Microdrop v1.0


#1

We're happy to announce the release of Microdrop v1.0:

http://microfluidics.utoronto.ca/update/application/data/microdrop-1.0.0.msi

A list of changes can be found here: http://goo.gl/PYhW4C

Note that if you have previously installed a pre 1.0 version, you
should uninstall it before installing the latest version.

This release is also available as a self-extracting portable version
which you can run from a folder on your computer or a USB thumb drive.
Note that this option does not require Administrator privileges and
allows you to easily switch between multiple versions of the software
on the same computer:

http://microfluidics.utoronto.ca/downloads/microdrop_portable-1.0.exe

We've also released an update to the DMF control board plugin (which
should update automatically, or you can manually update it from the
menu item "File/Manage plugins").

To see what changed in the plugin, check out the change log here:
http://goo.gl/zYRDpR

If you experience any problems, please send a message to the dev list:
https://groups.google.com/forum/#!forum/dropbot-dev

-Ryan

···

--
Ryan Fobel | PhD Candidate | Wheeler Microfluidics Lab |
http://microfluidics.utoronto.ca
DropBot: open-source DMF | http://microfluidics.utoronto.ca/dropbot
Institute of Biomaterials & Biomedical Engineering | University of Toronto
160 College Street Room 440 | Toronto, ON M5S 3E1
Tel: 416.946.5702 | Fax: 416.946.3865


#2

Hi Ryan,

I have installed MicroDrop v1.0 (Windows Installer) but when I launched it, I got this message in a second and it disappeared automatically (I have uninstalled the older version before):

Then I used the self-extracting portable version 1.0: everything executed normally (DMF Control Board plugin was upgraded to 1.0.10 and the firmware to 1.0.14) but I got the error message whenever I tried to initiate any calibration:

I also tried to run the old version MicroDrop v1.0 rc then upgraded the plug-in to 1.0.10 (and also the corresponding firmware) but I got the same error. So I guess the error was caused by the plug-in…

One more thing that I always get another error message after I launch the DropBot (in any version). I have read Michael’s post about this and fixed the board as instructed (cut the trace/leave the 3.3V power jumper unconnected) but still get this error:

Linh.


#3

Hi Linh,

···

On Mon, Apr 13, 2015 at 3:47 PM, Linh Tran dlin...@gmail.com wrote:

Hi Ryan,

I have installed MicroDrop v1.0 (Windows Installer) but when I launched it, I got this message in a second and it disappeared automatically (I have uninstalled the older version before):

It looks like you are running the file “Microdrop.bat”. This file is intended to be used to launch the portable version (not the installed version), which is why you are getting that error. If you are using the installed version, you should instead launch “Microdrop.exe” or use the “Microdrop” link that is added to the start menu. These both check for administrator privileges (needed to install the control board driver) and prompt you if necessary.

I will remove the “Microdrop.bat” file from the installer to prevent this point of confusion in the future.

Then I used the self-extracting portable version 1.0: everything executed normally (DMF Control Board plugin was upgraded to 1.0.10 and the firmware to 1.0.14) but I got the error message whenever I tried to initiate any calibration:

That error code you are getting (32) is a current limit error. This is consistent with the analog channel error you are seeing. When you get that “analog channels damaged” message, what are the values of A0 and A1 VGND? Look for lines like the following in your log file:

[INFO]: Analog reference = 4.35 V

[INFO]: A0 VGND = 2.17 V (49.80% of Aref)

[INFO]: A1 VGND = 2.17 V (49.80% of Aref)

-Ryan

I also tried to run the old version MicroDrop v1.0 rc then upgraded the plug-in to 1.0.10 (and also the corresponding firmware) but I got the same error. So I guess the error was caused by the plug-in…

One more thing that I always get another error message after I launch the DropBot (in any version). I have read Michael’s post about this and fixed the board as instructed (cut the trace/leave the 3.3V power jumper unconnected) but still get this error:

Linh.

