If you dispense it with a 1ml syringe, the pressure you can develop in the syringe is much higher (pressure = force/area) so it is much more controllable. Trying to squeeze it from a 10ml syringe is very difficult and not easy to control. However, I might try using the larger syringe to fill the 1ml one. However, as I said earlier, stencils are now so cheap that unless it is for a one off without too many components that’s the way I would choose.
Yeah, that is an obvious point when you stop to think about it, but also a trivial detail that hides in plain sight. I got out the old paste syringe and - like you said - it’s size is 10 mL . I guess I need to get a package of 1 mL syringes and keep them on hand for small, one-off, projects and rework.
Dale
2 Likes
We now know Dale’s wife does not have access to his computer…
I have been using laser cut frameless stainless stencils for all my recent boards, inspiration came from this http://www.hoektronics.com/2012/10/27/super-simple-smt-stencil8/ the two most important items are the “Precision Tooling Fixture Block” and the reflow oven. The oven performs better with some modification, start here if you are so inclined https://github.com/UnifiedEngineering/T-962-improvements/wiki and here http://www.electricstuff.co.uk/reflowoven.html with emphasis on removing the masking tape. I found credit cards make particularly good ‘squeegees’.
Two things I have learned,
Stencils come in different thickness’s if you are doing a lot of very fine pads thinner is better.
I order my boards with HASL finish, a couple of times a minuscule lump has been left on the solder surface causing the stencil to not seat flush to the board and allowing surplus paste to leak through. I now ‘scrape’ the the board prior to fitting the stencil, you will feel any lump.
Three things…
I have a lot of difficulty getting low temperature paste to work effectively, I was using it on the second side of double sided boards. Now I use the higher temperature paste on both sides by not putting any large components on the first side reducing the risk of departure from the board when it is heated after being flipped.
2 Likes
I think this is one of the reasons for ENIG pad finishes. If your source doesn’t charge extra over HASL you might want to consider that finish. (Yes, I know, soldering onto gold finishes feels wrong to us old-heads. But apparently the gold finish on ENIG is thin enough to mitigate many of the issues with gold embrittlement. (Though a top hit scan of a google search does show a wide range of opinion on this…)
I didn’t mention this but my modules have the counterfeit chip. Because of that I couldn’t take advantage of most of the features. I could only transmit one byte at a time and I think I should been able to transmit as many as 32 bytes at a time.
Sparkfun sells the IC. Now I realize all I need is a hot air gun and maybe a stencil. Thanks for the footprint!
No, the counterfeit ones can send 32 byte packets. They contain SI24R1 chips, which would be a legitimate NRF24L01 compatible part if not fraudulently labeled NRF24L01+.
I’ve actually never seen a REAL NRF module
Where did you buy your IC’s? Is SparkFun a reliable source?
Wow. I just did a search. Thanks.
Yes. Sparkfun is a reliable place. I usually go to ebay or amazon for NRF modules.
If you get the datasheet from Nordic, you’ll see the sample board layout is the NRF module you get from disreputable offshore sources.
If I was making a product, I’d get real NRF chips, or maybe real SI24R1 chips and lay out the circuit on the main board 'stead of buying Chinese modules. But the modules are so cheap, easy to use and work so well, I don’t mess with wires more than this long or so for hobby projects.
Thanks. I’ll do the same. I’ve been looking at SMD videos using paste with fine tip soldering irons. I’m pretty sure I could do it. The NRF is a fabulous IC. It has its own built-in set of instructions and the communication between devices is stable.
So I guess we have different opinions here.
Well, yeah, the chip can be beat into submission to get it working, but it’s full of “features” that are very annoying to work around to make the chip work with anything than the very weird and limited “6 pipe” model.
After a lot of pain to write a decent library for it I discovered it had a barely adequate range (30m or so)
So I had a working RF module, and the pain was past me cause I had a library to beat it into submission for me, but because of the bad range I kept looking further.
Then I found the RMF69 modules.
- None of that 6 pipe rubbish.
- Decent power reception measurement (so transmit power can be dynamically adjusted).
- Built in encryption modules.
- Different frequency bands (434 / 868 MHz) lots of channels in those frequency bands.
- Castellated PCB’s are probably also better for P&P machines.
There were more irritating issues with the Nordic chip, while the datasheet of the NRF69 looks much better. After I got the RFM69 modules my focus wavered to other projects and they are still lying unused in a drawer somewhere.
I never concerned myself with “pipes” and 30 meters was about what I got with the trace antenna.
Several years ago I built a small mobile robotic arm with pan/tilt for a camera.(proof of concept) I’m a hobbyist so I was excited to do that with such an inexpensive module. Shortly after that I started noticing problems and discovered I had a bunch of fake chips. I haven’t bought any since then because I was not sure they would be authentic.
I’m going to look into RFM69. It sounds interesting.
Yea. The 6-node “network” feature is lame.
I just give each a unique 5 character address and store the addresses of other “nodes” in either flash or eeprom. I use mainly AVR’s or Arduinos. Each gadget listens on its own address, but when it wants to send a message, switches the the address of the node it wants, sends the message, gets the ack, then switches back.
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.