Fluke 87V Frequency counter upper limit 1.2MHz +

[Dread]

I was fooling around with a 87V I got off of EBay used (Yeah thanks Dave that's more money I am spending because of EEVBlog  ) so I decided to check out everything to make sure it was working and found a (Blown mA fuse of course  ).

I got to the freq counter and fed in a 50Hz 1Vpp signal and started cranking the Generator up.
At 200KHz when I expected a sudden stop it kept going so I got to 400KHZ and said hold on let me check the voltage reading.  The voltage had dropped a little maybe 5% so I kept going.  At 600+KHz and the meter started to give erratic readings so I figured let me check the voltage again, this time it had dropped by a lot so I cranked up to 2Vpp and the reading was solid again.  This went on until I got to 1.250 MHz on the fluke with 7Vpp on the generator and at that point I decided that I was not going any further.    BTW the reading was spot on with my frequency counter all the way!

Interesting I never knew the 87V could read that high a frequency beyond it's specs.

BTW My board is Revision 10 and came with a perfect outer shell but a scratched as all hell display.

Thank the lord for Novus, it removed all the scratches and made the display look Brand new. Seriously I doubt anybody could tell if the meter was used or brand new after a 20 minute Novus session.

[korborh]

I am able to read 1.6MHz on my 87V ...the voltage has to be high ~20Vpk though.

[retiredcaps]

I was fooling around with a 87V I got off of EBay used (Yeah thanks Dave that's more money I am spending because of EEVBlog  ) so I decided to check out everything to make sure it was working and found a (Blown mA fuse of course  ).

Since my sellers seem happy to *cough* gouge *cough* on shipping prices to Canada for a Fluke fuse, I now ask Fluke multimeter sellers if the fuses are working.  If they don't know, I send them instructions on how to test it.

I'm sick of sellers listing everything as "working perfectly" and then having one or more blown fuses.  I have been burned 3 times with those listings.

[Dread]

I am able to read 1.6MHz on my 87V ...the voltage has to be high ~20Vpk though.

Was that the absolute highest reading or did you back off like me and decided to stop at 20v?  I am amazed at how accurate the counter is but I have no idea if some damage may occur at too high a voltage / freq.

[korborh]

I am able to read 1.6MHz on my 87V ...the voltage has to be high ~20Vpk though.

Was that the absolute highest reading or did you back off like me and decided to stop at 20v?  I am amazed at how accurate the counter is but I have no idea if some damage may occur at too high a voltage / freq.

After somewhere above 1.6MHz it started to give random values. The function generator was set on its max output which is 20V for open circuit .... but it may not be actually that high as it would be loaded by the 87V ac coupling. The meter was reading accurately well into the 1.6MHz freq though.  The AC input can take 1000V, so I was not worried about voltage.

[Wartex]

I was fooling around with a 87V I got off of EBay used (Yeah thanks Dave that's more money I am spending because of EEVBlog  ) so I decided to check out everything to make sure it was working and found a (Blown mA fuse of course  ).

Since my sellers seem happy to *cough* gouge *cough* on shipping prices to Canada for a Fluke fuse, I now ask Fluke multimeter sellers if the fuses are working.  If they don't know, I send them instructions on how to test it.

I'm sick of sellers listing everything as "working perfectly" and then having one or more blown fuses.  I have been burned 3 times with those listings.

I bought 87-5 with a blown fuse for $100 because the seller thought they damaged the meter and "amps didn't work". Then I sold it for $280 locally with a proper fluke $8 fuse and bought unopened Fluke 289.

[ilikepez]

Yeah the highest range is over 200 khz, but not necessarily the highest measurement.

[Dread]

Yeah the highest range is over 200 khz, but not necessarily the highest measurement.

The accuracy of the Vrms is rated to 200khz, after that the accuracy starts to degrade.  In most cases that matters but in some cases all you may need is the freq measurement, so it's nice to know that I can read up to a 1MHz signal if the Vpp is at least 5V.

[firewalker]

My Fluke 87 III with sine, 1 Vpp goes up to 210 kHz. With 10 Vpp goes up to 800 kHz. With the ttl out connected it will reach 605 kHz.

Alexander.

[alm]

The function generator was set on its max output which is 20V for open circuit .... but it may not be actually that high as it would be loaded by the 87V ac coupling. The meter was reading accurately well into the 1.6MHz freq though.  The AC input can take 1000V, so I was not worried about voltage.

Maximum voltage is derated with frequency, since the meter presents a mostly capacitive load at high frequencies. According to the specs, the maximum voltage - frequency product is 10⁶ V Hz, or 0.6 V at 1.6 MHz.

[korborh]

The function generator was set on its max output which is 20V for open circuit .... but it may not be actually that high as it would be loaded by the 87V ac coupling. The meter was reading accurately well into the 1.6MHz freq though.  The AC input can take 1000V, so I was not worried about voltage.

Maximum voltage is derated with frequency, since the meter presents a mostly capacitive load at high frequencies. According to the specs, the maximum voltage - frequency product is 10⁶ V Hz, or 0.6 V at 1.6 MHz.

I see that spec in the manual now. ...perhaps it is not meant to be for all frequencies. For higher frequencies, it would be relaxed otherwise 0.6mV at 1.6GHz would overload the meter! I guess the meter attenuates internally due to bandwidth limitation hence the need for having higher voltage for accurate reading.
Anyway, had I known this spec before, I would not have tried the experiment. The meter survived no problem....

[Dread]

The function generator was set on its max output which is 20V for open circuit .... but it may not be actually that high as it would be loaded by the 87V ac coupling. The meter was reading accurately well into the 1.6MHz freq though.  The AC input can take 1000V, so I was not worried about voltage.

Maximum voltage is derated with frequency, since the meter presents a mostly capacitive load at high frequencies. According to the specs, the maximum voltage - frequency product is 10⁶ V Hz, or 0.6 V at 1.6 MHz.

I figured that but was not sure of the numbers.  That why I backed off at 7Vpp which was still way over the limit.  I am always worried when working with RF and my test equipment.

[alm]

Obviously inductance and series resistance will become important at higher frequencies, especially if you use standard test leads. What would be the point of specifying it less conservatively beyond its rated bandwidth, however? It's not like this would allow you to measure larger 1 GHz signals.

The main point is that it's not rated for 1000 Vrms at 1 MHz, not even close.

[korborh]

Looking at the ACV input circuitry, it seems like there is pretty good attenuation at high frequencies...the 5.6pF cap in series with 100K.
So I am wondering why this conservative limit on 1000V-Hz? 
The schematic here is from 87 (not the 87V) but I reckon it would not be very different.

[alm]

There is about 400 pF in parallel with the 101 kohm resistor used for the 1/100 voltage divider. Presumably this means that there is about 4 pF in parallel with the 9.996 Mohm, since the voltage divider should be compensated. There may also be additional parasitic capacitance not shown in the schematic, for example in the FET switches inside the ASIC.

At the rated 400 Vrms for the 1/100 divider, the current at 1 Mhz would be about 10 mA, probably higher than the internal switches can handle. With 10 Mohm of resistance in the divider, the current at 400 VDC would be 40 uA. To limit the current to the same current at 1 MHz, the voltage at 1 MHz would need to be limited to 1.6 Vrms. This is quite close to the 1 V specified limit. Note that this was just for a single resistor divider at a single range, and under the assumption that the current carrying capacity of the analog switches is the limiting factor.

[korborh]

alm, thanks for the explanation - yes that makes sense, that the 5-10ma current could be too high for the internal switches.

[Dread]

Maybe a quick Dave video could confirm the current limitations of the chip.