DC dummy load circuit calibration

[VEGETA]

After seeing that I obviously won't go with this circuit as it is... but it is good that we came this far.

I put the current sink as you mentioned and yes it does ruin the supply for IDK reason.

In a panic, this is what I came with as a solution:

Suggestion 1: Using 7660.

I found it on JLCPCB supplier (https://lcsc.com/product-detail/PMIC-AC-DC-Converters_HTC_TJ7660D_TJ7660D_C126092.html) so they can deliver it with the PCB itself.

We could hook the negative of 12v supply to ground which in turns will eliminate all these peaks of noise as I tested. However, we use a simple resistor divider to get approx. 5v or so then feed it to 7660 to get -5v which becomes our negative rail.

I think this is very simple if it does not introduce anything else. I knew about this all along but I didn't want to use 7660 since it is not so available to others (even for me now xD).

We can simulate the -5v supply easily by making a 5v supply with its positive to ground and its negative is -5v. As for the LED I removed it and installed the 100 Ohm back... Now if i want it I can connect it anywhere else like on the main 12v rail with proper resistor dividers and so on.

If we want an extra cost we would use 5v linear regulator 7805 to give 5v and power the LED while at the same time give 5v to 7660.

I know 7660 charge pump has switching but I don't think it will affect us here, or will it?

How about this?

Suggestion 2: use another LM317

or better yet 7805 hooked to 12v and the negative of 5v regulator is hooked to ground not to negative rail. Here we can avoid the negative rail all together.


Suggestion 3: op-amp rail splitter

this is not even a solution to this design but rather a completely new way to avoid the dodgy negative rail of ours, perhaps it is the cause of all this. I didn't build this yet.


___

I saw from other designs that they use suggestion 1 since it gives them 12v full voltage while having also -5. Getting that extra IC maybe worth it if the design is working and all these problems are solved. Scullcom design uses suggestion 1.

If you go with this design without resolving the issue with the Vbe multiplier and the meter supply current, I'm washing my hands of this whole project and ignoring this topic.

Don't give up on meeeee! I am trying also xD. Originally I wanted this to finish in a week or so but looks like it took a lot more. I saw Dave's project so simple, also the one from mjlorton so I thought I could do it similarly.

I am good at electronics but not too good, that is why you see me asking a lot in details. I work as an instrumentation planning engineer right now so it is a different field.

Looking forward to read you opinion about the suggestions above. I hope we can finish it soon.


____

EDIT:

Suggestion 4: put the panel meter across 12v.

Meaning from positive (+9v) and negative rail (-3). This way it won't interfere in our dodgy negative rail and ground config.

However, I remembered that it must share a common ground with the measuring signal so this suggestion is bad... forget it.


___

EDIT 2:

I tried different resistors and caps with 2n2222 circuit and yet it refuses to work! looks like suggestion 1 is the way to go. getting 7660 is doable and if we are to do so then we would get tl431 to replace lm317 if we want.

I need to read your opinion about suggestion 1, if it works or not.

I really want to finish the project xD. Don't you?

[VEGETA]

I have tested the opamp splitter and it did not work! same problem as before.

I used 30k with 10k to get 9v and -3v but the same with other values.

Here, there are no transistors at all so what is the cause of this error?!

[Ian.M]

If you use a 7660 charge pump for the negative rail, it needs a good stiff power supply to work well, so don't even think about a resistor divider to power it.  You could go back to using a USB charger for 5V supply, but that limits the maximum positive swing of the OPAMP driving the gate and will limit the max current you can get with a decent heatsink.   Using a regulator is a far better choice, but I recommend a 9V one as 7660 chips are rated for operation at up to 10V, and 3V headroom is adequate for all common regulators.   Then you can use the regulator output for the OPAMP's Vcc supply which will keep all the noise from the LED meter's changing current draw out of the sensitive part of the circuit.    That also means you don't need the TL431.  You only need a really common 7809 9V regulator - preferably the low current 78L09 variant - or a LM317 + two resistors to set it to 9V.