You received this message because you are subscribed to the Google Groups “dropbot-dev” group.

To unsubscribe from this group and stop receiving emails from it, send an email to dropbot-dev...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.


#4

Hi Ryan,

We got 0V. How could we fix this? Thanks!

-Linh.

2015-04-13 14:10:35,707 INFO Analog reference = 4.89 V

2015-04-13 14:10:35,707 INFO Analog reference = 4.89 V

2015-04-13 14:10:35,707 INFO Analog reference = 4.89 V

2015-04-13 14:10:35,739 INFO A0 VGND = 0.00 V (0.00% of Aref)

2015-04-13 14:10:35,739 INFO A0 VGND = 0.00 V (0.00% of Aref)

2015-04-13 14:10:35,739 INFO A0 VGND = 0.00 V (0.00% of Aref)

2015-04-13 14:10:35,769 INFO A1 VGND = 0.00 V (0.00% of Aref)

2015-04-13 14:10:35,769 INFO A1 VGND = 0.00 V (0.00% of Aref)

2015-04-13 14:10:35,769 INFO A1 VGND = 0.00 V (0.00% of Aref)


#5

My control board:

And signal generator:


#6


#7

Hi Linh,

You’re going to have to probe the board with a multimeter to try and figure out why you’re getting 0V instead of the expected 2.5V:

http://microfluidics.utoronto.ca/git/kicad___dmf_control_board.git/blob/refs/heads/master:/docs/Control%20board%20v2.1%20documentation.pdf

Please measure the voltages at the following test points:1, 2 3, 4, 12, 14. Also check pins 4 and 11 on IC4 (they should be at +5V and GND).

-Ryan

···

On Wed, Apr 15, 2015 at 4:56 PM, Linh Tran dlin...@gmail.com wrote:

You received this message because you are subscribed to the Google Groups “dropbot-dev” group.

To unsubscribe from this group and stop receiving emails from it, send an email to dropbot-dev...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.


#8

Hi Ryan,

That helps! There was a loose pin at the U1. Now I got A0=A1=2.38V and no more “analog channels damaged” message! We could run the “high-voltage reference load” calibration:

Unfortunately, we were not able to run “device load” calibration and test channels. We got this error message:

Do you know what caused this error?

Thank you!

-Linh.


#9

Hi Ryan,

Somehow the other 2 calibrations are working now (loosen cables I guess). Here are the results: https://www.dropbox.com/s/cxblkt0ny1nuqv2/Mindy-LinhTran%20DropBot%20Calibration%20Results.zip?dl=0

Anti-aliasing circuit installed and checked; Auto adjust amplifier gain checked. Sometimes the LIMIT light on the Amplifier turned ON but not constantly. Normally it is OFF.

Do you have any thoughts?

Thank you!

-Linh.


#10

Hi Linh,

Looking at your switching board test results, there are still a few
switches that still look suspect:

17, 76, 77, 84, 91 and 108.

Please test all pins on these switches while in both their on and off
state (you can toggle the state of the switches as in this thread with
nothing connected to the HV line:
https://groups.google.com/d/msg/dropbot-dev/0R7FMsuSvDI/wCdfHjJ_fCIJ).

If one of the switches is giving a very low capacitance value, it is
likely not connected (or shorted to ground). If two neighboring
switches are bad (e.g., 76 and 77), they are potentially shorted
together.

In any case, you need to make sure at least one of your switching
boards is 100% free of shorts before the device load calibration will
give meaningful results. You may also need to reset the configuration
to defaults before doing the calibration if your initial parameters
are far from the optimum.

-Ryan

···

On Mon, Apr 20, 2015 at 5:37 PM, Linh Tran <dlin...@gmail.com> wrote:

Hi Ryan,

Somehow the other 2 calibrations are working now (loosen cables I guess).
Here are the results:
https://www.dropbox.com/s/cxblkt0ny1nuqv2/Mindy-LinhTran%20DropBot%20Calibration%20Results.zip?dl=0

Anti-aliasing circuit installed and checked; Auto adjust amplifier gain
checked. Sometimes the LIMIT light on the Amplifier turned ON but not
constantly. Normally it is OFF.

