Now we shall mod (Scratch) ... So that it generates the Blinky code ... Without us typing a single Lisp Parenthesis! (°-°)

github.com/AppKaki/blockly-uli

Show thread

(Scratch) needs a Code Generator for each language ... So we copy the Dart generator, paste it as Lisp ... And tweak the Lisp generator for 🛠️

github.com/AppKaki/blockly-uli

Show thread

Previously we've done a Code Generator for Embedded ... But it got complicated because of Type Inference ... Lisp should be simpler, I think 🤔

lupyuen.github.io/articles/adv

Show thread

Yep Web Serial API works with at 2 Mbps ... Soon we shall transfer some uLisp Scripts ... From Blocky (Scratch) to BL602!

googlechromelabs.github.io/ser

Show thread
Show more

@lupyuen I might have some insight into the SX1276? issue. It just occurred to me what to do.

Try using sudo and lsusb -v to measure the power being allocated to the BL602 board. I had readings of 98mA. An ammeter would work too. I don't think it's getting the power to emit more radiation.

@AmpBenzScientist Interesting... so we think BL602 might not have sufficient power for the LoRa Module?

Here's what I observed...

(1) BL602 + SX1276 seems to be emitting the right power (see the red chirp)...

lupyuen.github.io/articles/lor

(2) But BL602 + SX1262 seems to be emitting much less power (the faint chirp)...

lupyuen.github.io/articles/lor

If it's a power problem, why is SX1262 affected but not SX1276 🤔

@lupyuen Frequency and the power required to generate it possibly.

@lupyuen SX1276 is the one I can't use but it's close to 1GHZ. The other uses a frequency about half of that.

@lupyuen As an American I will have scary people hunt me down and put me in the back of a van if I don't abide by FCC regulations. The SX1278 is the only one I can use. If I'm not mistaken it's 915mhz. I also live within rough sensor range of one of the largest Air Force Bases on the East Coast in the US.

@AmpBenzScientist haha it's the same in Singapore. We don't even allow Satellite Dishes 😉

@lupyuen No satellite dishes? They could mandate a mesh size for the parabolic reflector that would mitigate any potential problems for improper use.

@lupyuen Your country is against Chinese influence or for it? My country is both.

@AmpBenzScientist My country is too small ... We won't dare to offend any other country sigh

@lupyuen I would say peaceful because cowards don't push boundaries in scientific advancement.

@lupyuen I don't like minesweeper irl. What are your thoughts on the SX chip problem?

@AmpBenzScientist Hmmm I'm a bit too distracted right now ... Gotta write up about Lisp on BL602 ... Then test the Pine64 LoRa Gateway.

I guess SX1262 works for now, even though it seems suboptimal

@lupyuen Just got done shaving with the safety razor and I finished it by shaving with the foil shaver. I don't think I've ever had such a close shave. I still need to clean and care for the shaver. The things we do to be accepted by society.

@lupyuen I will look into the SX1262 and SX1276 datasheets and report back if I find something and have a good idea.

@lupyuen The SX1276 has a lower power usage during transmission except for PA_Boost being used. At 13dbi it typically draws 29mA while the SX1262 is 45mA at 14dbi in the higher frequency range. It's 32mA at the same strength in in low frequency.

The SX1262 with 22dbi will draw 118mA in the high frequency mode. Keep in mind that is "typical". My EVBV1.1 was using 98mA.

@lupyuen I think that's per transmission and it seems that the SX1262 does this more often. 32.768kbps vs 62.5kbps in Lora mode. These transponders are quite different from each other. I would not recommend using the BL602 as anything other than a controller for this. The transmission power would be better supplied from a battery source.
They have the same frequency deviation but the increased transmission might make it less detectable to most SDRs. The deviation is programmable and 200khz.

@lupyuen DC-DC mode is used for the IC core but the PA is supplied from VBAT. This should be 3.3V for 22+dbi and can be configured with the LDO by connecting VDD_IN + VBAT and ensuring that the voltage is correct. This simplifies the circuit.

I can only assume that the transceiver wasn't getting the optimal voltage from the SDR graphing. Perhaps the weak strength and repeated lower strength sections were from sub or over 3.3V power and the sectors that were weaker are evidence that a component used more power than it did during the other areas. I would say that it's not the correct voltage from what I've seen.

@AmpBenzScientist Hmmm our RFM90 module doesn't expose the VDD_IN from SX1262...

github.com/lupyuen/bl_iot_sdk/

Maybe we put this on hold? I need to take this up with Pine64 before they start selling it 🙂

@lupyuen It would be nice to have a product that doesn't have hardware bugs.

Show more

@lupyuen I'm going to shave with my safety razor. Electric shavers are not acceptable for a proper shave. I will be away for the typical 15 minutes it takes to shave with a safety razor.

@lupyuen I love China and especially under Xing pin. He's done much good for the country. I think his history shows the future of China.

@lupyuen Sorry. I know almost nothing about your country. In the US we are essentially taught about our history, Russia/USSR and China.

@lupyuen It might be necessary to use a better power source for powering the SX chip.

@lupyuen
Idk but id try to do differential analysis wrt voltages or amperages etc as much as i know.

@AmpBenzScientist

@PawelK @lupyuen These are very different transceivers. The SX1262 might take more energy but the bandwidth is closer to dial up speeds. That's almost double the SX1276. The SX1276 looks like it's losing on all levels. The transmission is a bit more sophisticated and I will need to review the SDR readings to try to better understand what is going on here.

@PawelK @lupyuen It's just a product from Pine. I am not aware of a product they have made that has worked without a massive community effort. Pine makes hardware. The community makes the hardware work. When I got my Braveheart I thought it was broken after a couple of uses. I took it completely apart and photographed it.

It was not broken by me. The battery discharges when powered off and dies after a couple days. I found out why it was getting so hot during use and recommended thermal pads to the developers.

Lee is likely the reason why Pine didn't die. As far as I've seen, he's been involved with every product that they have made. Imagine buying a product that doesn't work at all, working to make it work and never getting paid for it. That's what we do. It's for the community, it's for the knowledge we gain and most importantly for the future.

@AmpBenzScientist
Pine isnt apple. Its hw manufacturer of early versions of google 20% and community fixes bugs. Then they produce closer approximatins to final product.

@lupyuen

@PawelK @lupyuen My Pinephone came with no Operating System which was a good thing. I tried to get involved in porting Debian to the device but Ubuntu circle jerk crew hoarded all the code so that they could get all the glory. They got left behind and eventually had to admit that they were not releasing the code. Pine just takes existing components and combines them into a product.

@PawelK @lupyuen My Pinephone Braveheart? Yes I do. There were only 5k sold in the world.

@AmpBenzScientist
If hw isnt unfixable id like to play with lineageos on it if you allowed etc. Or you can for yourself.

@lupyuen

@PawelK @lupyuen Pinephone Braveheart v1.1 it barely escaped the Covid lockdown even though I placed an order immediately after it became available.

Sign in to participate in the conversation
Qoto Mastodon

QOTO: Question Others to Teach Ourselves. A STEM-oriented instance.

An inclusive free speech instance.
All cultures and opinions welcome.
Explicit hate speech and harassment strictly forbidden.
We federate with all servers: we don't block any servers.