Confusing resistor package size choices in Fluke 289

[gamalot]

I was repairing a Fluke 289 with a faulty LCD, and the problem was actually easy to locate, the +20V rail was only a little over 15V. I disconnected the 0 ohm resistor R157 and used an external power supply to provide +20V, and now everything works fine.

I measured pin 3 of the boost chip U21 (LT1615), and the voltage there was only about 0.85V instead of the reference voltage of 1.23V I expected. A new LT1615 would save this Fluke 289, but the prices from Digi-Key and Mouser are too expensive, so I might go to LCSC.

It's just a simple 30 minute repair and nothing worth to talk about, what really bothers me is the choice of package size of the two voltage divider resistors R145 and R146 by the Fluke designer, R145 is 121K in package 0805, and R146 is 2M in package 0603.

Why? 

[robert.rozee]

likely Fluke were using component values and packages that are already used elsewhere on the PCB; the more different components (size or value) used in an assembly, the more machinery is required on a production line and hence a higher production cost. as an example, if a design uses many 10k and a few 20k resistors, production cost may be less if each 20k resistor is replaced with two 10k resistors in series. similarly, if a design uses some number of 10k resistors but in a mixture of 0805 and 0402 package sizes, it may be cheaper to just use 0805 sized throughout.


cheers,
rob   :-)

[gamalot]

likely Fluke were using component values and packages that are already used elsewhere on the PCB; the more different components (size or value) used in an assembly, the more machinery is required on a production line and hence a higher production cost. as an example, if a design uses many 10k and a few 20k resistors, production cost may be less if each 20k resistor is replaced with two 10k resistors in series. similarly, if a design uses some number of 10k resistors but in a mixture of 0805 and 0402 package sizes, it may be cheaper to just use 0805 sized throughout.


cheers,
rob   :-)

I had similar thoughts, but I checked the schematics and found that there was only one resistor in the 121K 0805 package and two in the 2M 0603 package (the other one was R293, unpopulated). If it was to reduce the BOM list, I can't explain why they used the 0805 size. Maybe they just had the device in stock.

[floobydust]

I think it's to make room for C93.

[coromonadalix]

ok  re - found schematics  loll   

https://www.eevblog.com/forum/testgear/fluke-189iian-(289)-service-manual/msg4745819/#msg4745819

[gamalot]

I was about to reply to your PM and saw this. 

[Kean]

I had similar thoughts, but I checked the schematics and found that there was only one resistor in the 121K 0805 package and two in the 2M 0603 package (the other one was R293, unpopulated). If it was to reduce the BOM list, I can't explain why they used the 0805 size. Maybe they just had the device in stock.

This is likely BOM optimisation across products.

While they may not appear elsewhere on this PCB, they probably use those resistor values on other older designs in different package sizes. The PnP lines will be used to assemble various products, and those particular resistor value/size combinations are already loaded in reels on the machines.

If the 121k parts were used multiple times on a new design then they might have chosen to move to 0603 and have to load up a reel of them taking up a new slot or requiring changeover.  It often just doesn't make as much sense for a single part placement to do that if there is PCB space to place the existing size.