New Keithley DMM6500 and now DAQ6510

[HendriXML]

I rewrote the script, which now only uses one measurement (shunt voltage R101) and thus is useless.
But look now the measurments are just fine (pretty precise tuning :
The rows indicate:
target current: total averaging time (minimum of 30 s), running average measured shunt voltage, running standard deviation measured shunt voltage

Code: [Select]

10 mA: 39.10, 5.07171, 0.000209
8,2 mA: 30.85, 4.15821, 0.000126
6,8 mA: 32.98, 3.44836, 0.000171
5,6 mA: 31.38, 2.83963, 0.0000477
4,7 mA: 31.28, 2.38344, 0.0000456
3,9 mA: 31.28, 1.97755, 0.0000568
3,3 mA: 31.13, 1.67344, 0.0000620
2,7 mA: 31.22, 1.36911, 0.000119
2,2 mA: 30.89, 1.11570, 0.0000319
1,8 mA: 31.12, 0.912941, 0.0000985
1,5 mA: 30.85, 0.760737, 0.000107
1,2 mA: 30.91, 0.608708, 0.0000843
1 mA: 30.89, 0.507387, 0.0000672
820 μA: 30.93, 0.415905, 0.0000390
680 μA: 31.66, 0.344852, 0.0000158
560 μA: 30.87, 0.284076, 0.0000963
470 μA: 30.89, 0.238457, 0.0000269
390 μA: 30.85, 0.197997, 0.0000311
330 μA: 30.93, 0.167316, 0.0000212
270 μA: 30.89, 0.136969, 0.0000465
220 μA: 31.37, 0.111924, 0.0000481
180 μA: 31.00, 0.0913745, 0.0000170
150 μA: 30.82, 0.0761302, 0.0000215
120 μA: 31.04, 0.0607542, 0.0000356
100 μA: 30.86, 0.0507703, 0.0000393
So it really seems that the DC ratio function has some strange quirks, under "certain conditions". Will run the simple script with different wires connected to rule out loose wires.
Edit: different wires run just fine as well.

[HendriXML]

The more straight forward way would be using a scanner card that is available for the 6500.

The scanner card I find too expensive,  as well as the digital io card, which both would be very nice to use for experiments though.

In another experiment I used a scope by tuning  the offset voltage to do precise measurements (0.1 mV accuracy @ around 1.5V), but that needs also a whole process of calibrating the offset.

So too keep it simple I made my 2 channel scanner wire  with just a relay and diodes and will control that using the 2nd channel of the awg. Switching will be done using DC offset, with or without polarity inverse. With my awg that switching is without relays. (Going from 0V dc to 10V dc has some range relay switching.)

Too bad the ratio switching seems to have some issues, but this way I'll get my data as well.

[Hydron]

A few people have made their own scanner cards compatible with the DMM6500, check out these:

https://github.com/macgeorge/SCAN2000STM32
https://github.com/cozdas/CozScan2020
https://www.eevblog.com/forum/circuit-studio/example-project-relay-scan-card-for-k2000-dmm/

The only reason a MCU is involved in 2 of the above designs is it avoids pissing about with latching relays - would otherwise be able to be done with a few logic ICs and the appropriate connector (see the last example).

[voltsandjolts]

...and another 20 channel SSR card:
https://www.eevblog.com/forum/circuit-studio/example-project-20-channel-solid-state-scan-card-for-k2000-dmm/

[HendriXML]

I ran an updated script. One fase (tuning) of a measurement cycle measures only current voltages, the last fase measures also Vbe voltages. This way there's less channel switching.
I would have like to use only one trigger model, which I would the steer with my own script. This seems almost impossible. There is a way to sync with the dmm: let it wait for a *TRG command. But that cannot pass info. And there's no way to way to sqeeze in info some other way, except for lan events. These can have 8 different id's, so they could be used for steering.
Did some research and with some more effort lan events (broadcasting a udp packet) could have worked.
However I went the route of swapping between different trigger models.
The funny thing is that without the ratio measurement,  I still get the jumps some times.
The solution is to have large delays between measurements. I think I went for 1 s, but that might not even be enough. Will experiment further, because it seems to depend on what is measured and at what range. This makes me also wonder of secondary measurements, in what way is are they different to what my trigger model is doing? Why would it result in either a jump, or no jump, but not much in between. (Except sometimes there seem to be 3 levels)
Fascinating..

