Products > Test Equipment

ET5410 Electronic Load mod: add external sense

<< < (19/21) > >>

DaneLaw:

--- Quote from: klausES on August 07, 2024, 02:54:41 pm ---There are very different information, even up to 4.3 kg in one case a dealer writes 5 kg.
All of the written does not close to the scale with your weight. So I don't go from most information on the Internet.
This seems to me sucked out of their fingers. Really measured, hardly any of these retailers who otherwise also sell women's stockings.

--- End quote ---
As highlighted.
There is a big difference in weight between the old versus new [A+model] its just how it is, it ain't down to site-wording limbo.
Looking at the internals.. it's been cost-optimised/enhanced.

That doesn't always have to be a bad thing (even on a programmable sinkload) if the prices also drop noticeably.


--- Quote from: klausES on August 07, 2024, 02:54:41 pm ---Incidentally, a nice button for the Rotary Encoder. Hopefully your device doesn't weigh so much because of this  ;)

--- End quote ---

No, its a beefy aluone, but not that weight potent... ;D

MF-jockey:
My ET5410A+ arrived on September 2nd, 2024.
The device without cable weighs 2923 g.
First I checked the accuracy of the voltage measurement.
When the inputs are open or short-circuited, 0.012 to 0.019 V are displayed. It changes slightly when switched off and on again.
With a connected battery, an exactly the same voltage of 13.755 V is displayed compared to the OWON XDM1241 (accuracy 0.05%). A 10V reference voltage is measured once at 10.006 V and a little later at exactly 10,000 V.
This was followed by a check of the current measurement in the range up to 11 A, the XDM1241 cannot do more than that. It was noticeable that the load does not always regulate the setpoint exactly. The largest deviation of 0.23% between the two displays was at 7 A. However, the OWON is only specified with an accuracy of 0.5% in this range.
The next test was the contact resistance up to the circuit board. To do this, I connected the terminals with a solid copper wire and measured the resistance between the circuit board points with the DE-5000 from DER EE; it was 3 mOhm.

Now for the modification with sense inputs. I don't like the solution with a switch because I expect contact resistance in the long term. You also have to pay attention to the switch position.
I prefer the solution with two low-resistance resistors and have calculated their measurement error. As a result, I decided on 10 Ohm resistors.

This results in a small additional error of only -0.005% for 2-wire measurements (e.g. -0.6 mV at 12 V or -2.4 mV at 48 V). This is negligible compared to the error that typically occurs due to the resistance of the wiring.
When measuring with 4 wires, you should also pay attention to low-resistance measuring cables, otherwise the voltage drops in the load cables will affect the measured value via the 10 ohm resistors. If you have to deal with longer cable lengths, you should choose the 22 ohm variant, as this will result in an additional error of -0.011% in 2-wire operation.
I am attaching the calculation with both variants here and hope that it is presented in an understandable way. Please see the attached "error-calculation.pdf".
In a 30 A load test with very good load lines (2x 10 AWG parallel, approx. 10.5 mm² Cu, only 0.27 m long), the additional sense lines produced voltage values ​​about 50 mV higher.

MF-jockey:
Better is the enemy of good.
Afterwards I noticed that you should connect the resistors closer to the source.

This avoids the voltage drops on the internal power rails (each has about 1 mOhm).
For 2-wire measurements, the voltage measurement is then better than in the original state (even with 22 ohm resistors).
For 4-wire measurements, the measurement error is also reduced somewhat.
At the moment I only use it about halfway.
When I open the device again, I will correct this.

mawyatt:
Have you considered what will happen when just the High Current cable/wire from the Source to Load is interrupted? This should cause the Load/Source to seek or force current thru the Sense Lead and thus the 10 ohm Resistor, which will result in High Currents attempting to flow thru the Resistor which may Open due to I^2*R heating and the Load will loose feedback!!

Maybe consider a "Safer" mod would include a High Current SBD in parallel with the Resistor and with a Higher Wattage Resistor. Also adding a Low Current PTC Fuse in Series with SBD and Parallel Resistor to not allow the High Current to be sustained thru the Sense Leads and potentially damaging them.

Doing some "Safety Tests" with just Power Supply Controlled Sources while disconnecting various leads from the PS to Load should reveal how the Load misbehaves under these conditions, and one doesn't "find out" with an expensive Source  :o


Best,

MF-jockey:
Yes, that is a sore point of this solution, that the load line must be carefully connected, especially at higher voltages. In the event of a fault at higher voltages, the affected 0.3 W resistor acts like a fuse and must then be replaced.
You should also connect a polyswitch fuse (e.g. RXEF010 from Littelfuse, is safe up to 60V, safe up to 240V is BOURNS MF-RM012/240) in series with the resistor and at the same time use a resistor with a higher power (>= 3 W). Due to the resistance of the polyswitch (2.5 to 7.5 ohms), the measurement error in 2-wire operation increases slightly.
Please do not place the polyfuse or PTC in the measuring line, this will significantly impair the voltage measurement.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod