LeCroy DDA5005A - Fan Replacement

[amacdon]

I previously followed the other DDA5005 threads but found little information regarding this subject. I have spent some time evaluating, planning, designing, and testing parts so that you may not have to in future. to begin, I upgraded the DDA5005A main 120mm fans because they were very loud and unnecessary. I compared different fan specs and found the original OEM fans to be pretty much the best performing fan outside of noise performance. See the attached excel file for comparison. there is obviously a better IPPC3000 fan in the list but these are 4pin PWM and I did not want to use an inline fan speed controller for this since the fans already receiver a modulated DC voltage to control fan speed.

Originally, as mentioned in another thread, it seemed simple enough to just buy two 120-140mm fan adapters and get two Noctua NF-A14 IPPC2000 3 pin fans. This was not the case for this oscilloscope. the fans are mounted to an aluminum-foam-aluminum plate sandwich podium and was designed to be installed in a one way method. the fans are mounted to the plate with flush pop rivets which made it so the fans had to come out with the plate. this was challenging and almost cost us a damaged ribbon cable. be careful removing the fan assembly. The new fans will also not fit directly back onto the fan podium so I designed a direct mounting 2 piece plate and printed it with solid infill on my 3d printer. see the attached 2 pieces and print them out of any material you want. I chose PLA+. Since the new Noctua fans have rubber isolators you no longer need the aluminum-foam podium to vibrationally isolate the fan. the larger 140mm fans also require a larger inlet so the new mount I designed is IMO better. the fans can now be mounted to the half mount before being installed into enclosure, unlike before. this also means they can be removed just as easily next time. I used new M3 screws to mount the fan mount to the enclosure since the mount is thicker than the aluminum. You can either tap the holes like I did or make them bigger and insert knurled brass heat set insert nuts using a soldering iron. both files are available, just print the correct one.

Section 1: getting inside
this breakdown will start from an untouched oscilloscope. the next section will cover the fan replacement. the top metal enclosure needs to be removed but first you must unscrew the top two back plastic feet and then remove the plastic front cover plate.
1. for the feet, pop out the rubber bumpers with a pick or flat screwdriver, then unscrew using a long torx T25 bit.
2. for the front panel, there are three T10 torx screws under the panel at the front. remove them. then lift the front panel off carefully. ours was already cracked.
3. now the top metal cover can be removed. there are three screws revealed by the front panel on top and two on each side at the front. all T10 torx.
4. if you have the printer in the top cover just remove the connection to it after gently pulling the enclosure cover off. ours did not have one so it was simpler for us.
5. You are now ready to change out your HDD for an SSD hard drive and begin the fan installation as well!

section 2: fan replacement
1. I suggest that you first try to break off the pop rivets with side cutters to easily remove the fans.
2. Then, cut all zip-ties holding down the wires/cables in the channel between the fan and motherboard. the EMI ferrite is the biggest obstacle which can cause damage to surrounding wires.
3. next, remove the top USB front panel using T10 Torx bit. this is so that it does not obstruct the removal of the fan mount. This USB front panel has a thin ribbon cable adhered to the underside which connects the acquisition card to the front panel controller. do not forget to reinstall this at the end. the ribbon came out with the front panel so I forgot to reinstall the ribbon in its FPC connector.
4. remove the aluminum-foam podium with T10 torx bit. when pulling it out, make sure to lift the side at the knob panel and try to pull it through the side hole instead of directly up. push all of the wiring which was released from zip ties out of the way and be careful not to cut or pinch them.
5. when you remove the fans, it would now be a good opportunity to clean up the dust and debris on all the boards. Our front panel board has a couple mechanical encoders that have failed from debris build up so we removed the board and cleaned it thoroughly with an ion gun (compressed air that doesn't produce static) and 99% alcohol. stick a screw driver into the fan impeller first before blowing compressed air.
6. mount the Noctua 140mm fan to the corresponding half base plate I designed. The original fans suck air into the oscilloscope section down from the PC section. both fans are clearly marked so make sure you follow the same orientation. install each mount plate with fan attached into the enclosure separately and lock them together first before screwing anything down. I recommend using tweezers in the hex hole on the screw to insert them into their spot. some locations are hard to reach by hand.
7. you may notice that the Noctua fan pinout is not directly compatible with the original Panaflo FBA12G12U1A fan. This was a simple fix. first, de-pin the Noctua connector and reinsert to match the color/ pinout scheme of the Panaflo fan. the Noctua connector can only be installed with the connector oriented with its exposed contacts facing upwards. next with pliers, wiggle off the shroud around the pins. these scopes are sufficiently old enough that the shroud is barely holding onto the pins soldered to the fan controller PCB. you do not have to unscrew this PCB to do any of these tasks. leave it be unless you have an issue with that board. ours was fine.

