Is a mosfet floating source ok?

You say "odd", but that is exactly the point you were confused about that y ou now should understand.

The ohms/volt rating of the multimeter is an expression of the loading of t he meter on the circuit when measuring volts. Or course a higher rating is better meaning it draws less current in voltage mode and is the inverse of the meter resistance. The resistance of the internal meter tells you how much it impacts the circuit when measuring current. A higher resistance re sults in a higher voltage drop. So it would appear the meter inside your m ultimeter is 5 kohms and 50 uA. If you do the math, you will find the 0.

25 V range and the 50 uA current range both simply connect the internal met er directly to the inputs.

What is a bit more difficult to understand in multimeters is how the switch es work to change the ranges. The ones I've looked at were interesting com binations of series and parallel resistors that allowed switching one input to different points in a circular arrangement of resistors and the meter. I can't recall the advantage this gave. It seems to be hard to find info on moving coil multimeters these days. I think it had to do with measuring either voltage or current from the same circuit. I wonder why they had th e coil moving rather than the magnet, but I suppose that's because the magn ets were so large in those days.

--

Rick C. 

+++ Get 1,000 miles of free Supercharging 
+++ Tesla referral code - https://ts.la/richard11209

for the IO with diodes to supply and gnd.

for 5V compatible IO they commonly say 5.5V and +0/-5mA

Peabody wrote:

I think that 1M resistor should work fine with your meter and give you 400 nA range with 0.1 nA resolution. Maybe also 10M for 10 times better resolution.

To give you same idea I tried few different meters. First, cheap DT830B. It has rather accurate 1M input resitance so on 200 mV range gives 200 nA current measurement. As extra bonus no math is needed to convert result: number on display just gives nanoamps. Second, DT9205A. It has

10M input resistance, so theoretically gives 10 times better resolution. Third AN8001. It has autoraging milivolt range, automatically switches between 60mV range and 600mV range. On 60mV range it gives 0.01mV resolution. IIUC on this range there are no explicit resistors, current flowing into/from meter is just leakage. Small, but probably nonlinear. This should be somewhat comprable to your meter. I used to to measure voltage drop on 10M resitor.

Fourth meter is AN8009. Somewhat similar to previous one, but more sensitive. On milivolt range it autoranges between

10mV and 100mV. On 10mV range resolution is 1uV. I also used this meter to measure voltage drop on 10M resitor.

I connected 3 meters in series: DT830B, DT9205A and one of AN meters + resitor. Also in series was 1N4148 diode in reverse direction. Whole chain connected to 9V battery. Agreement between meters was reasonable: DT830B gave only two digits (3.3 in first trial, 3.1 in the second) but they were reasonably stable. DT9205A gave three digits, but the lowest was constantly changing and also the middle digit was changing (but slower than the lowest one). AN gave four digits, firt to reasonably stable, the other to changing. Since readings were constantly changing it is hard to do accurate comparison, but AFAICS they agreed to 5% of value shown. Somewhat disturbing is change of current between two trials. In first trial I used AN8001, in the second AN8009, the rest was the same. I saw change on all 3 meters, so I think that change of current was real (and not artifact of meter inaccuracy). Also, instability of readings is disturbing, but at least in case of AN meters I have reasons to believe that it is real (voltage is changing). Namely there is noise and to fully utilize meter sensitivity one has to limit noise. In fact, after writing the above I did third trial, but this time puting 330nF foil capacitor in parallel to 10M resistor. Now, current is again different, AN8009 slowed down, after few seconds to stabilize result it shows 29.25 with last digit switching betwen 5, 4, and 6 (I take 5 as most likely value). DT830B steadily showed 2.9. DT8205A switched between 28.8, 28.9 and 29.0.

Remark: If you want to use capacitor to reduce noise it is important to take good (preferably foil) capacitor. Bad ceramic may leak fraction of nanoamper and potentially spoil accuracy.

Remark2: Another look at meters showed slightly lower current. So diode leakage seem to decrease with time...

--
                              Waldek Hebisch

Yeah, I guess it depends on what kind of resolution I need. It seems the lower the resistor value the less impact it will have on the circuit, but the less resolutin I would have. As I said in another post, even a 1K resistor would give me a number down to 0.1uA, and if that's all I need to know, that would be the best choice. For the girls I date, fractions of a nanoamp don't come up very often.

But I do need to revisit the original floating gate question, and see how much current is leaking through each stage.

Rick C wrote:

AFAICS there are much more different meters on Aliexpress than you think. There are severel major families, withing family typically there are several models with slightly different functionality. An, likely given model has several manufactures which produce meters which look identically and have identical specs but show significant internal differences (different PCB layout, different parts, sometimes different schematics).

Some meters that I have:

- DT830B, there are bunch of models in DT830 family, but IMO useful one is B, with 200uA range (other take out this range and replace it by something useless to me). One limitation of DT830 is switch which has 20 positions and one position is used to power it off. So number of possible functions is limited to 19. I bought few of them (at different times) locally and one on Aliexpress (from China). I am satisfied with locally bought ones, but the one directly from China failed quickly after arrival. I am not sure if I need to write more about this meter, it is well known with lot of info on the net. Just let me say that when I opened 4 of them bought from different sources each had different PCB...