[HendriXML]

Yesterday I had a successfull run of all target currents using 2x 1 s delay , which means good stability over 30 s at every target current.

Today with 2 x 2 s delay, I got this attached result. 

I'm pretty sure it has to do with doing 2 measurements, but what might be the underlying cause?

A warning says that using 2 reading buffers in a triggermodel might impact performance (says the eventlog), but I don't think it means this kind of performance, I hope..

[HendriXML]

I've been babysitting the measurements...

I've 2 series with a delay of 2 x 5 s.

Before the combined measurements they both reach a stable target voltage (is mentioned in filename) for at least 10 reads.

Then the combined measurements start. In one situation the jump up is the exception, in the other the target voltage becomes the exception. (Resulting in a "stable measurement" by the criteria I've set)

Hope some one can point out that I'm doing it all wrong.. When using a scanner card similar switching happens doesn't it?

[HendriXML]

I came to the marvelous idea to let my handheld meter read the voltages as well. And it seems that it's reading the same jumps. So they might actually be real.

In that case a measurement of Vbe, might disturb the circuit in such way that the constant current takes a jump. Another possibility could be that the awg, starts misbehaving. But nothing special can be seen at the Vbe measurements.

I actually prefer the circuit being at falt. It would hopefully mean that adding a buffer would make things ok.

Maybe capacitative coupling bewteen the powerlines plays a role?

[MegaVolt]

Maybe capacitative coupling bewteen the powerlines plays a role?

What does the oscilloscope show connected to this line?

[HendriXML]

I made a collection of screenshots. Vertically which measurement wires where connected differ. Horizontally which measurement was switched to.

Channel 1 just measures the awg voltage, which feeds the base of the transitor via a 560 ohm resistor. Channel 2 shows the voltage at the relay.

Switching my 2 channels causes a lot of disturbance at a point which has a buffer cap, and only 50 ohm output resistance on which the difference is measured (total: 560+50). So at the base the disturbance is larger..

This was not even a measurement which had jumps from what I know. This to me already shows that the measurement voltages need to be buffered. The strange thing is however that current (and not vbe) measurement wires (connected vs disconnected) seems affect the disturbances the most.

[Kleinstein]

The transitor circuit is oscillating. The cables add capacitance, and this alone may make enough difference to go from just stable to oscillating or strong ringing.

The circuit looks a little difficult to make stable. A first try would be a capacitor (e.g. some 1 - 100 nF)  of Q102 collector to ground. I am not sure this is enough - it is not one of the more usualy circuits. If not enough add some resistance before the base of Q102 and than the capacitor from collector to base.

[HendriXML]

The capacitor Kleinstein mentioned made the circuit stable.
The circuit was taken from this page https://electronics.stackexchange.com/questions/360175/leftover-voltage-when-using-constant-current.
It has some lengthy explanation on its workings. Putting it in my experimental setup with all the wires and awg to control the current probably made the circuit unstable (but the cap stays ).

Together with the characterized transistors, used in the validated current mirror, I'll be using it to have a lineair ramp circuit which can be speed up. (Will check that for stability!)

I added the results of the measurments. To me they look solid and give some idea what the accuracy will be and what base resistor to use.

When using the ratio functionality, more samples were taken and averaged, but that is a big pro of not having to wait the relay switching. The room temperature was quite different, so that probably explains the shift upward (B: colder). Would be a nice project to have the temperature measured (remotely) and added to the experiment report. Here's a page I found about creating selfmade scpi devices.
https://electronicprojectsforfun.wordpress.com/homebrew-scpi-controllable-instruments-with-arduino-controllers/
Besides the temperature measurement, switching relays via lan for example would have a ton of uses. To bad the digital io card is expensive, because those 6 lines could also be controlled from within the dmm. 