I will post more photos taken and files in next post.

[amacdon]

here are some pictures from the fan removal and installation process

[Coordonnée_chromatique]

Hello, why don't you use PAPST, they are more silent than Noctua.
PAPST 4312 NL : 100m³/h 25dB
 

[nctnico]

The key parameter is static pressure, not airflow. From my experience modifying the same oscilloscope, you need fans with high static pressure and neither fans mentioned (the PAPST one is even worse) will be suitable for this oscilloscope. I found that Nidec D1225c12b7 (http://whitefan.com/type6/item/showitem.php?car=130&cate_no=130&item_no=3774&cha=&) fan provided the same amount of cooling compared to the original fans by comparing temperatures through the maintenance menu. Per usual Lecroy MO, the construction of this oscilloscope is pretty bad and the designers did literally zero to optimise airflow. So what is left is just pushing lots of air through the frame hoping it gets where it needs to be.

[amacdon]

Hello nctnico, yes static pressure is the key here but since they pull air from the large enclosure into the analog section the static pressure is minimal. Our temps are still great using Noctua fans. I added the rest of your Nidec D1225c12b7 fan in my excel table and it only has a slightly better static pressure. Do you have measured temps to compare? see new table attached.

[Coordonnée_chromatique]

A high static pressure requirement is a sign of an undersized exchange surface, why don't enlarge it instead of putting a large fan with a air duct ?

[nctnico]

Hello nctnico, yes static pressure is the key here but since they pull air from the large enclosure into the analog section the static pressure is minimal. Our temps are still great using Noctua fans. I added the rest of your Nidec D1225c12b7 fan in my excel table and it only has a slightly better static pressure. Do you have measured temps to compare?

As I wrote: I have experimented with several fans until I found the Nidec model which resulted in the same temperatures compared to the original fans. I wanted the machine to be more quiet but not risk overheating it at the same time. All the hot parts are Lecroy specific chips and thus nearly impossible to obtain / replace.

[amacdon]

Hello, The replacement fan is the simplest modification I could do without destroying the oscilloscope sheet metal. the elimination of the aluminum fan podium already had a significant impact on volumetric flow to reduce the static pressure requirements. your suggestion is valid but this discussion just revolves around replacing the fan to reduce sound while maintaining low temperatures.

[nctnico]

Yes, but that is the goal of me replacing the fans as well. Reduce the noise. According to a sound meter, I got over 6dB less noise measured from the same distance & location from the front of the oscilloscope.

[DaJMasta]

I used some Scythe branded ones when doing a similar operation on a WP 7300.  You can only get so quiet since the airflow is pretty critical for the acquisition board, but they've got some lower sound options that still push quite hard.

In a related note, if you need to probe the top of the acquisition board, you need to have a plan for how to get airflow over it, since the only fans for it are mounted in this upper chassis.  What worked for me was a big sized box fan aimed down at the table, and for troubleshooting it was sufficient (checked with a thermal camera), but just leaving it open air is a recipe to cook your ADCs and similar.  You can pull out the power cables a little farther than is needed to get to the board and completely swing the acquisition board flat on the bench with the rest of the scope on its side, you just have to be careful doing it, and the cables don't offer much more slack than that.