Run the power LED with a series resistor from the regulator output - 1K would be good which will give about 7mA through the LED - plenty bright enough. 

For 2A max load, you need 0.11V at the top end of the pot, lets say 0.12V to have a little margin to be certain you can reach 2A.   Therefore it needs 12uA flowing through the pot.  This is a very low current, so the PCB must be cloean and if you mount the pot off-board, please use screened cable to hook it up with the screen connected to ground and the pot body if it is metal, but *NOT* to either end of the track or the wiper.     If you feed the pot via a dropper resistor from the +9V regulator output, you'll need 740K in series.   680K and a 100K preset will give you enough adjustment range. 

For the zero adjust at the bottom end of the pot,  as the 7660 output isn't very well regulated, I'd suggest using a 10K resistor and a diode to get a reasonably stable 0.6V-0.7V negative voltage, then using a 10K preset feeding a 10R resistor to divide that by up to 1000:1 to provide the zero adjust.

An OPAMP + BJT based rail splitter would be another possible option - if its designed to sink enough current it could cope with the varying supply current of the LED meter.  If you want to go down that road, I have some thoughts on the rail splitter design I need to develop in LTspice.

What OPAMP splitter are you experimenting with?  Schematic please with component values, measured voltages and exact description of the fault.

[VEGETA]

opamp splitter version below, it does not work.

____

How about sticking to LM317 with a fixed resistor + 10k pot for adjustment -> put it to 9v. Since 9v regulators are not as available..? Then feed that 9v to 7660 followed by 100uF caps for filtering... and also powering the meter from 9v lm317 connected to circuit ground as well as the power LED (1k series resistor as you suggested). This eliminates all problems on the expense of getting lm317 IC + 7660 IC + one extra 10k pot! 

^
until we could get a better solution if opamp splitter worked.


__

I will mount the 10-turn pot on the front panel then connect it by wires to the PCB, thus getting a screened cable is an extra annoying step. Is this due to EMI? You remind me of our instrumentation wiring that requires this type of cable but here I want it to be very simple. If this is gonna introduce some error then it is easily offset by the zeroing pot.

___

getting the 2A will change as you explained since now it is 9v not 1.25 volts... So 740k is a must now and getting 100k pot is also an extra pain but we have to tolerate it.

____

For the zero adjust at the bottom end of the pot,  as the 7660 output isn't very well regulated, I'd suggest using a 10K resistor and a diode to get a reasonably stable 0.6V-0.7V negative voltage, then using a 10K preset feeding a 10R resistor to divide that by up to 1000:1 to provide the zero adjust.

I didn't understand. Do you mean ditch the 7660 completely and rely on diodes? like putting 2 in series to get -1.4? but here it will connect to 12v negative which in turns is ground so it won't work unless we put a divider like before which got us in this mess in the first place.

[VEGETA]

I've been trying these ideas and I came up with this in attachments.

I hope this is the final or near final version, since I don't see any of the previous problems so far.

I've been able to get 1mA out of it which is fantastic, but you gotta know that current meter resolution is 10mA so we have more accuracy than our meter. It is completely safe!


Looking for your response.

[Ian.M]

A LM317 is quite predictable and you don't need a 10K pot to set it as the exact value of the 9V output voltage is non-critical.    Assuming you put the power LED + 1K resistor on the 9V rail, that will draw at least 6mA, assuming the LED Vf is under 3V.   Therefore you only need 4mA more to make up the 10mA minimum load requirement for the  LM317.   The top resistor in its feedback divider always has 1.25V across it so 270R would draw 4.6mA.   A 1.5K lower resistor would give an output voltage of 8.2V - near enough to 9V for this circuit.  However the reduced voltage means the LED Vf must be under 2.2V or we must decrease its series resistor or we must recalulate for more current in the LM317 divider to meet the 10mA minimum load requirement.   Try 220R for the upper resistor which will draw 5.7mA and a lower resistor of 1.2K for 8V out.

