Got my GDM-8251 from eBay today...one question

[alank2]

It had some scratches in the display bezel which I'm not thrilled about, but I guess it is what it is.

Now for my question:

I prepared a 50Hz 5V RMS signal from my SDG2082X with 2V DC offset and decided to try AC on the first display and DC on the second display.  In this case, it displays 30mV!!!  My Fluke 8845A has no problem showing 2V.  It seems to be very sensitive to the frequency.  If I change it to 1kHz, it goes to 58mV.  At 10k it is fine.

Now if I switch it back to 1K, it is fine, back to 50 Hz, it is fine.  It is an auto ranging type of problem it seems.  Same thing happens in DC mode.

Firmware is V2.30 - what is the latest?

[xrunner]

I'm going to use your thread to post some things abut this meter, as I'm trying to get the firmware and do that upgrade (still waiting on the file from GWinstek).

Firmware is V2.30 - what is the latest?

2.3

I prepared a 50Hz 5V RMS signal from my SDG2082X with 2V DC offset and decided to try AC on the first display and DC on the second display.  In this case, it displays 30mV!!!  My Fluke 8845A has no problem showing 2V.  It seems to be very sensitive to the frequency.  If I change it to 1kHz, it goes to 58mV.  At 10k it is fine.

Now if I switch it back to 1K, it is fine, back to 50 Hz, it is fine.  It is an auto ranging type of problem it seems.  Same thing happens in DC mode.

OK, I tried it as you did, and got the same results, but check this - after you use 10 kHz, go right back down to 50 Hz. On my unit it will then readout the correct DCV number after that. 

It does read a freq. of 50 Hz just fine also.

Also, starting from the beginning, in combined ACV + DCV, it will readout the correct value 5.38V

sqrt (AC² + DC²)

[alank2]

OK, I tried it as you did, and got the same results, but check this - after you use 10 kHz, go right back down to 50 Hz. On my unit it will then readout the correct DCV number after that. 

Yes, because it is in the right range at that point.  The problem is that with the slower frequencies, it doesn't select the correct range initially.  If you up the frequency and then it does, when you lower the frequency, it will stay.

[xrunner]

Yes, because it is in the right range at that point.  The problem is that with the slower frequencies, it doesn't select the correct range initially.  If you up the frequency and then it does, when you lower the frequency, it will stay.

Hmm, but that can't be the reason either, because if you use DC offsets in the mV range in the beginning, it doesn't display the correct value either, at least on mine.

[xrunner]

Well, having played with this now for an hour, it's pretty obvious that it works OK in auto range (displays the correct DCV) with an initial freq. of at least 2 kHz or higher. It also displays the correct DC or AC voltage reading if not in a dual-display mode, and it also correctly displays the calculated value of a single combined reading of ACV + DCV.

And, since you at least, have the latest firmware ...

Since it's obviously gone through several firmware revisions, and since, even though we are pretty smart people here, it's doubtful we are the only ones who have noticed this since it's been out.

The answer then is that it's acting within it's design parameters. The issue we observe - a dual-display ACV and DCV reading issue of the DCV being wrong (below 2 kHz initially)  --> in auto-ranging mode only, is a specification/limitation that is not explicitly listed in the manual. Many pieces of test gear have limitations and bugs, this is a limitation for this meter.

But, as you realize, and which works, is to simply acknowledge the issue and turn off auto ranging for the DCV second display. All you have to do is up-range it to a higher range and you get the correct reading. 

[alank2]

Response from GW Instek:

>The issue is caused by dynamic range. For GDM-8255A/8251A, customer need to manual set the range for it.
>The newer model like GDM-834x or GDM-8351 has already solve this issue.

[xrunner]

Response from GW Instek:

>The issue is caused by dynamic range. For GDM-8255A/8251A, customer need to manual set the range for it.
>The newer model like GDM-834x or GDM-8351 has already solve this issue.

Yep.

I ordered the GDM-01 cal key from Tequipment, but then found out that they couldn't ship it until March 28, so I had to cancel that order. I found the key for sale on Ebay so it should be here in a few days. I also got the firmware file from GW Instek, so I'm ready to try the upgrade as soon as I get the cal key, via the RS-232 serial connection.

[coromonadalix]

@xrunner   

Could you check  when the usb cal key arrive, if you have a short between pin 1 and 4  with an ohm meter ...  and report the results.

thks

[Housedad]

is it the 8251 or the 8251A you have?

[coromonadalix]

GDM-8251A     the cal key GDM-01 should be the same between these two I've been told

thks

On a few web sites  they sold it for 8 to 10$   but the minimum shipping fees are 23$   for a mere  few grams ...  sent email to gw instek support

[xrunner]

@xrunner   

Could you check  when the usb cal key arrive, if you have a short between pin 1 and 4  with an ohm meter ...  and report the results.