Do you have any thoughts?

Thank you!
-Linh.

--
You received this message because you are subscribed to the Google Groups
"dropbot-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to dropbot-dev...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


#11

Hi Linh/Mindy,

I was thinking more about the results you shared and decided to
compile some previous results we collected while troubleshooting
similar issues. I've added a new section to the wiki:

http://microfluidics.utoronto.ca/trac/dropbot/wiki/HighVoltageSwitchingBoard#Troubleshooting

On board 1 (channels 0-39), it looks to me like maybe you've got a
switch output that's shorted to HV or maybe one of the outputs is not
connecting to ground when the switch is off. Check channel 17 as it
looks like it may be an outlier.

Boards 2 and 3 look like they either have shorts, switches that are
not turning on, or broken relays that will need to be replaced.

It's probably a good idea to sort out these issues before applying
high voltage to your system (the "Test channels" diagnostic uses 10V).
I've added a note to the build instructions which makes this
recommendation. If you do have shorts on any of your switching boards,
performing the "reference" or "device load" calibration may cause
damage to the switching board (i.e., relays) and/or the amplifier. The
Trek amplifiers do have current limit protection, but it's not 100%
perfect, and we have fried more than one (a couple hundred bucks to
fix and a few weeks without an amplifier).

-Ryan

···

On Wed, Apr 22, 2015 at 4:44 PM, Ryan Fobel <ry...@fobel.net> wrote:

Hi Linh,

Looking at your switching board test results, there are still a few
switches that still look suspect:

17, 76, 77, 84, 91 and 108.

Please test all pins on these switches while in both their on and off
state (you can toggle the state of the switches as in this thread with
nothing connected to the HV line:
https://groups.google.com/d/msg/dropbot-dev/0R7FMsuSvDI/wCdfHjJ_fCIJ).

If one of the switches is giving a very low capacitance value, it is
likely not connected (or shorted to ground). If two neighboring
switches are bad (e.g., 76 and 77), they are potentially shorted
together.

In any case, you need to make sure at least one of your switching
boards is 100% free of shorts before the device load calibration will
give meaningful results. You may also need to reset the configuration
to defaults before doing the calibration if your initial parameters
are far from the optimum.

-Ryan

On Mon, Apr 20, 2015 at 5:37 PM, Linh Tran <dlin...@gmail.com> wrote:

Hi Ryan,

Somehow the other 2 calibrations are working now (loosen cables I guess).
Here are the results:
https://www.dropbox.com/s/cxblkt0ny1nuqv2/Mindy-LinhTran%20DropBot%20Calibration%20Results.zip?dl=0

Anti-aliasing circuit installed and checked; Auto adjust amplifier gain
checked. Sometimes the LIMIT light on the Amplifier turned ON but not
constantly. Normally it is OFF.

Do you have any thoughts?

Thank you!
-Linh.

--
You received this message because you are subscribed to the Google Groups
"dropbot-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to dropbot-dev...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


#12

Hi Ryan,

Thank you for your troubleshooting instructions. That is very helpful!

For all of our 3 boards, I have measured the voltages of the **pins 1-4 **(referred to the ground at the 2 middle pins of J1) and the resistance of the pin 5-8 for all switches and I got:

Do those values look normal for you? For me I guess it is not normal because in ON state, the GND should be 0V. So I have removed the switch CH76 & CH77 (which were giving no capacitance in the previous test) and measure the voltages again but nothing changed. And I am wondering why in OFF state, pin 1 was not 5V but only 1.37V.

Thanks,

Linh


#13

Hi Mindy,

Those values look reasonable to me. I’m not 100% clear which points some of your measurements refer to, so I’ve modified your images to include arrows pointing to where I think you’re measuring.