Feeding 8V to the 7660 will give a negative rail of about -6V to -7V depending on the load on it.    That's great for the OPAMP Vee (V-) supply, but its uncertainty is less good for the zero adjustment.    If you put a 2K2 resistor from Vee to a small silicon diode (e.g. 1N4148), anode to the Gnd rail, you'll get an approximately -0.6V rail that's ten times more stable than the Vee supply to the OPAMP from the 7660 (neglecting the temperature coefficient of Vf for the diode).  Tap down on that with a 10K preset feeding a 10R resistor and its easy to get the few mV of negative bias required to get the MOSFET to zero current cut-off, and avoid fluctuations in the 7660 output significantly affecting the load current.

Your latest circuit (#8) has a fatal bug - you've put the panel meter on the output of the 'regulator'.  Unfortunately your unrealistic regulator sim as a voltage source is fooling you because it doesn't draw its load current from +12V, so the return current of the panel meter isn't going through the diode that provides the negative Vee rail.   Connect the panel meter so it draws from +12V and see how horrible it really is.

[VEGETA]

A LM317 is quite predictable and you don't need a 10K pot to set it as the exact value of the 9V output voltage is non-critical.    Assuming you put the power LED + 1K resistor on the 9V rail, that will draw at least 6mA, assuming the LED Vf is under 3V.   Therefore you only need 4mA more to make up the 10mA minimum load requirement for the  LM317.   The top resistor in its feedback divider always has 1.25V across it so 270R would draw 4.6mA.   A 1.5K lower resistor would give an output voltage of 8.2V - near enough to 9V for this circuit.  However the reduced voltage means the LED Vf must be under 2.2V or we must decrease its series resistor or we must recalulate for more current in the LM317 divider to meet the 10mA minimum load requirement.   Try 220R for the upper resistor which will draw 5.7mA and a lower resistor of 1.2K for 8V out.

Feeding 8V to the 7660 will give a negative rail of about -6V to -7V depending on the load on it.    That's great for the OPAMP Vee (V-) supply, but its uncertainty is less good for the zero adjustment.    If you put a 2K2 resistor from Vee to a small silicon diode (e.g. 1N4148), anode to the Gnd rail, you'll get an approximately -0.6V rail that's ten times more stable than the Vee supply to the OPAMP from the 7660 (neglecting the temperature coefficient of Vf for the diode).  Tap down on that with a 10K preset feeding a 10R resistor and its easy to get the few mV of negative bias required to get the MOSFET to zero current cut-off, and avoid fluctuations in the 7660 output significantly affecting the load current.

Your latest circuit (#8) has a fatal bug - you've put the panel meter on the output of the 'regulator'.  Unfortunately your unrealistic regulator sim as a voltage source is fooling you because it doesn't draw its load current from +12V, so the return current of the panel meter isn't going through the diode that provides the negative Vee rail.   Connect the panel meter so it draws from +12V and see how horrible it really is.

So what do you suggest to overcome such issues?

So far we knew all the resistor values necessary, so how can we solve the panel meter issue? I thought the regulator will solve it but you seem to back down from this idea due to simulation.

Yes, simulation shows a bad thing as you said but how can we be sure?

If we don't want the current to pass through the diode or the negative rail, then using 7660 is a must. However, you don't seem to like that idea and you are right since I would need to keep re-adjusting the zero point due to the switching of 7660.

If you put a 2K2 resistor from Vee to a small silicon diode (e.g. 1N4148), anode to the Gnd rail, you'll get an approximately -0.6V rail that's ten times more stable than the Vee supply to the OPAMP from the 7660 (neglecting the temperature coefficient of Vf for the diode).  Tap down on that with a 10K preset feeding a 10R resistor and its easy to get the few mV of negative bias required to get the MOSFET to zero current cut-off, and avoid fluctuations in the 7660 output significantly affecting the load current.

