ELEKTOR FORUMS

After searching the internet I eventually found a suggestion to enter the ID code of: 00 (that is two zeros) in each of the ID code boxes, and then selecting the R8C MCU option allowed me to program the MCU.

After completing this step, the Curve Tracer worked immediately.

Perhaps those people familiar with this range of MCU devices would have known this, but it has had me confused and frustrated for far too many hours!

I should also say for anyone else looking for help on this project, T2 is shown on the parts list as a BC547, but on the circuit as BC557 (I fitted a BC557 for T2 and it appears to work correctly).

K1 on the main PCB and on the controller PCB are listed as being a 20 Pin Plug (Controller PCB) and 20 Pin Socket (Main PCB). I found that the populated Controller PCB came fitted with a socket, making 1 socket on the Controller, and 1 on the Main PCB. making two sockets, and no plugs! I then had to carefully replace the socket on the Main PCB with a Pin Header, not an easy task!

Here is an update for you poor guys who have trouble making this project work. I tried using the Flash Starter program that the Elektor article says to use. It is included in the Elektor firmware download.

But guess what. It does not work. I kept getting error windows saying the "baudrate" could not be set. I wasted a couple hours on that one, trying all different combinations of baudrate, interval and so on. Total waste of time.

Then I went to Renesas Technology's website and downloaded the same program flashsta.exe plus a bunch of other stuff included in the zip file. This apparently is a later version, and it worked the first time with no problems.

Thanks to the previous poster in this forum I knew I had to set all the ID boxes to 00.

Now, when I run the Curve Tracer application program, it connects within a few seconds and seems to work. I have not connected a tansistor yet - that will have to wait until I mount the PCB in an enclosure.

My question is, where is the quality control at Elektor? There obviously is none. The daughterboard was supposed to be pre-programmed and ready to use, but it was not. I'm very disappointed I had to waste an afternoon fooling around trying to fix their problem.

I have also tried to load the FW with the ID all ZEROs but, the same.

The Z PWM signal does not move, that means that the application gives an error

I got it work.

There was one of the quad Analog switches not working.
Substituted it and now everything works fine.

thanks for your attention
regards to all

r

Jan,
I haven't looked into this project for a while but I did send comments to Elektor directly and to the German forum on May 7, 2009. Unfortunately I never got any response reagrding my analysis from Elektor and I didn't follow up as I got busy with other projects and also moved back to the US.

Here is my summary from the German forum:

I did some measurements and analysis of the curve tracer's analog part and noticed that when in the negative voltage mode (for PNP, P-Mosfet, P-JFet) the curve tracer does not output the full voltage range between 0 to -10V but only goes up (or if you want down) to -6.6V! The reason is a wrongly designed circuit. Looking a the circuit, you will notice that in the positive mode everything is ok as analog switch IC9 switches the output voltage throughh R33 to the input of IC10. Due to the high input resistance the full voltage of +10V is fed to IC10. However, in negative mode, this doesn't work as now R32, R33 and R51 are forming a voltage divider. Thus, only 2/3 of the output voltage and thus approximatey 6.6 volts are fed to pin 3.

I recommend to get rid of R51 and the tracer will work just fine, at least mine does.

Unfortunately, his bug cannot be detected easily as the PC software apparently only assumes the correct voltages are output by the device. Hence, the supplied PC software just displays these assumed values. For the positive mode everything is peachy but as mentioned above not so for the negative mode.

Andy

Hello,

I am experiencing two problems with my curve tracer, which may be related, don't know yet.

1) All measured transistors are low on gain. I think that largest hfe I measured is around 35-40. Thats not all, the curves displayed does not look at all as in the datasheets. At low collector currents the hfe~1-5, where a small signal transistor should be maybe ~100. see pic (bc547a).

2) When measuring some transistor types, mostly npn i get the response - no collector current at maximum base current. Even though I am totally sure that I connected the correctly.

I have currently no clues what the problem can be.

I have: checked all resistors, capacitors, changed opamps, analog switches. Have to do some more measurements later.

Are these problems of bad chips and no loaded firmware common to the 'assembled and tested' boards, or just to the kits ?

I want to characterise a tube of 6n139 split-darlington opto-isolators with a view to extending the published 6n136 spice model (*). A simple curve tracer which can uplink to PC+Excel seemed to fit the bill.

I don't have the skill to trouble-shoot the circuit or program a PIC, so I'm concerned that I might have to send the unit back if it does not 'plug and play'...

I'm also concerned that the basic circuit may be 'buggy', even a year after publication-- This isn't how I remember Elektor !!
===

(*) Took a LOT of research, but a HP document (c)1977 mentioned that LED-->diode CTR is ~0.16% for their range of optos. This figure is quoted for 6n136, but not for 6n139. I'd planned to use it as my default value, but nice to have it confirmed...

the problem you have is the same i had in the past, when one of the analogue switches was not working. see my previous post were I say:

"The thing is that the TRACER still does not completely work, this because the PWM signal from TIMER Z stay stady, while the PWM signal Y works correctly (oscilloscope display).
Of course this gives an error"

The error that was given was the same you are mentioning

NOw I suggest you to have a look to the ana switches, or maybe substitute them all if ever. If one does not correctly work, you won't be able to have the collector current, this is ONE of the reason of that particular error.

take care..
bye to all