The important thing to calculate is the current through resistors R5 and R11. In the ON state, the top switch should be off and therefore the current through R5 should be close to 0 and the current through R11 should be ~10mA. You mentioned that the GND should be 0V in the ON state (I agree), but I’m assuming that the 3.29V you measured is on pin 4? I expect the GND node (the opposite end of R11) to be at 0V. Is that correct? If so, that would make the current through R11 ~9mA, which would indicate that the bottom half of the switch is in the ON state (which it appears to be, since the resistance between pins 5 and 6 is 76 Ohms). So far, this looks good to me.

I’m not clear where the 1.37 V measurement refers to, but if it’s at pin 1, that would make sense. In this case, the current through R5 would be ~10mA (which means that the top switch is on, in agreement with the 76 Ohms between pins 7 and 8). The bottom resistor, R11, looks like it has 0mA across it which is also the expected behaviour.

Assuming my interpretation of you’re measurements is correct, things looks ok to me. If the output pins have the correct resistance (~76 Ohms between the pair that is supposed to be on, and > 1 MOhm between the other pair), I wouldn’t expect you to have any problems.

Have you re-run the “Test channels” diagnostic after swapping out those relays? Did you see any changes?

-Ryan

···

On Fri, Apr 24, 2015 at 4:01 PM, Linh Tran dlin...@gmail.com wrote:

Hi Ryan,

Thank you for your troubleshooting instructions. That is very helpful!

For all of our 3 boards, I have measured the voltages of the **pins 1-4 **(referred to the ground at the 2 middle pins of J1) and the resistance of the pin 5-8 for all switches and I got:

Do those values look normal for you? For me I guess it is not normal because in ON state, the GND should be 0V. So I have removed the switch CH76 & CH77 (which were giving no capacitance in the previous test) and measure the voltages again but nothing changed. And I am wondering why in OFF state, pin 1 was not 5V but only 1.37V.

Thanks,

Linh

You received this message because you are subscribed to the Google Groups “dropbot-dev” group.

To unsubscribe from this group and stop receiving emails from it, send an email to dropbot-dev...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.


#14

Looking back at your results again, it looks to me like switching board 2 is the closest to getting up and going. If you’ve removed or replaced switches 76 & 77, it may be worthwhile to try setting the i2c address of switching board 2 to 32 (i.e., using it to replace switching board 1). Then disconnect the other 2 boards from your system (double-check the “Test channels” diagnostic again to confirm that there are no shorts), and try running the reference and device load calibration routines with that single switching board connected.

-Ryan

···

On Fri, Apr 24, 2015 at 5:18 PM, Ryan Fobel ry...@fobel.net wrote:

Hi Mindy,

Those values look reasonable to me. I’m not 100% clear which points some of your measurements refer to, so I’ve modified your images to include arrows pointing to where I think you’re measuring.

The important thing to calculate is the current through resistors R5 and R11. In the ON state, the top switch should be off and therefore the current through R5 should be close to 0 and the current through R11 should be ~10mA. You mentioned that the GND should be 0V in the ON state (I agree), but I’m assuming that the 3.29V you measured is on pin 4? I expect the GND node (the opposite end of R11) to be at 0V. Is that correct? If so, that would make the current through R11 ~9mA, which would indicate that the bottom half of the switch is in the ON state (which it appears to be, since the resistance between pins 5 and 6 is 76 Ohms). So far, this looks good to me.

I’m not clear where the 1.37 V measurement refers to, but if it’s at pin 1, that would make sense. In this case, the current through R5 would be ~10mA (which means that the top switch is on, in agreement with the 76 Ohms between pins 7 and 8). The bottom resistor, R11, looks like it has 0mA across it which is also the expected behaviour.