Isn't that what I did in #8? or I am misunderstanding you?

Your latest circuit (#8) has a fatal bug - you've put the panel meter on the output of the 'regulator'.  Unfortunately your unrealistic regulator sim as a voltage source is fooling you because it doesn't draw its load current from +12V, so the return current of the panel meter isn't going through the diode that provides the negative Vee rail.   Connect the panel meter so it draws from +12V and see how horrible it really is.

Ok, so what to do now? I really don't understand why the voltage peaks like this? it is a very low current and we have a high power supply.

If we cannot find a solution, then using 7660 is the only way to go as far as I can see. So I am waiting your input since I tried to go around the diode with no use 

_____________________

EDIT:

Do you mean like the image?

Here I've got 7660 giving -7 and the diode + resistor resulted in around -0.7v which is enough for zero biasing. Then this -0.7 is fed to the 10 + 10k pot as usual for zero adjustment.

However, I feed opamp negative rail with 7660 directly.

How about that? Now, panel meter goes from 9v directly to ground or better yet from 12v to ground. No diode and no 2n2222.

I guess now this issue is resolved, is there anything else?

[Ian.M]

Without a 7660 or some other way of getting a negative rail that doesn't have to carry the return current for the LED meter, the problem is the drastically varying current through the diode or Vbe multiplier that's used to offset the negative side of the 12V supply to get the negative rail.   To be reasonably stable, the variation must be no more than 10% of the average.   

I suspect that most of the meter's supply current goes to light the LEDs.  Least current will be with both displays reading 1.1 and most, 88.8.   1.1 has four segments lit and 88.8 has 21 segments lit (ignoring the d.p.)  so, assuming 88.8 on both displays is the max 60mA consumption, 1.1 on both probably draws about 12mA.    However those will be average currents as its highly likely that it uses multiplexing so it can use a cheap MCU with fewer pins.  Rather than needing 42 pins for direct drive it only needs 13 to multiplex 6 digits x 7 segments.   That alters the situation considerably - peak current will be when the multiplexing is outputting an 8 digit, and least when it is displaying a blanked leading zero.   If there's enough decoupling built into the display it will average out, but it would need about 1000uF to do that, which is a physically large capacitor, and there isn't much room for it so its consumption may well swing between near zero and up to 60mA at the multiplexing refresh rate.   To know for sure, power up the display with a 1R resistor in series with its negative supply, and feed its voltage measurement wire from a pot or variable supply so you can easily change the reading.  Scope across the 1R resistor and calculate the current fluctuations from the peak to peak amplitude of the scope trace in mV.

Yes, your edit above is what I've been suggesting.   If you want to continue experimenting while you wait for your parts/board order, you could build a charge pump with a NE555 squarewave oscillator driving a charge pump (2 diodes, 2 capacitors) to do the same job as the 7660.   Ideally use a CMOS 555 clone not a real bipolar NE555 as the CMOS ones have rail-to-rail output and will more accurately emulate a 7660.  Sim attached.

N.B a practical 555 charge pump will have 100uF electrolytic + 0.1uF ceramic decoupling between pins 1 (Gnd) and 8 (Vcc) to handle the very large current spikes when it switches, and should have 0.1uF to ground on pin 5 (CV).

[VEGETA]

It is done then, we'll consider this the final version and I will start modifying the KiCAD project accordingly. Getting 7660 is easy from China so people wouldn't face a problem. I will use 1N4007/1 in the bias negative voltage since it is what I have and what is available to most people.

Looks like current fluctuations of the panel meter are accurately assumed to be 10-60mA and yes the panel meter doesn't have a big capacitor but lots of tiny circuitry. We were correct to assume the problem in the diode or Vbe in the first place but looks like it won't be solved in this design as long as panel meter exists. However, I guess if it is not in the design and replaced by LCD character display then it would be better. However, that is for another project.

