to 500mA which is way too much.  It has mostly solid/poly caps.  There are some electrolytics which appear cosmetically healthy but you can never tell.  None of them get hot to the touch.  A similar Advantech MB only draws 137mA in standby and all of the ones I tested from my junk pile also consume less than 200mA in standby.

As for the power supply, I did some load testing with resistors and it starts doing the clicking dance somewhere between 200mA and 250mA.  I feel like that's a little weak but it wouldn't be an issue if the MB were operating properly.

Also FWIW the +5VSB spec in this MB's datasheet is over 300mA.  There are jumpers for routing SB power to USB (for keyboard and mouse waking) and somewhere else so I'm sure this is the worst-case spec with all optional peripherals active, but it's worth noting that my power supply would not be able to handle that much standby load.

I have two of these motherboards now but the new one lacks the PS/2 ports so I'll try to find the defect in the original and report back.

[Alex_twn]

I fixed a broken PSU on a WR6100 with exactly the same symptoms as you.
The +5VSB was not there because the +15V on the primary side was not able to reach +15V.
The reason was 2 dry capacitors and 1 diode.
I can show you were on the WR6100 psu schematics if there is a chance that your WP7300 psu is similar in this area.