I have inspected the database there is no difference apart from the DeviceID and the voltages - which is correct.
I do not understand how the LF is working. As the software will detect the F version and not operate. As you have the F version only I do not understand how you know the LF works.
you are using Manual Select ?
any other switches set or unset?
a clone PK3?
I think if you check the voltages you may find a local issue. The PK+ software inspects the voltage and it the voltage is below 4v8 and is expects 5v0 it will not detect the chip.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Evan, thanks you so much for all the work you put into this, for my part I can work with what I have so far, for an 18 years old device maybe I'm the only one in 2025 to use it.
I build a stand alone test PCB for the 16F1937 and use 2 PICkit3 programmer clone
to re-test with the target device powered by the programmer to avoid differential voltage
still same result from both software old Microchip, and the PICKit3+.
Both software find: (if any can help)
selecting 16F1937 with Target VCC Source: auto-detect, when checking Communication
Microchip software,
it return this message: 'Parts in this family are not auto-detect.'
PicKit3 +
set back the Device: PIC16LF1937
If needed It will be my pleasure to send you at no charge the PCB test
to help you get further on this chip
Let me know.
by mentioning you used a PDIPs as test , I got an old circuit that I design 8 years ago
that had a PDIP.
I removed it and tested again the same maner as with the TQFP, and it's working fine
on both software.
I guess this could mean the batch of TQFP IC's from aliexpress are a bad copy or all the 20 IC's are defective only on that specific F or LF detection.
Let me know your thought about.
the package is ready to ship today, let me know if you still needed for other research
@JB. I really appreciate you sending the PCB with the chip for testing—it was incredibly generous of you, and it has made a real difference in getting to the bottom of this issue. 🛠️
After running tests across multiple environments—including PICKitPlus, MPLAB IPE, Northern Programmer, and K150—all consistently identified the chips as PIC16LF1937 rather than PIC16F1937. This discrepancy is certainly worth noting, especially for anyone relying on specific device characteristics. ⚠️
I would not want to make assumptions about the AliExpress supply chain, but given our findings, I would advise caution when using this batch of chips. If you have any further thoughts or insights, I would love to hear them.
Thanks again for your support and for going the extra mile in sharing the test board—it is truly appreciated. 🙌
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I was worried the package will never get to your place, since the clerk at canada post told me it should be in your hand in about 6~7 days , (posted may 10) when checking last week I asked how long should it take then a another clerk says it should take 2 to 3 weeks.
about aliexpress you're advise to be caution is well fonded since it's not the first time I had some issue with aliexpress supply chips.
many times when ordering timer or other IC's I check the manufacturer datasheet eg: National , Fairchild, Texas to match the received devices, and after some test I can ask for a refund if not meeting the specifications. since they claim all the IC are 'new' but not original from the original manufacturer.
but for my hobby I can work with this unless I need something realy on spec's.
Last edit: JB 2025-05-28
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello,
I'm pretty sure this is not realy a problem,
with PicKit3Plus software when selecting '16F1937' and trying to read or program the chip
it return 'no device detected' but when selecting the '16LF1937' it's work fine.
does the software need an update or both 'F1937' and ' LF1937' are basicly the same
except for the VDD 3.3V / 5V since I only have on hand the 16F1937.
Thank's for helping
Hello,
What software component? PICkitPlus GUI for PICKit3 programmer or PK+commandline? or both?
Evan
PICkitPlus GUI for PICKit3 programmer only I don't use command line.
Last edit: JB 2025-04-27
A few thinks to clarify. I cannot reproduce here.
I have inspected the database there is no difference apart from the DeviceID and the voltages - which is correct.
I do not understand how the LF is working. As the software will detect the F version and not operate. As you have the F version only I do not understand how you know the LF works.
you are using Manual Select ?
I think if you check the voltages you may find a local issue. The PK+ software inspects the voltage and it the voltage is below 4v8 and is expects 5v0 it will not detect the chip.
my Pickit3 programmer is a clone, and I manualy select the F, or LF from the PicKitplus3 software.