Yes, your edit above is what I've been suggesting.   If you want to continue experimenting while you wait for your parts/board order, you could build a charge pump with a NE555 squarewave oscillator driving a charge pump (2 diodes, 2 capacitors) to do the same job as the 7660.   Ideally use a CMOS 555 clone not a real bipolar NE555 as the CMOS ones have rail-to-rail output and will more accurately emulate a 7660.  Sim attached.

Yes I saw such circuit of using 555 timer but it is just another IC so we are better yet get the 7660 if we want to get another extra IC. Originally I thought we can use the charge pump without switching but that was when I didn't understand the concept behind it.

One last issue is the 10-turn POT. I will mount it on front panel, and will put wires from it to the board. I will make it very short so I will put its footprints very close to the front panel. I hope this won't ruin everything, I don't mind few errors that could be removed by zeroing POT though.

I will build the project again in KiCAD and will post it here. Thanks for your awesome collaboration so far, gonna be credited in the videos and documentation wherever possible.

[Ian.M]

I suggested the 555 circuit for testing because I expect you to either have a few already or be able to find them locally.

I'd still recommend prototyping it before doing the final PCB.   Use that 555 circuit to stand in for the 7660.  We've had enough nasty surprises that I'm reluctant for you to commit to the PCB design till you've actually had it working.   It could be simplified a bit for testing though - only two MOSFETs and a 1A full scale limit.  That will still let you see how well the heatsinks cope.

Due to the very low sense resistance of only 1/5 ohm per MOSFET, you'll need to put those parts + the OPAMP and its whole feedback loop on protoboard as solderless breadboard will have too much contact resistance.   The rest of the circuit should be fine on solderless breadboard.

As you are using a metal case, short unscreened wiring to the front panel pot should be OK.   Screened wiring would be essential in a plastic case as otherwise any nearby RF sources would be far more likely to cause it to malfunction.  The case should be connected to circuit ground.

There are lots of improvements that could be introduced but they all add complexity.   An Atmega328P controlled version, programmed using the Arduino toolchain, with a LCD, and an opto-isolated serial <> USB PC interface for control and data logging would be interesting, but I agree that the best way forward would be to build this version and get experience using it before designing anything more complex.

[VEGETA]

I will need time to get heatsinks from my friend, so by the time PCBs arrive I will get them together.

However, I could try one mosfet + opamp (lm358) with low currents (200mA or so) just to see how stuff works. I don't have enough 1R resistors to begin with if not never since I used lots of them previously so I doubt I still have any left.

For 10-turn pot, should I put a wire from its outer body to ground? or from one of the pins to ground? Also, if plastic case is used can't we get away with the same method of connecting the body of pot to ground?

The case should be connected to ground? that means i wire from the ground pin or pad to the case itself. I don't know if I could solder it or so.

[Ian.M]

Assuming the pot has a metal mounting bush, if the case is grounded, its body will be as well.   If you are testing it loose on the bench, wrap a wire round the mounting bush and clamp it with the nut to ground it.

To ground the case, the easiest way is via the PCB mountings if you are using screws for that.  For one of the mountings, simply use an internally threaded brass spacer with a serrated spring washer between it and the case, and a Belville spring washer on top of the PCB under the screw to keep it in firm contact with the brass spacer.   Use a plated through hole in the PCB with a bare tin or ENIG plated pad both sides larger than the spacer and washer diameters, and within the pad surround the hole with a ring of vias to tie the pads together even if the hole plating gets damaged.  For an example look at any PC motherboard.

It gets a bit harder if you are sliding the PCB into guides.   You'll either need a leaf spring soldered to the PCB, or better, a ground wire with a ring terminal at the case end, bolted down with a serrated spring washer between it and the case.

For testing you can improvise the resistor with a long loop of insulated thin copper wire.  e.g. 2.4m of 24AWG wire should give you about 0.2 ohms.   Fold it in half, and wind it on a cotton reel or similar then tape it in place to make it compact enough to work with.  By folding it in half before you wind it, the winding is non-inductive.   It will have a horrible temperature coefficient and won't be very accurate, but will certainly do to test the circuit.

