A look at the Uni-T UT210E

[flywheelz]

Remember that problem when you Zero the meter in 2A mode and it reverts to 2000 count. Well,  in the dotless mode you get 6000 count or whatever you count you set in eeprom.  So you will see say 5670 which means 5.670amps.

[hugatry]

Adding bytes 1C and 1D to locations A7 and B7 will add 6000(/2000)DCA and 6000(/2000)ACA modes to 2A switch position. (For reference, check attachment "BytesModes.jpg": https://www.eevblog.com/forum/testgear/a-look-at-the-uni-t-ut210e/msg925581/#msg925581)

The firmware in the OTP memory hasn't been programmed to make these modes revert back to 2000 counts after zeroing, because these modes are not used in the meter, unlike the lower ranges. The IC doesn't actually know if it measuring 6 or 6000 amps, rest of the circuit will handle scaling the input... Tell IC that you are measuring higher range, but don't change anything else -> decimal point moves, making reading 10x, 100x, 1000x higher/lower.

[bitseeker]

Remember that problem when you Zero the meter in 2A mode and it reverts to 2000 count. Well,  in the dotless mode you get 6000 count or whatever you count you set in eeprom.  So you will see say 5670 which means 5.670amps.

Yes, I remember the count reset issue. However, I missed whatever discussion there was of dotless mode. Now it makes sense. Thanks.

The firmware in the OTP memory hasn't been programmed to make these modes revert back to 2000 counts after zeroing, because these modes are not used in the meter, unlike the lower ranges. The IC doesn't actually know if it measuring 6 or 6000 amps, rest of the circuit will handle scaling the input... Tell IC that you are measuring higher range, but don't change anything else -> decimal point moves, making reading 10x, 100x, 1000x higher/lower.

Got it. Thanks for the clarification.

[DAIRVINE]

For DC why not go for 9999 or even 12000 if it will display over-range?
The useful thing about the meter is 1mA DC current clamp, so we should optimise for that??

On AC we can know that we lose accuracy above 6000 counts or even lower if there is a big crest factor. Though could maybe read a square wave above this?

[flywheelz]

There are only have 4 digits so 9999 is possible but I remember someone posting here that maximum should be something around 7000 because the mcu is not able to accurately count past that. 

[bitseeker]

From what I recall, and what DAIRVINE was referring to, there is an issue if you extend the maximum count too far for AC because the value displayed is RMS (i.e., you have to leave headroom for the peak). So, 7000-8000 would be safe there.

However, for DC, unless there is excessive noise, non-linearity, or some other issue at the upper end of the range, it might be fine to 9999 (maximum displayable value).

When I get around to modding mine, I'll probably set everything to 6000 or 7000 count for simplicity.

[joeqsmith]

Anyone willing to test two new eeprom mods?  It gives two new functions in 2A mode.  Its reported on the Russian site that these functions still have the full count even after zeroing out (i.e. doesn't go back to 2000).  The LCD does not show a "." in the read out.

Update:  Tested the DCA 0000 function in 2A and it truly works even after zeroing out. 

Code: [Select]

       old      new
 A7: 00 1C - DCA 0000 in 2A mode / SELECT slot 3
 B7: 00 1D - ACA 0000 in 2A mode / SELECT slot 4

That group is really digging into it.  So is the decimal gone as they suggested?

With my programmer working again, I tried these changes.  Currently I have the meter default to DCA/DCV. 

If I turn on the 210E, then select the 2A mode, the decimal is shown.  Zero out the meter, the decimal is still shown.  Putting more than 2A through the meter will over range it. 

Turn the meter back off and on, then again select the 2A mode.  Don't zero out the meter and place more than 2A through the meter.  It does not over range and reads the correct value.

Turn the meter back off and on, then again select the 2A mode.  Select AC, the decimal is shown.  Now select DC, the decimal goes away.  Don't zero out the meter and place more than 2A through the meter.  It does not over range and reads the correct value.  However the meter now starts beeping at about 600mA. 

Turn the meter back off and on, then again select the 2A mode.  Select AC and then DC.  Zero out the meter and place more than 2A through the meter.  It does not over range and reads the correct value.  However the meter still beeps at about 600mA and there is no decimal.

So it's not a real clean setup IMO.   Maybe I missed something.

[flywheelz]

Turn the meter back off and on, then again select the 2A mode.  Select AC, the decimal is shown.  Now select DC, the decimal goes away.  Don't zero out the meter and place more than 2A through the meter.  It does not over range and reads the correct value.  However the meter now starts beeping at about 600mA. 

Turn the meter back off and on, then again select the 2A mode.  Select AC and then DC.  Zero out the meter and place more than 2A through the meter.  It does not over range and reads the correct value.  However the meter still beeps at about 600mA and there is no decimal.

So it's not a real clean setup IMO.   Maybe I missed something.

