New Keithley DMM6500 and now DAQ6510

[Mr. Scram]

Touch screen display and higher processing power created the need for improved cooling over the other instruments you mentioned. The fan type, location and speed was carefully chosen to be considered quiet enough in an office environment for *most* people. Of course there are those who can still hear it and be bothered by it. To that end, individual opinions about it generally won't be helpful to another individual. YMMV. Hope this is helpful.

I understand choices had to be made, it still means adding a fan to the lab. More importantly, it's adding a fan to a device which can be used for extended periods of time. It's not the end of the world as some other devices have fans too, but it's definitely a consideration. I also do understand that noise is very personal. I can stand some, but try to minimize it whenever I can. I know I tend to use noisy devices less, often unconsciously. Your posts here are much appreciated. I really welcome the engagement and helpful replies. It's good to see engineers care about the products they make.

[PTR_1275]

If I’m in the home office and turn on the DMM6500 I can definitely hear it.

But turn on the rigol dp832, any of my HP 6632b power supplies or for a lot of noise my 6032, 6012, 6011 or any of the several kw dc loads and the noise is lost.

At the end of the day, I don’t think I’ve used the dmm6500 many times without another piece of equipment going.

[KedasProbe]

Saying I have another device that is even louder on my bench will never get us silent designed devices.
The general office noise gets louder so it's 'ok' to add another loud device in the office....
Everyone is starting to wear noise-canceling headphones to 'fix' it.

It's certainly possible to do it look at PC building most of the time silent only for intense processing there is a fan. This means you can leave it on like it should (since it has a warm up time) for simple measurement and people will not mind the fan speed up for heavy tasks.

De 50/100Hz hum is not something I can forgive, that's just going too cheap or bad design. 
I wonder if there was an engineer going "I found one we saved 5 dollar", "good job because hum is not in our spec sheet!"  And I assume it will only get louder over time.
(FYI I can hear the hum even with my WH-1000XM3 headphones on but it's gone as soon as I power the headset on, no music)

Since more people will be opening up their DMM for a battery replacement, we may as well trying to figure out a good modification for the fan noise, since it was said it blows harder than needed.

E-Design, is there a command to get the temperature of the hot parts that need this cooling?
or maybe just an warning flag in diagnostics. This way we can verify if a modification is ok or not.
Even external modification like closing x% of the holes can reduce the noise much.
(We can make it silent enough but it's harder to know if the temperature is ok or not, unless we have a probe)

The hum/noise source in one picture: https://xdevs.com/doc/Keithley/DMM6500/lukier/expconn.jpg
edit: http://www.sunon-fan.com/uploads/sunon_Date/pdf/MF50100V2-1000C-G99.pdf

[Mr. Scram]

What's the story with the battery? Is there any recourse for people who are stuck with a prematurely drained battery?

[KedasProbe]

The transformer hum was an issue on early units.. when did you purchase yours? Did you contact anybody about a replacement at that time? Transformer hum has been resolved for some time now.

Its first calibration was on 4/14/2019 and it's my understanding that the Hum problem should be fixed, so I assume the 'fix' is present and it was even worse before. (I have no absolute reference since there is no hum spec)

or is it possible that some old unit got returned and calibrated again and sold, or a batch without fix?

Since you asked, there might be a command to get an internal temperature (not sure a user accessible one, I could check) but you asked about temperature of hot parts? - no there isn't one that can tell you about what parts are hot. Most of the measurement circuits are actually not hot, but rather want a temperature stability in their area.

Any indicator that the device is still working within spec as before the change.
You think a lower CFM would make it go out of spec?

Also, transformer hum and fan noise are two different sources of audible noise and not related.

Obviously

[KedasProbe]

Such 0V/100mV test seems fair enough.

Any extra hum advice?
I see the battery is within 3year warranty so maybe I take the opportunity to let them check the hum, if nothing about the hum is changed at least the battery is replaced.

[Mr. Scram]

Has anyone had their DMM6500 or DAQ6510 calibrated yet and what did it cost you? I don't have much experience with Keithley calibrations but I hear they can be comparatively expensive.

[JxR]

Has anyone had their DMM6500 or DAQ6510 calibrated yet and what did it cost you? I don't have much experience with Keithley calibrations but I hear they can be comparatively expensive.

