ESR meter upgrades

[stj]

another update.

took a while.
this one has a new makefile.
also a new testing mode - you want a 10-30pF cap for stage-2 calibration.

[stj]

fresh update
this one is a must,

The programmer Pieter-Tjerk (radio amateur PA3FWM) has added some additional code
for testing of Unijunction transistors and Progammable Unijunction Transistors.

now some bad news.
there are so many features that you cant fit *everything* into the memory.
so for now there is no 16MHz builds wih the crystal tester.

once i get a good 16MHz crystal i'll start experimenting to get the best stuff to fit.
probably by removing the rotary-encoder tester as it seems pointless,
and maybe the serial port - as the pc side seems to be unforthcoming at the moment.

[roko]

Hi.
Capacitors for calibration must be 10-30pf or 10-30nf?

[SteveNielsen]

I'd like a bit of clarity on the serial port. Is that for adding a USB port or? I'm just curious. I recall some earlier versions of the TTester had USB possible. How was that used with a PC? Is it not useful at all in this one yet or just not moving ahead?

I'm not progressing past SVN630 for the time being, I'm in a wait and see mode.

What quality of 16MHz xtal is needed? I haven't seen anything below 30ppm accuracy available online.

[stj]

farnell/element14 has 10ppm crystals.

the serial port is on pin26 and forwards all the display data.
you can connect it through a max232 serial convertor or a usb2serial adaptor.
it was intended to be used with a lookup table of transistor specs to see if a part was good.

the usb port on some models was just for battery charging.

[stj]

Originally Posted by roko

Hi.
Capacitors for calibration must be 10-30pf or 10-30nf?

i see the error.
it's nF

SamplingADC is now enabled by default. The calibration procedure asks
for a capacitor for L-measurement with the message "Cap for L meas?".
You should use a capacitor value between 10nF and 20nF for future
measurement of little inductances. If your coil is only detected as
resistor, or with little inductance, you should connect this known
capacitor (10nF-20nF) parallel to the coil and repeat the measurement.
The measurement is now possible between any test ports.
You can only measure little capacitors between TP1 and TP3!

[SteveNielsen]

Originally Posted by stj

farnell/element14 has 10ppm crystals.

There is a difference between freq stability and freq tolerance. 10ppm tolerance doesn't mean 10ppm stability. I've seen some say 10ppm tolerance with 30ppm stability. I presume stability relates to operating temperature?

I'm not sure how much it's gonna matter for the tester though. What are your thoughts? What are you looking for? Curious Steve here

[stj]

i suspect the original part is 30, but as there isnt much price difference - why not use 10
i think i'll try building some 16MHz code later so i can see how to save space without losing any of the part tests.
i know using a smaller font works - but it dont look so good.

[stj]

managed it.
you can build a 16MHz version with all component functions including crystal testing by removing the rotary-encoder tester and the serial port and using a smaller font.

[ben7]

I'm thinking ... I have two different STM32 "nucleo" dev boards, I wonder if I can translate/port this code to one of those ... I'm becoming more and more interested in building one of these
(I've never ported code before, so that would get interesting. I haven't fiddled too much with microcontrollers beyond "arduino" crap XD)

I do have some ATMega328P chips, but as you said, there really isn't any more room for more features ... which the STM32 chips would help significantly with.
Also, I suppose that certain measurements (capacitance?) might be more accurate with a "blazing" 48Mhz clock speed

-Ben

[stj]

48MHz? that's baseline, i was looking at the NUCLEO-F746ZG the other day.
218MHz, hardware accelerated code and 144pins!
i have to have one of those!!! specially now they announced this:
http://www.st.com/web/en/press/p3781

i dont think you can convert this project though for 2 reasons.
1 - the STM is 3.3v and you need 5v i/o to drive stuff like fets.
2: the design is very reliant on the specifics of the atmel A/D and D/A convertors.

if your designing from scratch, you could use an atmega644 or atmega1284, unfortunatly they cant directly replace the atmega324 because they jhave more pins.

btw, i managed to build the 16MHz code without losing the serial port now.