Assuming my interpretation of you’re measurements is correct, things looks ok to me. If the output pins have the correct resistance (~76 Ohms between the pair that is supposed to be on, and > 1 MOhm between the other pair), I wouldn’t expect you to have any problems.

Have you re-run the “Test channels” diagnostic after swapping out those relays? Did you see any changes?

-Ryan

On Fri, Apr 24, 2015 at 4:01 PM, Linh Tran dlin...@gmail.com wrote:

Hi Ryan,

Thank you for your troubleshooting instructions. That is very helpful!

For all of our 3 boards, I have measured the voltages of the **pins 1-4 **(referred to the ground at the 2 middle pins of J1) and the resistance of the pin 5-8 for all switches and I got:

Do those values look normal for you? For me I guess it is not normal because in ON state, the GND should be 0V. So I have removed the switch CH76 & CH77 (which were giving no capacitance in the previous test) and measure the voltages again but nothing changed. And I am wondering why in OFF state, pin 1 was not 5V but only 1.37V.

Thanks,

Linh

You received this message because you are subscribed to the Google Groups “dropbot-dev” group.

To unsubscribe from this group and stop receiving emails from it, send an email to dropbot-dev...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.


#15

HI Ryan,

Thanks for your super quick reply :slight_smile: Sorry for the confusion in my diagram but you were correct. The voltages were measured exactly at the pins (1 to 4) of the switch, referring to the ground (at the 2 center pins of the power supply connector J1)

I’ll try to do the test and calibration as you suggested with only that board on Monday and let you know when we chat.

-Linh.


#16

Hi Ryan,

I have replaced the relays which showed no capacitance in the channel test and that did fix the problems.

Board #2

Board #3

Board #2 - Device Load Test

Board #3 Device Load Test

However, for the board #1, even though all the channels have capacitance in the result of the channel test but the channels CH0-15 have ~20x capacitance than expected. To track down the cause of this, I have done the following:

  1. I observed that the problem happens with whole blocks of channels so the issue might be from the Shift Register ICs (U1-U5). So I have replaced the U1 (controls CH0-CH7) and repeated the channel test but nothing improved => eliminating the problem caused by shift registers.

  2. Because CH0 has the largest capacitance so I decided to remove the switch of CH0 and repeated the channel test but… the channel test was NOT able to detect this change (the result still showed a capacitance signal in CH0)! Then, I removed the switch of CH1 and did the test, again the channel test was not able to detect this change. The capacitance of CH0 and CH1 seemed to be decreased.

Board #1 - Initial state

Board #1 - After removing CH0

Board #1 - After removing CH0 and CH1

Thanks,

-Linh


#17

Alright, looks like you are getting close!

The shape of those curves tells me that there is a a baseline capacitance of ~15-20pF present even when all switches are off (channel 0 is supposed to be 1pF and the ratio of c_measured/c_expected is reading 15-20).

I can think of 2 possible causes:

1.) one of the channels is on when it is supposed to be off (probably one that has a capacitance of ~20pF, i.e., channel 18 +/- a couple of channels)
2.) one of the channels is not grounding properly when it is in the off state

To test for case 1, I would try turning off all switches, set the actuation voltage to 10 V, and probe the output of each channel on that board with an oscilloscope. All of them should measure 0V, but you may find one that measures 10 V.

If not, you can test for case 2 by turning off all switches and disconnecting the high voltage (or turn off the amplifier). Then use a multimeter to measure the resistance between each channel output and HVGND (you can probe the HVGND test point from one of the gold SMA connectors on the switching board).

-Ryan

···

On Wed, Apr 29, 2015 at 5:21 PM, Linh Tran dlin...@gmail.com wrote:

Hi Ryan,

I have replaced the relays which showed no capacitance in the channel test and that did fix the problems.