[Hydron]

I think (but am not sure) that you can close/open arbitrary relays on a scanner card if they don't have a measurement function assigned. Would mean that a DIY scanned card can act as a LAN controller for some relays if you wanted (wired to not connect to the measurement terminals). Can probably be tested using the virtual scanner you can activate for script testing.

[HendriXML]

I think (but am not sure) that you can close/open arbitrary relays on a scanner card if they don't have a measurement function assigned. Would mean that a DIY scanned card can act as a LAN controller for some relays if you wanted (wired to not connect to the measurement terminals). Can probably be tested using the virtual scanner you can activate for script testing.

According to the manual I think you're right. That's something to keep in mind. 

[voltsandjolts]

I think (but am not sure) that you can close/open arbitrary relays on a scanner card if they don't have a measurement function assigned. Would mean that a DIY scanned card can act as a LAN controller for some relays if you wanted (wired to not connect to the measurement terminals). Can probably be tested using the virtual scanner you can activate for script testing.

Eh, no.
The relays connect input terminals to the measurement bus.
Activating multiple relays can create a dangerous situation where voltage on one input is routed out of another.

[Hydron]

I think (but am not sure) that you can close/open arbitrary relays on a scanner card if they don't have a measurement function assigned. Would mean that a DIY scanned card can act as a LAN controller for some relays if you wanted (wired to not connect to the measurement terminals). Can probably be tested using the virtual scanner you can activate for script testing.

Eh, no.
The relays connect input terminals to the measurement bus.
Activating multiple relays can create a dangerous situation where voltage on one input is routed out of another.

Key passages in my post were "DIY scanned card" (*scanner) and "(wired to not connect to the measurement terminals)" - the Keithley one will have the issue you describe, but a DIY one can be designed to have a few isolated relays with no connection to the measurement bus. You get up to 20 channels, so you can have a fair bit of flexibility - I'm planning on a reference temp sensor channel plus a few thermo-couple, relay, SSR, current sense and isolated channels when I get around to building mine.

[voltsandjolts]

As I said, activating multiple relays can create a dangerous situation where voltage on one input is routed out of another.
You seem to think the dmm firmware will allow this to happen. Be prepared for a dissappointment.

Edit:
Hmm, OK, I concede. Having read the manual it does sound like it allows multiple channels to be closed. I find that surprising.
The firmware on my 20 channel card wouldn't allow that, it opens all relays before closing any, with safety in mind.

[Hydron]

You seem to think the dmm firmware will allow this to happen. Be prepared for a dissappointment.

I said "I am not sure" as this is certainly something I'd want to at least try using the virtual card option before concluding that is is possible, regardless of the fact that the manual says that it will work. Did you try with your scanner card on a DMM6500? Or an older card compatible instrument?

I would say that there some sense in allowing arbitrary channel closure (though as you say, care is required) - you could have some selectable currents shunts or something (filter capacitors?) available that you want to be able to switch across the measurement bus, maybe this is why it's even an option in the manual (and presumably DMM).

In theory I should have a DMM6500 in my hands in about 2 weeks time - hopefully will be able to do more than just infer things based on the manual then!

[maxspb69]

I accidentally discovered that my new DMM6500 has an 11A Bussmann fuse inside, and not 3.5A like everyone else. Absolutely new meter, production date: January 2021. Is this normal or a manufacturing error and should be replaced urgently with 3.5A?

There is also a wire. connecting the K7 optorelay to U32 chip under shield. The wire is installed very neatly (at the factory?) By the way, the chip (U32) to which the wire goes is not installed at all on earlier boards (according to the photo from the forum). Do I have some kind of new revision with improvements?

The device was produced in January 2021, purchased from an authorized dealer as completely new. I was the first to open the box (the box was sealed with Keithley's branded  tape).
Not refurbished.

[voltsandjolts]

The 3A fuse is inside the banana jack on the front & rear panel.
The 10A banana jack is on the rear panel - you already found the fuse for that.

