Yes.... Another Dummy Load

[wigman27]

Well,

 I have replaced the LM324 with the Bordodynov library one followed Kevin.D's values exactly and stepped the DAC voltages and load voltages and its working perfectly

I have also tidied up the spice schematic so its a little more readable and lets go with that I think

Thank you all VERY much!

Lee

[Jay_Diddy_B]

Lee,

You should add a damping circuit of 2.2uF and 2.2 ohms from the MOSFET drain to ground.

If you add some inductance, 1-10uH between the MOSFET source and the Power being tested, the circuit might oscillate.

This inductance is to represent the inductance of your test leads.

Jay_Diddy_B

[TerminalJack505]

So far as I can tell you have two problems. 

The other problem is voltage offset.  By using a 0.1R resistor you're getting a gain of 10x as well as seeing 10x worse offset.  Which could account for up to 70mA of error on the output with the LM324's worse-case offset of 7mV.

Hi TerminalJack505,

Thanks for your reply

My original circuit did have an offset problem as it was using a x10 gain op amp picking up the voltage drop across the 0.1ohm resistor. However with this circuit I don't have any gain in the op amp and offset isn't really a problem a the threshold voltage of the MOSFET is around 1.7v so 7mV offset hopefully won't worry me too much.

...

You still have a gain of 10.  1V on the input leads to 10A of current on the output.  That's a transconductance gain of 10. 

Note how the op amp's voltage offset is multiplied by this gain in my attachment.

(I had to use a negative rail to avoid the fact the LM324 can't quite get down to 0V with a single supply which, otherwise, makes the offset seem worse than it really is.)

[wigman27]

Lee,

You should add a damping circuit of 2.2uF and 2.2 ohms from the MOSFET drain to ground.

What does this do? I see that your schematic has the same values also, is it a fairly standard practice?

Lee

[Jay_Diddy_B]

Lee,
You asked, so let me explain.

When I was developing my Dynamic load I was trying to get high bandwidth for doing transient load tests. I did my control loop without considering the characteristics of the test leads that are used to connect the power supply to the load.

When I tried the load I found the current oscillated at around 200kHz.

So I thought may be the inductance of the leads going between the power supply and the load were causing problems.

I estimate that the inductance is between 1 and 5uH.



I can make a small signal model:



These are the results, the current splits between the source and the damping network depending on the frequency.




Without the damping network, the source impedance rises.

Adding the damping network let me get a high bandwidth dynamic load. The damping network is really a function of the lead inductance and resistance.

Jay_Diddy_B

[wigman27]

Hi again!

That makes perfect sense, thanks Jay_Diddy_B

As a matter of interest to others, Neilm posted this document in another topic. It explains what happens to op-amps when driving capacitive loads perfectly!

Thanks again everyone!

Lee

[wigman27]

Hi,

I have been having a play with the power supply for the control circuit and I would like to incorporate the 5V auto selector type system from the Arduino Leonardo to swap between VIN and VUSB. (schematic attached)

As I will be powering the circuit directly from 5V adapter, I will not be running the supply through a regulator and then "isolating" VIN from VCC 5v and I don't have 3.3v, so the Arduino circuit won't work.

I have had a play with a few things, the second attachment below is using a Schottky diode to "isolate" VIN from VCC which works but I really don't want the 300mV (or more) volt drop across the Schottky.

So the Third attachment is an Oring circuit that I have played with but seeing as though the voltages are very similar, both MOSFETs work in parallel instead of isolating the USB voltage.

I know the OP amps aren't setup yet for driving capacitive loads but I am just trying it out at this stage

Does anyone have any other ideas I could try?

[TerminalJack505]

You need to give one of the supplies priority.  You might look at the circuit in this post for some ideas.

[wigman27]

Thanks TerminalJack505

I had actually seen that post in my original searching for an answer but I didn't really see anything without a Schottky or your version, which to be honest, I am having a little trouble working out exactly how it works.

My priority supply is defiantly the VIN 5V Jack so I think I will either bite the bullet and go with a Schottky, or use a relay. the relay I was thinking is attached below, very similar to the one suggest in that other post.

The main reason I don't want to use a Schottky in series with VIN is because the op amp I am using to drive the MOSFET needs to go up to about 3.2V to drive 8A through the FET so I don't really want to drop VCC any more than I have to.

I could use 9V power on the opamps and a 5V reg for the rest of the circuitry but I really want to use 5V USB as a secondary supply. There are many many ways to skin a cat but I think these two may be the best for this purpose.

What would be your pick for this?

[TerminalJack505]

The relay seems like a good solution.  I don't know if it would work properly for "hot switching" (or whatever the term is.)