- DT9205A, this is classic 3.5 digit meter using (probably clone of) ICL chip. It has switch with 32 positions for various functions. There are AC voltage and current ranges using opamp based rectifier. Apparently PCB has places for capacitors to compensate voltage divider on AC range, but those places are empty. On voltage ranges it has 10M resistance. There is also capacitance measurement. Actually I have two different meters which are nominally the same model. They differ internally and functionally: one has 2nF capacitance range, the other 2mA AC current range (again 32 positions on the switch limit number of available functions). Power switch is separate (whic I like).

- Aneng AN8001 and AN8009. Those are autoranging meters with some functions selected using button. AN8001 display goes to 6000, AN8009 display is full 4 digits. Both have 8 positin switch for main function, on AN8001 two postions of the switch are taken by "OFF", on AN8009 there is one OFF position. I bought AN8009 because it is quite sensitive: milivolt range has 1uV resolution (actually it autoranges between two subranges, one 10mV with 1uV resolution, the other 100mV with 10uV resolution). There is nasty feature: on milivolt range anything below 5uV in absolute value is shown as 0 (presumably to hide offset error). Still, AFAICS resolution is real, meter gives resonable result measuring voltage from theromocuple. On milivolt range current to/from meter seem to be leakage of measurement chip. My estimate is that this is few picoampers. On current ranges it has very low voltage drop. I did not measure actual voltage drop, but on 10A range clearly test leads are main contributor (internal shunt is of order 10mOhm). Resolution on A/mA range (using 10A socket) is 0.1mA (again it autoragnes between full 10A range and 1A range). I have noticed that my meter has -0.4 mA offset error ont this range (meter is now 1 year old, I do not remember if it always had such offset or if this degraded with time). Still, small current seem to produce expected change, so resolution seem OK. There is uA range with 10nA resolution (autoranges between two subranges, 1mA and 100uA), but is seems to have offset error, with no current shows -50 (- and 5 on lowest digit), with 136nA shows 60 (positive sign and 6 on lowest digit). There is gap on miliamper range: 1mA of more overlad uA range, but have poor accuracy on A/mA range. On AC ranges this meter nominaly is true RMS, but this is via 10000 samples per second ADC and digital processing, so limited to about 1kHz. And due to this AC bandtwith is rather small. Beside voltage and current there is frequency counter, duty cucle, resistance (up to 100MOhm) and capacitance measurement. There is also temperature measurement.

AN8001 has less resolution, it mV range has 10uV resolution (autoranges between 600mV and 60mV subranges), on 10A range has 1mA resolution (autoranges between 10A range (should be 60A but presumably limited by meter heating/durablity) and 6A range). Miliaper range autoranges between 600mA and 60mA subranges, so 100uA resolution on lowest subrange. Resitance goes to 60MOhm.

As I wrote, I bought AN8009 due to its resolution, in price range up to $200 I was unable to find anything with better resolution. Low resolution in the middle of current range probably means that it can not serve as only meter. AN8001 is probably better for "normal use". To say the truth, I do most of my measurements using DT830B, using other meters when DT830B can not do: AC currents or voltages (DT830B has AC voltage ranges but is quite inaccurate for low AC voltages), capacitance or when I need more accuracy.

I looked at several other meters on Aliexpress and no of more expensive meters looked better to me than Anengs: many had lower resolution than AN8001 and each has some limitations in available functions.

Concerning quality, since apparently there are several independent manufactures which make identically loking meters, quality will vary...

--
                              Waldek Hebisch

king at 50 vendors on eBay or Aliexpress selling the same crap meter. Did I say 50? More like 500! It's hard to find useful variation or get any idea of quality.

Aliexpress is the worst for searching. If you enter two words they return everything that matches EITHER word. They return as much crap as possible or that matches in any way possible. Ebay may not be great, but I can usua lly manage the searches ok.

I have one that is like the AN8001, but it's designated RICHMETERS 102.

6000 count, but just one off position. The other off position is for a temperature probe. Looks like this one is the AN8002. The AN8009 picks up non contact voltage and a square wave output. Otherwise I'm not sure what else it offers over the 8002.

Thanks for all the info. These AN meters are nice, but I'd like to have on e that is a full size just so it sits on the bench better. The one larger device I had (before the Costco batteries leaked) was smaller at the bottom than the top and tended to tip over, so it had to lay flat. At least the numbers were large and easy to read. That was the one with nearly invisibl e writing on the controls. Some values used a light blue ink that faded in to the grey case.

I might get an 8009 just for the fun of it. At $26 it's hard to go wrong.

--

Rick C. 

---- Get 1,000 miles of free Supercharging 
---- Tesla referral code - https://ts.la/richard11209

First off all

1/ from where 3V for the DS chip comes from ?

VCC=12V ; So when

2/ SQW = 3V --> Vgs = -3.1V --> Vgs is likely near Vgsth ... so the P-MOS fet could be "on-state", this circuit is not working as intended.

3/ SQW = 0 --> Vgs = -12V --> Vds=0 ; Ok

Good Lord ! How this design is intended to work ?

H