[E-Design]

I accidentally discovered that my new DMM6500 has an 11A Bussmann fuse inside, and not 3.5A like everyone else. Absolutely new meter, production date: January 2021. Is this normal or a manufacturing error and should be replaced urgently with 3.5A?

There is also a wire. connecting the K7 optorelay to U32 chip under shield. The wire is installed very neatly (at the factory?) By the way, the chip (U32) to which the wire goes is not installed at all on earlier boards (according to the photo from the forum). Do I have some kind of new revision with improvements?

The device was produced in January 2021, purchased from an authorized dealer as completely new. I was the first to open the box (the box was sealed with Keithley's branded  tape).
Not refurbished.

You do not have a refurbished instrument.
These are both perfectly normal and are product improvements. However, it is not a new hardware revision (which we are working on).

Your internal 11A fuse is a safety item only and is present in case of some accidental high voltage connection to amps or other catastrophic internal event. The panel 3A fuses are for your convenience if you accidentally blow it during normal use, you can easily access it and the fuses are low cost. You shouldn't ever need (hopefully) to get to the internal 11A.

Hope this helps clear up your concerns.

[Mr. Scram]

You do not have a refurbished instrument.
These are both perfectly normal and are product improvements. However, it is not a new hardware revision (which we are working on).

Your internal 11A fuse is a safety item only and is present in case of some accidental high voltage connection to amps or other catastrophic internal event. The panel 3A fuses are for your convenience if you accidentally blow it during normal use, you can easily access it and the fuses are low cost. You shouldn't ever need (hopefully) to get to the internal 11A.

Hope this helps clear up your concerns.

What does this hardware revision entail? And does this mean recertifying the model? I've always wondered about that.

[maxspb69]

The 3A fuse is inside the banana jack on the front & rear panel.
The 10A banana jack is on the rear panel - you already found the fuse for that.

The 11A fuse for "10A" terminal is located in the rear terminal module. Inside there is a second fuse for the 3A terminals, common for the front and rear inputs. Most of the DMM6500 have a nominal value of 3.5A. In my meter, the internal (for the 3A measuring circuit) is 11A.
The purpose of the fuse in the current measuring circuit is to save the shunt from burning out when the current is significantly exceeded. An 11A fuse will not save a 3A shunt, it is more correct to have a 3.5A fuse there. (I know about the 3А glass fuses inside the current terminals).

[Hydron]

The shunt _may_ have a voltage clamp across it, plus it has the 3A fuse as a first line of defence - I assume something has to go badly wrong to put the shunt at risk (e.g. you've shorted out a source higher than the meter's rated CAT II 300V).

I suspect the change to 11A is to improve the burden voltage (the 11A fuse will have a significantly lower resistance than a 3.5A one), or to reduce the chance of it blowing before the external 3A fuse. I personally would prefer the 11A in that location if it is judged suitable as it seems it has been (unless the hardware has been updated to enable the higher rating to be used).

I would also like to know the details of the hardware changes/improvements - I have a DMM6500 scheduled to arrive next week but it is an ENCORE (refurb) model, so manufacture date may not be so recent.

Edit: yep, looks like there is a chunky bridge rectifier literally glued to what looks to be the 4-terminal high current range shunt. I would assume this is to limit the voltage across the selected shunt in a fault condition to ~1.4 volts (two diode drops), and should easily hold up until the fuse clears with any reasonable overload.

[Hydron]

I accidentally discovered that my new DMM6500 has an 11A Bussmann fuse inside, and not 3.5A like everyone else. Absolutely new meter, production date: January 2021. Is this normal or a manufacturing error and should be replaced urgently with 3.5A?

There is also a wire. connecting the K7 optorelay to U32 chip under shield. The wire is installed very neatly (at the factory?) By the way, the chip (U32) to which the wire goes is not installed at all on earlier boards (according to the photo from the forum). Do I have some kind of new revision with improvements?

If you get the chance, could you check what U32 is? (I assume you've put it back together though, I am not asking you to open everything again!). Also would you be able to check the burden voltage of the 3A range at say 1A (source 1A from a PSU, measure voltage across the input terminals of the DMM). Would be interesting to compare with a 3.5A fused instrument.