If you've got thinner wire, you'll need less length.  If you know its actual diameter its easy, we can calculate it, but if not, simply rig a LM317 + a 15 ohm resistor in parallel with a 100 ohm preset as a current source, adjust to 100mA,  put a long length of the wire in series and measure the voltage drop across the wire so you can calculate its resistance per meter.

[VEGETA]

I will slide the PCB inside the case but for grounding I can connect a cable from PCB ground pad to the inside bottom of the case at a place where I scratch it with a screw driver to remove the anodizing\painting then solder the wire directly. This could be even better than putting a washer with bolt.

This is my 10-turn POT:

https://www.banggood.com/3Pcs-3590S-2-103L-10K-Ohm-BOURNS-Rotary-Wirewound-Precision-Potentiometer-Pot-10-Turn-p-1061331.html

So I screw a small wire to the metal part of it and connect it to circuit ground? that is easy.

However, what if we didn't do any of that? what could happen?

I mean will it introduce some error in current like not zeroing out? or will it introduce huge amounts like 10s of mA?

[Ian.M]

Soldering to Aluminum alloys is very difficult without special techniques.  A bolted connection is usually much easier.   If you need a flat surface on the outside of the case, use a countersunk screw, which is usually not too obtrusive.

If you don't ground the pot bush, the current will probably jump about when you bring your hand near it to adjust it.   1mV of noise will cause a 20mA fluctuation of the current in the final four MOSFET design or 5mA for a single MOSFET so it would be very difficult to adjust accurately.

[VEGETA]

As you see, the case has no internal screw holes but only front panel ones, so where would I bolt the ground wire? Here we have no choice but soldering unless the back panel screws can do the job without the wire being exposed outside.

Actually, back panel should have a cut for the DC jack input which in turns have a ground pin. So I could wire it to the back panel (or even bottom part) the way I told you. I could also try to put a wire from ground pin of DC jack to one of the 4 screws if it works fine. I don't think this anodization can conduct current, does it?

as for the pot, I think I can put a wire from the place shown in picture to ground pin... Soldered from both ways since it is easy.

I will try without grounding first to see how stuff works, just in case.

As for grounding the case, is it necessary? also, it won't be mains earth referenced but only to ground of the circuit itself. So what if we didn't ground it?

[Ian.M]

As the circuit relies on millivolt level signals, any noise capacitively coupled from nearby metalwork is likely to upset its operation.   That's the price you pay for a simple circuit with low cost, low-value current sense resistors and a low minimum load voltage.  Bonding the case to the circuit 0V rail prevents any disturbance to the circuit when it is touched.

Anodising forms a fairly tough insulating layer, but its usually broken in tapped or countersunk holes unless they were formed before the anodising process was performed.

You can get washers with solder tags that are big enough to slip over the pot mounting bush to make attaching the ground wire easy.  If you cant get one easily, you can always cut one out of thin brass or copper - shim stock or heavy foil or even out of thin tin plated steel, though that's a lot more work to cut.   If you sand off the anodizing on the inside of the front panel round the pot hole before fitting the pot, it should ground the front panel well enough.

However, personally, I'd drill a hole in the case bottom, countersink on the outside, and sand off the anodising round the hole on the inside then use a countersunk head machine screw, with a serrated washer, a ring terminal for the ground wire and a plain washer and nut to firmly bolt the ground wire to the case bottom, then apply some lacquer to the bolted connection to exclude air and moisture so the serrated washer's connection to the aluminum doesn't deteriorate.

[VEGETA]

I remembered that I have a hand drill! I can use it to make a hole in bottom layer then solder ground wire in it very easily! I just don't make he hole reach the other side but enough to expose the internal aluminum and remove anodizing. As for POT, I would just solder a breadboard jumper wire to it then to ground pin. No need for more complications.


I don't get why you need to go outside the case.