OK I just got up, stand by ...

On a few web sites  they sold it for 8 to 10$   but the minimum shipping fees are 23$   for a mere  few grams ...  sent email to gw instek support

I got one off Ebay for $6.88 + $10 shipping. I agree the shipping is stupid as it could be sent in a small envelope. 

GDM-01

I did complete the firmware update yesterday and all went according to the document, now I'm at v2.3.

[xrunner]

OK I checked pins 1 & 4 (the outside pins) of the cal key - they are essentially shorted together. I also plugged in a USB test connector I have into the back of the DMM into the cal key USB slot, and I measured 4.88V between pins 1 & 4 (the outside pins). So the cal key is connecting 5V back to a sense terminal of some component inside?

I'll be interested in what you all do with this info. 

[alank2]

Wait, let me get this straight, they decided on a design decision to short the + and - lines in a USB A connector?  Is that right?

[xrunner]

Wait, let me get this straight, they decided on a design decision to short the + and - lines in a USB A connector?  Is that right?

  Apparently - yes they did.   

[edpalmer42]

I was looking at the specs for the 8251A and there seems to be some confusion on the input resistance.  The GW website shows 10M for DC and 1.1 M for AC.    The manual doesn't say anything about input resistance.  But there's a datasheet floating around that reverses the two numbers!    For a 5.5 digit meter, 10M is marginal.  1.1 M is depressing. 

Can someone measure their 8251A and report the results?

[xrunner]

I was looking at the specs for the 8251A and there seems to be some confusion on the input resistance.  The GW website shows 10M for DC and 1.1 M for AC.    The manual doesn't say anything about input resistance.  But there's a datasheet floating around that reverses the two numbers!    For a 5.5 digit meter, 10M is marginal.  1.1 M is depressing. 

Can someone measure their 8251A and report the results?

Just measured it using an external 1M R and 10.0V. The input R is 10M.

[sequoia]

Wait, let me get this straight, they decided on a design decision to short the + and - lines in a USB A connector?  Is that right?

  Apparently - yes they did.   

This is probably why there is this note in the firmware update procedure document:

"Check to ensure that the CAL Key is inserted into the CAL Key port and NOT the USB port. Failure to do so may cause failure of USB function."

 


Seriously, if a device has decent USB port implementation it should survive plugging this "CAL Key" into it, but I wouldn't be surprised if there is lot of computers and other devices nowadays that don't have any protections on USB ports....

[xrunner]

This is probably why there is this note in the firmware update procedure document:

"Check to ensure that the CAL Key is inserted into the CAL Key port and NOT the USB port. Failure to do so may cause failure of USB function."

 

Yea kind of a dumb design. You make the design with two identical ports next to each other with the knowledge that the electrical design with the cal key (a simple short) stuck in the "real" USB port "may cause failure of USB function".

People get in a hurry, it's on the back and might not be well lit, and you think just stick a blurb in the manual and call it a day. C'mon Gw Instek - why not just use the B type USB port, or some other type of port?

[coromonadalix]

thks for the tips, saved me money and i will be able to calibrate it. 

[cdev]

Does this meter measure inductance in any way? Also, how precise is the capacitance measurement - does anybody know what frequency its done at, and does it measure ESR?

[xrunner]

Does this meter measure inductance in any way? Also, how precise is the capacitance measurement -

It don't do none a' that stuff ...

But it does do a lot of other nice measurements -

http://www.gwinstek.com/en-global/root/Discontinued_Products/Discontinued_Meters_LCR/GDM-8251A

[xrunner]

thks for the tips, saved me money and i will be able to calibrate it. 

Cool beans, wish I'd have known about the trick. Ah well, it's all good. 

[bitseeker]

Kudos to xrunner for taking one for the team and sharing the solution to the cal key.

Plus, we get to have a good laugh at Instek for a very poorly conceived key "solution".

[SeanB]

Could they not have put a 1k or so resistor in series with the "key" lines. would solve that problem, and I doubt the cal input on the meter is drawing any current other than that for a pull down resistor keeping the line low. They could have put a red LED on the cal key that lights up to show it is in Cal mode, wank gimmick, but useful, and free as you already have the USB socket needing a small PCB, just add the 2 resistors and put the LED out the rear of the overmould where you normally have that cable, saves needing a non standard injection mould in production.

[bitseeker]

They could've put a resistor between pins 1 and 4 to prevent a complete short circuit when plugged into a real USB port. In one of the related threads, someone used a 100 Ohm resistor in their DIY cal key, IIRC.