[ben7]

Originally Posted by stj

48MHz? that's baseline, i was looking at the NUCLEO-F746ZG the other day.
218MHz, hardware accelerated code and 144pins!
i have to have one of those!!! specially now they announced this:
http://www.st.com/web/en/press/p3781

i dont think you can convert this project though for 2 reasons.
1 - the STM is 3.3v and you need 5v i/o to drive stuff like fets.
2: the design is very reliant on the specifics of the atmel A/D and D/A convertors.

if your designing from scratch, you could use an atmega644 or atmega1284, unfortunatly they cant directly replace the atmega324 because they jhave more pins.

btw, i managed to build the 16MHz code without losing the serial port now.

1. You can drive certain fets with 3.3V ... the FDV304P, a P-channel device, has a typical gate threshold voltage of 0.86V! Granted, it's not a power fet, but it can be used to switch other things on and off ... I'm quite sure N-channel devices are out there with similar gate thresholds.

2. True, I'll have to study the code to see how that all works. It'll likely require some modification, but I think it would be worth the effort... or it would at least be educational for me.

-Ben

[stj]

we are not talking about driving "certain fets", we are talking about driving whatever fets you need to test!

and driving triacs is even harder!

[ben7]

Originally Posted by stj

we are not talking about driving "certain fets", we are talking about driving whatever fets you need to test!

and driving triacs is even harder!

Ohhhhhhhhh......
LOL XD

I see ... hmm ...
Welp, I could do without the fet and triac testing feature for the moment I guess.
I'll have to take a look at just how it measures those, to see if there is a way around that limitation. I'm quite sure I recall there being some pins that are 5V tolerant, but they'd have to be externally driven up to 5V...
Perhaps this could be integrated within the zener testing circuitry, which would have a high voltage supply (Though it could possibly damage FET gates if it's too high).

It's mainly the capacitor testing that I want to do... inductance measurement would come in handy as well, for all those unmarked inductors I have...



-Ben

[stj]

and, 8 & 16MHz builds again.

[SteveNielsen]

The readme says:
-
you also need a good stable capacitor between 10 and 30nF to calibrate the new small inductor mode.
if you dont have one, just ignore the cal message till it times-out

to test small inductors you MUST connect the small cap used to calibrate the unit in parallel with the inductor.
----------------------
SamplingADC is now enabled by default. The calibration procedure asks
for a capacitor for L-measurement with the message "Cap for L meas?".
You should use a capacitor value between 10nF and 20nF for future
measurement of little inductances. If your coil is only detected as
resistor, or with little inductance, you should connect this known
capacitor (10nF-20nF) parallel to the coil and repeat the measurement.
The measurement is now possible between any test ports.
You can only measure little capacitors between TP1 and TP3!

Questions:
How does SamplingADC enabled effect testing operations?
It says use 10nF-30nF or 10nF-20nF for inductor cal and testing, which is it?
What value is considered a little capacitor that need to use TP1 and TP3?
Does calibration for and testing small inductors need to use TP1 and TP3 as well?

[stj]

no idea to the second question.
sampling adc makes readings more accurate for caps and maybe inductors, what it does is test the cap lots of times at different frequency's instead of using one fixed speed.

[SteveNielsen]

Thanks man.

I suppose I'd use 10-20nF and just TP1 + TP3 just in case, not like it's a big deal, just my curiosity and wanting precise info. I do have some used inductors I've not been able to measure and maybe this new development will help.

Have you got a good source for a 10ppm 16MHz xtal yet? I might go for that soon. It's gotta be surface mount right?

[stj]

i got my 16MHz xtal fitted - works great.
it even i.d.'d the original crystal i took off as 8000KHz

i also upgraded my voltage reference from a TL431 (1% drift) to an LM4040A (0.1% drift)
**one thing you cant see in the foto - i had to cut the top pin off the LM4040A to isolate it from the board.

where can you order from with free delivery?

btw, about the small cap thing.
i'v seen a guy on a different forum fit the cap under the socket through a reed-relay.
i'm looking into that.
he switches it with a switch though - i'd rather use a spare pin on the Atmel.

code.

[stj]

some more improvements to the autodetection.