[Ian.M]

Good luck with that.  Soldering aluminium is *DIFFICULT*.   You have to exclude  all oxygen from the surface, scrape off the oxide layer (which reforms almost instantly at elevated temperatures in the presense of oxygen) and solder while maintaining an oxygen free environment.    Its possible to do that in an inert atmosphere, or under oil (but the oil must have a smoke point significantly higher than your soldering bit temperature), and you may have some success with aggressive flux, a large pool of molten solder + scraping through the solder pool, but it will be very difficult due to the massive heatsinking effect of the large thick case walls.

Soldering to the pot bush is *NOT* recommended for plastic body pots as the heat is likely to damage them.  Its acceptable for metal body pots - just solder the ground wire to the back near the terminals.

[VEGETA]

But as you see there is no way to screw something on the main bottom case so I need to get dirty.

As for POTs, I didn't quite understand. You don't like soldering its metal part to ground? OK, but the only option left is to put the ground wire under the associated nut\washer while fixing it. I thought solder would be a better connection. What ground and black wire are speaking about?

[Ian.M]

But as you see there is no way to screw something on the main bottom case so I need to get dirty.

Looking at the case again, it looks like there's enough wall thickness in the sides to drill and tap them so you've got threaded holes for grounding screws.  Due to poor access  you'll have to drill right through and tap from the outside, but it will be easy enough to cut the screws to length and clean up the ends so they are flush with the outside when tightened on a ring terminal + toothed washer from the inside.   With a drop of black lacquer paint on the ends, they'll be barely noticeable.    I'd add grounding screws to both case halves due to the risk of poor contact beween them.     If you've never tapped aluminum before, use WD40 as lubricant and practice on something extruded that doesn't matter.   Personally I'd go with M3, but M4 would have a lower risk of breaking the tap off in the hole.  The tap *MUST* be held absolutely straight as if it starts on an angle it will cut a 'drunken' thread that will be much weaker as the hole forces the tap into line and part of the initial drunken thread is cut away.   It can be helpful to drill a slightly tight guide hole through a block of wood, tap the guide hole and clamp it to the part you are tapping so the tap is held straight and encouraged to advance as it first starts to cut.   See https://en.wikipedia.org/wiki/Tap_and_die for more details.

Another method that requires less skill with precision hand tools would be to use self-adhesive copper tape with conductive adhesive.   Sand off the anodising round the end plate screw holes on the inside of the end plates and the ends of the two case halves.  Stick a strip of tape to the end of the case half, extending a short distance down the inside of the case side, covering the hole the screw will go in,  pierce it for the screw and clamp it firmly between the case half and the end plate using the screw.  Dismantle again, trim the outside edge with an Xacto knife, and solder your ground wire to the copper tap on the inside of the side.   Similarly, stick a strip of tape to the inside of the face plate over the pot hole, cut out the hole with the Xacto knife and solder the ground wire to the tape clear of the mounting bush footprint.  If you decide to improvise with copper foil and ordinary adhesive tape, only put the tape to stick down the end of the foil you are going to solder to as its an insulator and mustn't be between the foil and the area round each hole from which you've removed the anodizing.

As for POTs, I didn't quite understand. You don't like soldering its metal part to ground? OK, but the only option left is to put the ground wire under the associated nut\washer while fixing it. I thought solder would be a better connection. What ground and black wire are speaking about?

I said 'back' not 'black'

A metal bodied pot usually has a tin plated pressed steel back cover.  Because its fairly thin, it doesn't take a lot of heat to solder to - just scrape off any lacquer to expose a small spot of bare metal, tin it then solder the wire to it.  OTOH soldering to the bush of your pot would take a *LOT* of heat, which would cook the lubricant and may melt plastic parts of the mechanism or loosen the bush in the body moulding.

If you clamp something between the bush and the front panel it needs to be thin and *FLAT*.   Trying to directly clamp the wire is likely to make the pot sit at an angle, and due to the small contact area there is a high risk of the wire squishing and the pot loosening.  Its much easier with a tag washer even if you have to make it from brass shim yourself.