So far as how my circuit works, the two leftmost comparators are acting like a window comparator with a little feedback via R4 for a Schmitt trigger type of action.  They sense the primary voltage via R5 and R6 and output a low voltage only if the primary is between the two voltages set by R1, R2, R9 and R3 (I just picked an arbitrary low and high voltage.) 

The diodes on the first two comparator's outputs allow the outputs to be tied together so that they work in a wired-AND fashion.  This output is feed to the next stage which turns on either the primary or secondary, as appropriate. 

Note that the MOSFETs are back-to-back to prevent the body diodes from conducting.  If this isn't done and just a single MOSFET is used then you risk having one supply force current into the other OR (if the body diodes are the other way) having the supply with the higher voltage do all the work, despite which one is switched on.  (The body diodes would just act as OR-ing diodes in that case.)

[wigman27]

One day someone will search the forum before posting, today is not that day...

add 100 ohm between the op amp output and the mosfet gate, and 1n between the op amp output and the inverting input, this will slow down the loop and make the op amp more happy with driving a capacitive load,

Hey Rerouter,

Today is that day!! I just needed an answer, searched the forum and found the answer!! thought you may be happy to hear it!

Happy Tuesday

Lee

[wigman27]

Hey All

I have finally got a schematic together, I haven't fully tested every component together yet so it may change without notice.

Some feedback would be awesome!

Thanks Lee

[Harvs]

For the fan, I would be inclind to just use a regular PC fac (which maybe that is?), as they include their own driver circuitry.  Then you just send it a 25kHz PWM signal to set the speed, no need for additional circuitry.

I would encourage you to provide a bias supply for the opamps, check out the jelly bean 7660.  For some reason people seem to be shying away from this, but it'll add about $1 worth of parts to add a cap voltage inverter and it'll mean you can actually operate down to 0A.  You can also then add a small negative offset to the DAC output so it can go down to zero as well, because it'll only go to several mV above zero at the moment.

[wigman27]

Well I have finally finished my first PCB design!!

I'm pretty happy with it, I would love to hear suggestions on stuff I could do better and differently

I have attached the top and bottom layers separately.

Thanks again ;-)

 

[Jay_Diddy_B]

Hi,

You might want to consider a larger connector in the PL1 position. The connector you have looks a little small for your maximum load current.

Jay_Diddy_B

[wigman27]

Hey mate,

Thanks :-) as I was designing the board I had the same thought, my max load is 8a and the trace width is 5mm. Molex do a connector that will do 10a with the 100thou pin pitch. So I decided to run with that for now and may go bigger for rev b if it needs it.

What do you think?

[Jay_Diddy_B]

Hi,

I think 8A on a 0.1" connector is a stretch. You can always solder some wires directly to the tracks if it becomes a problem. I did a quick check on Digikey and I found that the 0.025" square post were 4A max. The mating connector was sized for 22 awg max.

On the 0.1" connectors they are specifying 15m Ohms. At 8A this 960mW per pin 

The larger 0.156" center connectors with a 0.045" square post are specified at 7A and will take an 18 awg wire.

Jay_Diddy_B

[wigman27]

Yes, very good point!!

Thanks, I will go with a soldered wire this time and fix it in b. that will make it easier to get the tracks in.

Where I have put the larger track into the MOSFET at the top of the board, is that the best way to shrink them down to fit the pin pitch? Have you got any examples of how better to do it if not?

Thanks again :-)

[Jay_Diddy_B]

Hi,
In the Dynamic Load Project:

https://www.eevblog.com/forum/projects/dynamic-electronic-load-project/msg288313/#msg288313

 I used both sides of the board for the MOSFET connections. I had the Drain connections on the same side as the MOSFET and the Source connections on the other side. If you do not use an insulated device or a silpad, the heatsink is at the Drain voltage.

Here are the pictures:





Jay_Diddy_B

[wigman27]

Well the PCB is here!!!

I cant wait to try it out!!

[wigman27]

Yes, It has been a very long time in the making.. but here is the finished version 1 product

I have learned a lot!! which is the main reason for doing it!! But all in all it works as designed (after a number of newbie issues).

I had a few footprint issues which is my fault because I check nearly everyone except the processor!! dir!! Murphy strikes again and I stuffed up...

There is still a bit more refining of the software and hardware but im pretty happy with my first real electronics design and build!

On to version 2! try and implement isolation plus a number of other things 

A big thanks to everyone who helped!

Lee

[wigman27]

Version 2 is now finished!

To see the finished product I have started a thread in the open source hardware section that shows a link to my instructable and videos that I have done

https://www.eevblog.com/forum/oshw/arduino-programmable-constant-current-power-resistance-load/

A huge thank you to everyone that helped me with this! Most of all Dave that inspired me to do it in the first place!

Thanks again

Lee