Board #2

Board #3

Board #2 - Device Load Test

Board #3 Device Load Test

However, for the board #1, even though all the channels have capacitance in the result of the channel test but the channels CH0-15 have ~20x capacitance than expected. To track down the cause of this, I have done the following:

  1. I observed that the problem happens with whole blocks of channels so the issue might be from the Shift Register ICs (U1-U5). So I have replaced the U1 (controls CH0-CH7) and repeated the channel test but nothing improved => eliminating the problem caused by shift registers.
  1. Because CH0 has the largest capacitance so I decided to remove the switch of CH0 and repeated the channel test but… the channel test was NOT able to detect this change (the result still showed a capacitance signal in CH0)! Then, I removed the switch of CH1 and did the test, again the channel test was not able to detect this change. The capacitance of CH0 and CH1 seemed to be decreased.

Board #1 - Initial state

Board #1 - After removing CH0

Board #1 - After removing CH0 and CH1

Thanks,

-Linh

You received this message because you are subscribed to the Google Groups “dropbot-dev” group.

To unsubscribe from this group and stop receiving emails from it, send an email to dropbot-dev...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.


#18

Hi Ryan,

Today I have a chance to work on the boards. I have done the troubleshooting like you instructed and it works now! :slight_smile: Here is the result:

Unfortunately, I have encountered another problem when I set back the address for the board #2. Usually when I open the Serial Monitor in Arduino, the below message would show up and I use this command to set the address: set_i2c_address(x)

However, this time there was nothing showing up so I tried to use the FTDI cable to re-flash the firmware. And I got this error:

I also burned a bootloader to the ATMega328P chip again (successfully) but still stuck with the above error. I have set the right board (Uno) & port as well as run Arduino as administrator.

Do you have any thoughts on this problem?

Thanks,

Linh


#19

Hi Linh,

Is it not the USB power jumper, do you have power when you upload the sketch? When you burn the bootloader the chip is powered through the isp, so you wouldnt notice if there is no power. You would also get this error if the reset is not connected to the FTDI.

Hope this helps!

Best,

Csaba

···

On Wednesday, May 6, 2015 at 8:40:24 PM UTC-4, Linh Tran wrote:

Hi Ryan,

Today I have a chance to work on the boards. I have done the troubleshooting like you instructed and it works now! :slight_smile: Here is the result:

Unfortunately, I have encountered another problem when I set back the address for the board #2. Usually when I open the Serial Monitor in Arduino, the below message would show up and I use this command to set the address: set_i2c_address(x)

However, this time there was nothing showing up so I tried to use the FTDI cable to re-flash the firmware. And I got this error:

I also burned a bootloader to the ATMega328P chip again (successfully) but still stuck with the above error. I have set the right board (Uno) & port as well as run Arduino as administrator.

Do you have any thoughts on this problem?

Thanks,

Linh


#20

Thanks Csaba, I found out that somehow I had burned the IC so I just replaced it with the new IC and now it can be flashed.

-Linh.

···

On Thursday, May 14, 2015 at 6:19:17 PM UTC-7, Csaba Konrád wrote:

Hi Linh,

Is it not the USB power jumper, do you have power when you upload the sketch? When you burn the bootloader the chip is powered through the isp, so you wouldnt notice if there is no power. You would also get this error if the reset is not connected to the FTDI.

Hope this helps!

Best,

Csaba

On Wednesday, May 6, 2015 at 8:40:24 PM UTC-4, Linh Tran wrote:

Hi Ryan,

Today I have a chance to work on the boards. I have done the troubleshooting like you instructed and it works now! :slight_smile: Here is the result:

Unfortunately, I have encountered another problem when I set back the address for the board #2. Usually when I open the Serial Monitor in Arduino, the below message would show up and I use this command to set the address: set_i2c_address(x)

However, this time there was nothing showing up so I tried to use the FTDI cable to re-flash the firmware. And I got this error:

I also burned a bootloader to the ATMega328P chip again (successfully) but still stuck with the above error. I have set the right board (Uno) & port as well as run Arduino as administrator.

Do you have any thoughts on this problem?

Thanks,

Linh