In post #182 I wrote some info about beeping.  Perhaps you can try set FFh at address 19h.  Are you sure its starts beeping with 600mA or 6000 which is actually 6A? 

[hugatry]

...
However the meter still beeps at about 600mA and there is no decimal.

In post #182 I wrote some info about beeping.  Perhaps you can try set FFh at address 19h.  Are you sure its starts beeping with 600mA or 6000 which is actually 6A? 

Correct location would be 0x1C. That sets the alarm limit in hundreds of milliamps for "dotless 2A" mode. Value is originally 0x0A or 10, making the limit 10*100mA = 1A = "1000" on the display.

So you can have the alarm in volts mode when measuring mains, without having ridiculously low alarm limit in current mode. It was probably just coincidence that original value was so close to the value you set for voltage mode.

[flywheelz]

...
However the meter still beeps at about 600mA and there is no decimal.

In post #182 I wrote some info about beeping.  Perhaps you can try set FFh at address 19h.  Are you sure its starts beeping with 600mA or 6000 which is actually 6A? 

Correct location would be 0x1C. That sets the alarm limit in hundreds of milliamps for "dotless 2A" mode. Value is originally 0x0A or 10, making the limit 10*100mA = 1A = "1000" on the display.

So you can have the alarm in volts mode when measuring mains, without having ridiculously low alarm limit in current mode. It was probably just coincidence that original value was so close to the value you set for voltage mode.

Thanks for clarifying it.  I test the address 19h is for V mode.  Now I wonder what 16h to 18h affect.

[hugatry]

Thanks for clarifying it.  I test the address 19h is for V mode.  Now I wonder what 16h to 18h affect.

16h - "OL" limit for DCV.
17h - "OL" limit for DCA.
18h - Alarm/beep limit for DCV.

For each of those, limit = value * 10V.

Sources:

• Kerry Wong's translation of the main IC's datasheet:
  http://www.kerrywong.com/blog/wp-content/uploads/2016/04/DTM0660DataSheet.pdf
• Chinese datasheet, which was shared by user Puomjw in the thread of Russian forum that has been discussed earlier.

[joeqsmith]

...
However the meter still beeps at about 600mA and there is no decimal.

In post #182 I wrote some info about beeping.  Perhaps you can try set FFh at address 19h.  Are you sure its starts beeping with 600mA or 6000 which is actually 6A? 

Correct location would be 0x1C. That sets the alarm limit in hundreds of milliamps for "dotless 2A" mode. Value is originally 0x0A or 10, making the limit 10*100mA = 1A = "1000" on the display.

So you can have the alarm in volts mode when measuring mains, without having ridiculously low alarm limit in current mode. It was probably just coincidence that original value was so close to the value you set for voltage mode.

Yea, I still had it set to 1A.   Turned off all the alarms.  Too bad the thing just didn't handle the zero in normal without reverting back to 2K counts.    Not a bad work around (the dotless mode). 

[DAIRVINE]

This post seems to imply the maximum ADC is 12000.
Is it possible to set the max count to 11999? Is there some sensible indication of above 9999?
 For ACA its obviously a problem, but I don't really need ACA.
It may also be a problem with capacitance. Can anyone determine capacitance limits?

I spent a little time playing with modifying the EEPROM.

First, a comment about setting up 9999 counts. This absolutely works fine for DCV and resistance measurements. For ACV (or ACA), there is an issue. The problem is caused by the fact that this meter uses fast sampling and DSP to calculate the RMS value of an AC input, rather than using the traditional approach of an RMS-to-DC converter (e.g. AD637). The peak value of a sine wave is 1.414 the RMS value. So to get an RMS reading on the meter of say 9.999 VAC, the meter needs to sample instantaneous voltages up to 14.141 V. It can't. The ADC tops out at roughly 12000 counts. This means that, for AC, the maximum displayed counts should be limited to around 8000 count. This allows enough headroom in the ADC to sample the peaks of sine waves (and clipped sine waves which you will likely see on AC mains). Still, any waveform with a higher crest factor may show a RMS reading below actual, when the reading is reaching the top of a range.  A workaround might have been to manually range higher, but this meter has no range button, so you can't do that. All this was to say that I recommend leaving the maximum counts to 8000 or less.

I have also used what I learned here (big thanks to hugatry) to modify another DTM0660-based meter that I own. On that meter I added a mV range with high impedance input.

[joeqsmith]

There appears to be another difference with the dotless mode.  Curious if this is normal.

Using a fixed current source of 2.0A.  Power up the 210E. place into DCA 2A scale.  Zero as normal and attach the clamp.  Write down your reading.   Remove the clamp and set to dotless mode.  Zero and attach the clamp and write down this reading.

Now power cycle the meter and go back to 2ADC range.  Don't zero and attach the clamp.  Take the reading.  Now without disconnecting the clamp, switch to dotless mode and take the reading.