(Also the Microchip PicKit3 software does the same for this PIC)
The LF selection is working on a 16F1937 chip only when the 'PicKit3 ' programmer
is powering the target PIC at 3.3V
If the target is set at 5Vdc the device is not detected either 'F' or 'LF'
I did verify my test jig, and the VDD is clean & regulated at 5Vdc (7805) Regulator
see from the screen capture.
maybe this 16F,LF1937 is made for only 3.3V.
Many thank's
OK.
What are the setting for Tool/Target VDD Source? This controls the voltage.
What happens when you power externally?
As this happens with the old software.. this does point to the clone as I am not get the issue here.
The confuses me "maybe this 16F,LF1937 is made for only 3.3V" do you have a F or LF part?
sorry for the confusion
Tool/Target VDD Source = Auto-Detect
when target is powered externally = Device is not detected 5.1Vdd or 3.3Vd
I only have 16F1937 chip, and the only way I'm able to write to it is by manualy select 16LF1937
either Microchip old software or PICkit3 +.
Interesting... send me your hex file please.
PIC16F1937 Pulse Test.HEX
file was compiled with PBP3 Pro.
Also included GCB files
GCB 16F1937 TEST.hex
GCB 16F1937 TEST.gcb
Last edit: JB 2025-04-29
I have tried and tried. Different 1937 chips, different boards. different programmers. I still cannot not reproduce.
This has to be voltage or decoupling issue locally.
Hard to fix as I cannot reproduce.
Can you send me something that will fail here?
Evan, thanks you so much for all the work you put into this, for my part I can work with what I have so far, for an 18 years old device maybe I'm the only one in 2025 to use it.
I build a stand alone test PCB for the 16F1937 and use 2 PICkit3 programmer clone
to re-test with the target device powered by the programmer to avoid differential voltage
still same result from both software old Microchip, and the PICKit3+.
Both software find: (if any can help)
selecting 16F1937 with Target VCC Source: auto-detect, when checking Communication
Microchip software,
it return this message: 'Parts in this family are not auto-detect.'
PicKit3 +
set back the Device: PIC16LF1937
If needed It will be my pleasure to send you at no charge the PCB test
to help you get further on this chip
Let me know.
@JB.
Thanks for update and your feedback.
Do send to me. If you need details please send me a personal message. Send the chip on the board please. Interestingly, I have been testing PDIPs.
Evan
I think you found the answer.
by mentioning you used a PDIPs as test , I got an old circuit that I design 8 years ago
that had a PDIP.
I removed it and tested again the same maner as with the TQFP, and it's working fine
on both software.
I guess this could mean the batch of TQFP IC's from aliexpress are a bad copy or all the 20 IC's are defective only on that specific F or LF detection.
Let me know your thought about.
the package is ready to ship today, let me know if you still needed for other research
Thank's again
Interesting, very interesting.
Do send . I will try to figure out exactly what is happening.
Evan
I assumed you've got specialize equipment to find out.
will send today.
Last edit: JB 2025-05-11
@JB. I really appreciate you sending the PCB with the chip for testing—it was incredibly generous of you, and it has made a real difference in getting to the bottom of this issue. 🛠️
After running tests across multiple environments—including PICKitPlus, MPLAB IPE, Northern Programmer, and K150—all consistently identified the chips as PIC16LF1937 rather than PIC16F1937. This discrepancy is certainly worth noting, especially for anyone relying on specific device characteristics. ⚠️
I would not want to make assumptions about the AliExpress supply chain, but given our findings, I would advise caution when using this batch of chips. If you have any further thoughts or insights, I would love to hear them.
Thanks again for your support and for going the extra mile in sharing the test board—it is truly appreciated. 🙌
my pleasure,
I was worried the package will never get to your place, since the clerk at canada post told me it should be in your hand in about 6~7 days , (posted may 10) when checking last week I asked how long should it take then a another clerk says it should take 2 to 3 weeks.
about aliexpress you're advise to be caution is well fonded since it's not the first time I had some issue with aliexpress supply chips.
many times when ordering timer or other IC's I check the manufacturer datasheet eg: National , Fairchild, Texas to match the received devices, and after some test I can ask for a refund if not meeting the specifications. since they claim all the IC are 'new' but not original from the original manufacturer.
but for my hobby I can work with this unless I need something realy on spec's.
Last edit: JB 2025-05-28
I am aware that there are clone 16F690 but yours is a new clone
Added to list of mysterious products.
:-)