This was pulled from the Tek.com website


I thought I remember reading in the manual/data-sheet that the calibration period for the DMM6500 was 2-years.  I know mine was factory cal'd in 10/2008, so it isn't technically due yet.  Tek's website is now saying recommended cal period is 1-year.  Maybe they changed their mind

I'm considering putting mine up for sale...

[jancumps]

Has anyone had their DMM6500 or DAQ6510 calibrated yet and what did it cost you? I don't have much experience with Keithley calibrations but I hear they can be comparatively expensive.

Welectron asks 147.06 € excl. VAT.
VAT 19%
free shipping within Germany. I don't know the price for shipping outside Germany.

I plan to drive to them. Not that far away from Brussels and I can turn it into a nice day trip.

[Mr. Scram]

This was pulled from the Tek.com website
(Attachment Link)

I thought I remember reading in the manual/data-sheet that the calibration period for the DMM6500 was 2-years.  I know mine was factory cal'd in 10/2008, so it isn't technically due yet.  Tek's website is now saying recommended cal period is 1-year.  Maybe they changed their mind

I'm considering putting mine up for sale...

That seems to coincide with the notion Keithley is comparatively expensive to maintain. Spare parts seem to be quite expensive too. It's surprising how much manufacturers differ in this regard. TTi is suprisingly reasonable. Keysight often has very affordable spares. Keithley seems to be quite expensive to the point of regularly not making sense and Fluke is just ridiculous, if they even grant you the privilege of buying a spare at all. Guess that's the Danaher Tax.

[KedasProbe]

At the office we only calibrate one 6.5digit yearly.
So we can use that one to check if all the other DMMs are still ok or use the 2 year specs in case of the DMM6500.

When we return to normal office work after covid-19 I will check my office and home DMM6500 against each other again, like I did when I purchased them, it's over 1 year now.

[Mr. Scram]

Is there any news in regards to fixing the issue with simultaneous VI measurements? I understand this requires a hardware revision. The ability to do power measurements in a single go seems quite useful.

[Kleinstein]

There is a principle problem in measuring voltage and current at the same time: they share a common terminal - so there will always be some error from the current flow. It looks like in the current HW this error is relatively large (e.g. quite some cable and maybe even more to the internal common ground point. Chances are it can get better by not perfect, unless one has an extra terminal.

There are already 2 ADCs: the fast (likely SAR chip) and slow one (likely multi-slope). It may be possible to used both at the same time if they don't need the fast one as part of the slow conversions. With some averaging the fast ADC may be good enough for the anyway limited accuracy in the combined mode.

[KedasProbe]

Can you provide the tolerances now of measuring current and voltage at the same time?
(for each current range)

[SaKhan]

Just got one for my birthday from the Swiss official distributor. I visited them, we had a chat and although they weren't aware of the battery drain issue, they checked their stock and gave me a very recently produced unit (12.04.2020). Really nice guys.

[hwj-d]

Just got one for my birthday from the Swiss official distributor. I visited them, we had a chat and although they weren't aware of the battery drain issue, they checked their stock and gave me a very recently produced unit (12.04.2020). Really nice guys.

Did you get a detailed first calibration report?

[SaKhan]

Just got one for my birthday from the Swiss official distributor. I visited them, we had a chat and although they weren't aware of the battery drain issue, they checked their stock and gave me a very recently produced unit (12.04.2020). Really nice guys.

Did you get a detailed first calibration report?

The certificate inside the box states only that the meter is in calibration. I haven't requested any calibration services.

[hwj-d]

Just got one for my birthday from the Swiss official distributor. I visited them, we had a chat and although they weren't aware of the battery drain issue, they checked their stock and gave me a very recently produced unit (12.04.2020). Really nice guys.

Did you get a detailed first calibration report?

The certificate inside the box states only that the meter is in calibration. I haven't requested any calibration services.

This is usually part of buying a 6.5 DMM. Keithley takes the liberty of withholding the calibration data, which must be available, as an extra service. Other well-known manufacturers of prof. DMM naturally supply such a thing. In my opinion this is not so serious.

[Kleinstein]

With modern meters, that are adjusted in software the test / cal just after the adjustment of the ranges should be very close to nominal. So there would be limited values to most of the numbers. If at all the AC ranges may show a little more, as not every test point may have it's own corresponding CAL constant. It would be interesting only if they do the adjustment in the factory and do the calibration / check just before delivery, so with more time in between.