[VEGETA]

I have made a new schematic in KiCAD and here it is as .pdf in attachment.

Key points to be taken:

1- Made maximum current set resistor 750K since it is a standard value which now gives 2.11A as max current, no POT needed.
2- Adjusted LM317 circuitry to be 240 Ohms || 1.5K which gives 9.06V but the 1.5K is 1K in series with 1K || 1K.
3- Added ICL7660 negative rail generator with our agreed diode + resistor reference which feeds 10K POT and 10R to ground. 10K wiper is V_bias while other pin is not connected since it is in variable resistor configuration.
4- Power On diode is now available which takes its current from LM317.
5- Panel meter shunt resistor is now connected as it should be.
6- Added some caps here and there.
7- Labelled power rails correctly. Vdd = DC input, Vcc = 9v, Vee = -9v, V_bias = negative bias from zeroing POT.
8- Arranged Op-amps + MOSFETs circuitry in a civilized way (or so I thought) for easier reading.


Looking forward to your remarks on the schematic, so I can dig into PCB right away.

[Ian.M]

I haven't spotted any obvious mistakes, but I haven't put hours into checking it.

IMHO the modern trend in amateur schematics to make heavy use of net labels instead of drawing the actual interconnections is reprehensible and should be strongly discouraged.  At it worst you get schematics that are nothing more than a pinout for every chip with a net label on each pin which are absolutely horrible to trace signals on except in the original schematic editor by hiliting the whole net.   Fortunately yours is much better than that.

I would break it down into far fewer sections.  e.g:

Draw straight through from the pot + full zero bias adjustment circuit, then the the current sink section (OPAMP, MOSFET, sense resistors and feedback loop),  then paste three more copies of the current sink section below it with I_set as their input.  The negative rail generator, positive regulator, DC in and power LED all belong together, which just leaves you with one block for the D.U.T. PSU and meter connections.

I don't know if KiCAD has usable hierarchical blocks without having to set up a hierarchical sheet.  If it has, it would be preferable to use them rather than having four  clones of the current sink section.   

[VEGETA]

I will try to re-arrange it if I have the time tonight. In terms of good schematics, can you mention few points to follow?

I would rather prefer if you have any last thing or note on the schematic to say it now before digging into PCB. Probably I will wait for another day to be able to work on it.

In KiCAD there are global labels and local labels. I've used global since they appear nicer and there is no other schematic embedded within this one. Instead of doing hierarchical schematics, it is better to have just another page right? I still don't know how to do that yet in KiCAD. I don't mean another .sch file but just another page within the same one.

 

[VEGETA]

V_load v0.2

it is updated schematic as we agreed above, kindly check it out.

now it is less-newbie with wiring instead of labels... more organized blocks too.

I will consider it final, to begin in PCB soon. Hopefully before Ramadan ends (~ 2 weeks) since I won't do a thing during feast.

I still need to know how to define board shape in KiCAD as well as how to know the actual board size which depends on the case itself. I received the case but it is not with me now, I will get back home and then I may be able to measure it using my digital caliper.


Aside from that, wouldn't putting some thermal vias help (even a bit) in thermal dissipation? just saying.

EDIT: Fixed the PDF.

[ledtester]

You might be interested in this kit:

https://www.aliexpress.com/item/DIY-Kits-150W-10A-battery-capacity-tester-adjustable-constant-current-electronic-load-discharge-Test/32870007246.html

The schematic looks very similar to what you are doing (see attachment).

Another forum member turned this kit into a completed project. You may be interested in what he and other forum members had to say about it:

https://www.eevblog.com/forum/projects/upgraded-lm324-based-150w-72v-10a-electronic-load/msg1506610/

Note, also, that the PCB has footprints for both TO-220 and TO-247 MOSFETS - this makes it easy to experiment with different load transistors.