It appears there are different calibrations for dotless and normal.  With dotless mode and 1A applied I get about 38mA of error.   

[hugatry]

There appears to be another difference with the dotless mode.  Curious if this is normal.

Using a fixed current source of 2.0A.  Power up the 210E. place into DCA 2A scale.  Zero as normal and attach the clamp.  Write down your reading.   Remove the clamp and set to dotless mode.  Zero and attach the clamp and write down this reading.

Now power cycle the meter and go back to 2ADC range.  Don't zero and attach the clamp.  Take the reading.  Now without disconnecting the clamp, switch to dotless mode and take the reading.

It appears there are different calibrations for dotless and normal.  With dotless mode and 1A applied I get about 38mA of error.

I would have totally missed that, if you hadn't mentioned it!

Copy bytes from 50h and 51h (locations of 2A calibration data) to locations 56h and 57h (2000A/ dotless 2A calibration data). That should fix it.

[joeqsmith]

There appears to be another difference with the dotless mode.  Curious if this is normal.

Using a fixed current source of 2.0A.  Power up the 210E. place into DCA 2A scale.  Zero as normal and attach the clamp.  Write down your reading.   Remove the clamp and set to dotless mode.  Zero and attach the clamp and write down this reading.

Now power cycle the meter and go back to 2ADC range.  Don't zero and attach the clamp.  Take the reading.  Now without disconnecting the clamp, switch to dotless mode and take the reading.

It appears there are different calibrations for dotless and normal.  With dotless mode and 1A applied I get about 38mA of error.

I would have totally missed that, if you hadn't mentioned it!

Copy bytes from 50h and 51h (locations of 2A calibration data) to locations 56h and 57h (2000A/ dotless 2A calibration data). That should fix it.

Just tried it and yes it does work.  Thanks.

[DAIRVINE]

 
DTM0660L datasheet (Via Kerry Wong website) includes:
DC Voltage: 60.00mV/600.0mV, (600.0mV)/6.000V/60.00V/600.0V/1010V • AC Voltage: 60.00mV/600.0mV, (600.0mV)/6.000V/60.00V/600.0V/750V • DC Current: 600.0uA/6000uA, 60.00mA/600.0mA, 6.000A/60.0A • AC Current: 600.0uA/6000uA, 60.00mA/600.0mA, 6.000A/60.0A • Resistance: 600.0?/6.000k?/60.00k?/600.0k?/6.000M?/60.00M? • Capacitance: 9.999nF/99.99nF/999.9nF/9.999uF/99.99uF/999.9uF/9.999mF/99.99mF • Frequency: 9.999Hz/99.99Hz/999.9Hz/9.999kHz/99.99kHz/999.9kHz/9.999MHz • Duty cycle: 1% ~ 99% • Diode: 0.000V ~ 3.000V. OL when above 3.0V • Continuity: Beep when < 60?, OL when > 600? • Clamp current meter: user can set range, decimal. Single or dual automatic ranges. • Temperature: C/F • Transistor: 0 ~ 2000 (hFE) • Non-contact Voltage detection (NCV)

DC volts 60mV range: Is this settable?
Frequency: Is this settable?

[joeqsmith]

Used it enough now and it proven to be very useful but I have jacked it up enough times that it's time to make a degauss system for it.   

Not planning on anything real special.  AC line with a switch and PTC to provide the decay.  Doing it by hand now.    Starting out making the coil form.  Cut and machined some phenolic.   

[joeqsmith]

Epoxied the one side and let it harden.  Then attached the other.  Will let it dry for a couple of days.   In the mean time, still have some of the plywood left over from building the transient generators that I plan to use for the case. 

If you have made one, post some pictures. 

[bitseeker]

Fancy toroidal glue up weight.

[Macbeth]

De-gauss. I remember that hand held white tube "tool" well as an apprentice. Plug it right in the 240V mains, you could feel the AC throb as you scanned it across a CRT to get rid of all those magenta patches. It really didn't feel safe to me especially as it warmed up!

Oh those CRT's had a factory fit degauss strap around them wired direct to mains through a PTC anyway. The good old days!

[bitseeker]

Joe, what's that cool-looking shiny thing in the background (two handles, several buttons)?

[joeqsmith]

Joe, what's that cool-looking shiny thing in the background (two handles, several buttons)?

Visit here often?   

[joeqsmith]

You could use a large transformer like the one I showed with a single turn.  Variac to do the decay.  Lots of ways to set something up.   Mines just a line cord, fuse, PTC, pushbutton, indicator and coil.    PTC gets me about a 200ms decay.  The PTC sits inside the Pyropel. 

[bitseeker]

Joe, what's that cool-looking shiny thing in the background (two handles, several buttons)?

Visit here often?   

Still not enough, apparently.

Thanks, Joe. Beautiful work. Going to check out your other vids too.