[hwj-d]

With modern meters, that are adjusted in software the test / cal just after the adjustment of the ranges should be very close to nominal. So there would be limited values to most of the numbers. If at all the AC ranges may show a little more, as not every test point may have it's own corresponding CAL constant. It would be interesting only if they do the adjustment in the factory and do the calibration / check just before delivery, so with more time in between.

Yes, maybe I'm a bit biased as I have no other choice but to measure my LTZ1000 references with the DMM6500, which is of course a bit tricky anyway as I measure beyond the given nominal values. The more exact data of the initial adjustment and calibration would help there, even the dmm is comming from relative stock.

[exe]

a non burned in LM399

I'm curious, do these references go through any selection process, such as a temperature coefficient or voltage?

PS crazy idea, is it possible to downclock the unit so the front panel creates less heating and doesn't need as much cooling?

[MegaVolt]

PS crazy idea, is it possible to downclock the unit so the front panel creates less heating and doesn't need as much cooling?

In the brightness settings of the display there is an item: "Display, key lights, and all indicators off: BLACkout "
My experience is that when the temperature of the front panel significantly decreases with time.
But you have to work with the device through a remote interface.

[eplpwr]

It was found that most users / customers did not have any use for the full data report so that practice stopped in order to save processing time / cost. However, some people would still like the full report so it is available as a paid service and ordering option.

Sorry, but this "trust us" argument just doesn't cut it. After nagging support i got the cal reports for my DMM7510 and DAQ6510 respectively. Keysight gives you the factory cal report in an enclosed letter, no fuss. Who, in their right mind, would not want to have the "starting point", e.g. the initial factory calibration, of a meter with 6.5 digits or better resolution? Are customers using it as a interior decoration because of the nice colour screen, or somesuch? Most probable it's marketing dept BS made up to create the illusion of meeting customer expectations, while the real motive is increasing profit (saving a few minutes in mfg, getting some sheepish customers to pay extra for a report they should have received upfront). This is exactly like the statement "everybody has battery anxiety with electric cars, so we will only supply the cars on a lease basis. No one wants to outright buy the cars, including the 'risky' batteries, so we don't offer that option" - yeah, sure.

OTOH, I think you are doing the LM399 in the meter a disservice - my experience from a KS 34465A and a DAQ6510 is that, yes, the meters drift a little initially but the references are definitely pre-burned-in or selected for lowest drift, which is one of the things you do pay for with premium brand meters. I have another meter from a chinese mfg, SDM3065X, that very obviously had the LM399 just soldered in without any selection or pre-aging done; measurement was spot on day 1, not so much after a month (drifted out of 1y spec, even.)

I'm a bit sensitive to word weaseling, as you may have guessed, and I am quite specifically allergic to "the customers wanted this, so we listened to them" inverted argument, when the end result is bad for the customer and more $$$ for the mfg.

[hwj-d]

It was found that most users / customers did not have any use for the full data report so that practice stopped in order to save processing time / cost.

How can you find out that your customers would rather do without this data from your comparatively very good measuring device?

However, some people would still like the full report so it is available as a paid service and ordering option.

That's what im saying. This is annoying if one has to send back his just unpacked device for an extra payment, although such a thing should be part of the basic service and self-image of such a high-quality device. Especially since there is nothing described in the advertisement about it, as far as I can see.

That being said, hwj-d PM me your unit serial, I can probably get the 10V range adjust data you are using for your instrument since you seem to want it. Also, FYI you are measuring an LTZ1000 with a 6500 which has a non burned in LM399. Are you doing a time study or a temperature study?

Thanks for the offer. The device is two years old now, and it works great.The DMM satisfies me otherwise, and the LTZ references can be measured with it.  Especially in long-term tests with several references against each other, and if one knows the characteristics of his DMM.

I know the disadvantages of a LM399 measurement on a LTZ1000 reference. After 2 years nonstop 24/7 use, it should be burned in by now. And as you can see, I also get traceable repeatable data. 

After this period of time it is advisable to have the DMM calibrated anyway, if this will give me access to the missing qualified protocol. Of course, for my needs, the complete resistance measuring ranges are included in addition to all DC measuring ranges.

But I also go the other way first, and have some of my selected references calibrated, now that I know their drift and temperature behaviour 24/7 after also >2 years.

[SaKhan]

It was found that most users / customers did not have any use for the full data report so that practice stopped in order to save processing time / cost. However, some people would still like the full report so it is available as a paid service and ordering option.

If it's not a secret, how much does it cost to get the full report?