Forum - Recent Threads

as7421 problem with SMUX and GAIN configuration

TU1234 — Fri, 30 Aug 2024 10:11:38 GMT

Hello,

I'm trying to configure the SMUX and ADC gain of the as7421, but the files explaining the configuration specified in the datasheet are missing and the company's support does not respond to my messages. I have already readout the specified RAM, but have not yet been able to recognize a pattern in the data that has already been entered in order to know where I have to write which values into the RAM. Has anyone solved a similar problem? I would appreciate it.

RE: as7421 problem with SMUX and GAIN configuration

Akropolidis — Sun, 10 Nov 2024 09:04:40 GMT

Hello,

I'm working on a sensor library for the AS7421 as well to use for a project and I was just wondering how you decided on the configuration ({0x21,0x21,0x21,0x21,0x43,0x43,0x43,0x43}) for the ram registers

RE: as7421 problem with SMUX and GAIN configuration

TU1234 — Tue, 24 Sep 2024 06:35:57 GMT

Thank you very much for the answer. In the meantime, I sorted the photodiodes using the LEDs. I got the same result.

I don't know why, but autozero doesn't seem to have any effect. I have tried autozero with different settings, but the result is always the same as when autozero is off. The channels are only zero, if I write certain values in COMPDAC. So I think I can use COMPDAC as an offset. However, the last 16 bytes of COMPDAC seems to have no effect. The measurements are always the same, no matter what I write in there.

RE: as7421 problem with SMUX and GAIN configuration

primalfido — Sat, 21 Sep 2024 10:29:46 GMT

Hi there,

I do not know what P2RAM does. I think COMPDAC is related to temperature compensation but I cannot know for sure.

You might be missing autozero, enable it and then set all of the SMUX RAM locations to 0 (i.e. ground), and you should see low/zero output from all of them.

Additionally, using the default setup I stated above, you can arrange the channels in wavelength order using this ordering:

1, 48, 2, 34, 16, 32, 18, 51, 4, 49, 3, 35, 17, 33, 19, 54, 0, 13, 50, 63, 52, 6, 38, 20, 36, 22, 55, 5, 53, 7, 39, 21, 37, 23, 40, 26, 42, 24, 56, 10, 58, 8, 41, 27, 43, 25, 57, 11, 59, 9, 44, 30, 46, 28, 60, 14, 62, 12, 45, 31, 47, 29, 61, 15

This is what I have gotten from observing the LEDs on/off and it seems to match the shapes given in the datasheet. The 830 photodiodes might not be necessary in the right integration order but you can always average them .

RE: as7421 problem

CorinnaR. — Mon, 16 Sep 2024 06:33:24 GMT

Hello,

thank you for showing interest in our products.

Please use the following link for your question.

Technical support | ams OSRAM (ams-osram.com)

Best regards

as7421 problem

fjj19990227 — Fri, 09 Aug 2024 09:34:01 GMT

I am trying to control the as7421 to read spectral data using python's smbus, but there is a problem that I can't change the value of LTF_EN on the enable bit 0x60, so I can't take spectral measurements, has anyone solved a similar problem, I'd appreciate it!

RE: as7421 problem with SMUX and GAIN configuration

CorinnaR. — Mon, 16 Sep 2024 06:16:38 GMT

Hello TU1234,

thank you for showing interest in our products.

Please use the following link for your question.

Technical support | ams OSRAM (ams-osram.com)

Best regards

RE: as7421 problem with SMUX and GAIN configuration

TU1234 — Tue, 03 Sep 2024 10:39:31 GMT

Many thanks for the informations. I also noticed that there are 5 LEDs.

Do you know what the "P2RAM" is for? Does the bit "REG_BANK" have to be set to high for the configuration? I initially read the RAM at the offsets with both variants and there are only a few bits different.

Do you know what "COMPDAC" is for? Does this RAM section also need to be initialized? I tried changing this section and the channel values seem to change.

The ASETUP seems to work. I tried to set all photodiodes in the SMUX to low in order to then assign individual photodiodes to the ADCs, but some channels still change when all values in the SMUX are low.

RE: as7421 problem with SMUX and GAIN configuration

primalfido — Fri, 30 Aug 2024 20:12:32 GMT

doing this from my phone so please bear with me. I have spent a good month figuring this chip out.

They’re not explained in the datasheet. in fact, much is not explained. Did you know there are 5 LEDS? Not 4!!

The smux is basically junk at boot up and needs to be set properly so there is no pattern.

From what I’ve discovered through reversing it. The ASETUP ram locations are for the GAIN. ASETUPAB correspond to channels in the A and B integration and CD for the C and D integration. Both combined give the full 64 photodiodes.

You can set the values on eaxh cfg_ram between 0 and 8 giving a gain of 2^x. Setting 6 on CFG_RAM_x gives you a 64x Gain on photodiode x. I usually just fill all ram registers with the same value.

SMUX is a bit harder. On each SMUX_x location. Each of the 32 CFG_RAM register represents 2 photodiodes as 4-bit nibbles. Each of the 4-bit nibbles can be set to either 1,2,3 or 4 representing each of the 4 available ADCs in each integration annd 0 as ground Each photodiode can be remapped to any integration cycle using this. However I am not quite sure which photodiode is mapped to which nibble.

here is my procedure to set the smux with what works for me.

go to each smux offset

set all cfg_ram to zero

then I use the configuration {0x21,0x21,0x21,0x21,0x43,0x43,0x43,0x43}

set the offset to smux_a

write the config only at cfg_ram_0 -> 7

smux_b, write the config to 8->15

smux_c, config 16-> 23

smux_d config 24-> 31

now the channel data read out is not in order of the datasheet PD listing and needs to be sorted correctly. I have some indices figured out. I will post them in a reply when I find them

RE: as7421 problem

fjj19990227 — Mon, 26 Aug 2024 06:41:59 GMT

thanks it helps me a lot！

RE: as7421 problem

primalfido — Mon, 19 Aug 2024 10:37:55 GMT

PON must be enabled before LTF_EN can be enabled.

Additionally there is a library here for Python that you can use which might be helpful: github.com/.../as7421

RE: as7421 problem

fjj19990227 — Mon, 12 Aug 2024 01:25:53 GMT

as your suggestion，i check the i2c address.besides 0x64(correct address)，0x50-0x5F also exist at the same time. Weird.

RE: as7421 problem

bidrohini — Sun, 11 Aug 2024 16:02:50 GMT

The problem may exist somewhere else other than the LTF_EN itself. I think it is better to verify the I2C address with an I2C scanner code.