Test Equipment Anonymous (TEA) group therapy thread

[PA0PBZ]

Vince,

Since you are going to operate the mosfet in the linear region rdson is not really important since this is the minimum resistance when fully on. What you have to watch out for is the DC Safe Operating Area, which should be in the datasheet. Looking at the 2SK2461 DC SOA it can only do like 3.5A with VDS 10 volt, so that would not even be enough to test this PSU. (https://pdf1.alldatasheet.com/datasheet-pdf/view/6123/NEC/2SK2461.html , page 3) The IRF630A would be a better choice here, it looks like it can do something like 7-8A with VDS 10V. (https://pdf1.alldatasheet.com/datasheet-pdf/view/52980/FAIRCHILD/IRF630A.html , page 4). However, none of those are made for linear operation and it could have a dramatic end... ymmv. 

[Vince]

Thanks.

Didn't know there MOSFET "designed" to work in the linear region...
I am not famiiar with this situation, never studied the subject in any particular detail...so I was hanging to the basic stuff : if it can do X amps then Vds should be able to be anything up to max voltage rating, as long as the max power handling capability of the transistor is not exceeded. I do'nt understand how a transistor could not deliver it's current if its still within the allowable power curve ?
See, I am clueless.

However as I said the idea was to spread the load across several identuical MOSFET precisely so they are far away from their max current and power envelope.

Hmmm, OK. So let's be transparent here... here is everything I have power MOSFET wise (N type)... if someone thinks one of them would be a decent candidate.

Part Number   Qté   Package   Specs / Description
2SK903      TO220   850V 5A 2.4 ohms
2SK1819      TO220   450V 5A 2.5 ohms
2SK2461   8   TO220   100V 20A 0.1 ohm
2SK2545      TO220   600V 6A 1.25 ohm
2SK2750      TO220   600V 3.5A 2 ohms
2SK2761      TO220   600V 10A 1 ohm
2SK2771      Plastic Big   900V 9A 1.1 ohm 100W
2SK2996      TO220   600V 10A 1 ohm
BUZ73      TO220   200V 7A 0.4 ohm
FQP630   4   TO220   200V 9A 0.4 ohm
IPP60R280P7      TO220   600V 12A 0,28 ohm
IRF530      TO220   100V 10A 0.16 ohm
IRF630A   3   TO220   200V 9A 0.4 ohm
IRF630M   5   TO220   200V 9A 0.4 ohm
IRFBC30      TO220   600V 3.6A 2 ohms
IRFBC40      TO220   600V 6A 1.2 ohm
ON4521      TO220   600V 4.5A 1.6ohm
STP3NA60   TO220   600V 2.9A 3.3 ohms
STP3NK60ZFP   TO220   600V 2.4A 3.6 ohms


From what you say the IRF630 is good enough for the job here, so that's good to know. I will check the datasheet to see how many in  // I have to put so that the load can dissipate 55W very comfortably.
Oh, just looked at the datasheet.. it can do up to 70/75 Watts ?! 
Hmmm... so me asking 55W, it might survive, given that I would I would ask so much power for only a few sconds, just enough to see how the output voltage behaves, then dial it way back down immediately.

Hmm cool, so for my immediate trouble shooting needs, a single IRF630 could get the job done, nice ! Not ideal of course, but still good enough for my purposes here.


Now I need a schematic, I am listening 

[BU508A]

Electronic DC LOAD
[..]
So it would be fun if I could cobble something together with what I have at home. So just a really crude, minimalist, super simple, ugly as fuck, " does the job" kind of thing !
[...]

Use some fat resistors and throw then in a water bath. That is a very easy, cheap and effective solution.

Edit:
Added an article from IXYS about power MOSFETs working in linear mode.
Source: https://www.ixys.com/Documents/Articles/Article_Linear_Power_MOSFETs.pdf

[Vince]

I need to vary the load continuously and finely to test that PSU, using resistors is not practical.

You are missing the point, as I said I WANT to build a little load and this PSU loading situation is just the ideal pretext / excuse to make one 

EDIT : just saw your edit... thanks for the article, looks interesting, will have a read.

[Messtechniker]

Simply search for "Electronic load circuit".
Will keep you busy for a few hours....

Probably as simple as it can get for130W briefly. See attachment.
You may even ignore the 555 part, if dynamic testing is not needed.
 

[BU508A]

I need to vary the load continuously and finely to test that PSU, using resistors is not practical.

You are missing the point, as I said I WANT to build a little load and this PSU loading situation is just the ideal pretext / excuse to make one 

That is why those things were invented:


One can easily add more load by simply clipping more resistors in parallel to the load. No need for continuous mode in this case. Current can be measured with a DMM of your choice.

Btw, building a DC load may not be as easy as it sounds. ;-)

[Vince]

Simply search for "Electronic load circuit".
Will keep you busy for a few hours....

Yes I am sure ! That's why I asked here about suggestion on what circuit / design to use, precisely to save me valuable time !.. and avoid picking unknowingly some obsolete or faulty or flawed design, when there was probably a much better overall design out there that every body new about and was using for a long time, tried and tested.

That's what I call " asking for advice " ! 

[Vince]

I need to vary the load continuously and finely to test that PSU, using resistors is not practical.

You are missing the point, as I said I WANT to build a little load and this PSU loading situation is just the ideal pretext / excuse to make one 

That is why those things were invented:


One can easily add more load by simply clipping more resistors in parallel to the load. No need for continuous mode in this case. Current can be measured with a DMM of your choice.

Btw, building a DC load may not be as easy as it sounds. ;-)

Those test leads aren't quite up to the job 
And again you are missing my point.... I want to make a load, and you are denying me this, bad boy !

Difficult ? Come on, the one Dave made from memory had a tiny PCB the size of a matchbox, with a single TO220 MOSFET on a heatsink, and a pot to adjust the load, that's all...
That's all I want to make, no more.

I think I asked the wrong audience... I did say I wanted something crude quick and dirty, that does the job, nothing more.

I guess I will have to open a dedicated thread about it....

[Vince]

OK did a quick search, looks like the basic topology is this :

So a MOSFET, with a current sense resistor in the source, and an op-amp that drives the gate of the MOSFET in whatever way it sees fit, comparing the input from the pot, and the voltage from the current shunt.

[Zoli]

I need to vary the load continuously and finely to test that PSU, using resistors is not practical.

You are missing the point, as I said I WANT to build a little load and this PSU loading situation is just the ideal pretext / excuse to make one 

EDIT : just saw your edit... thanks for the article, looks interesting, will have a read.

I have  couple of spare boards(38X48mm)/parts(all SMD, except the FET and pot), and fits your 2SK2771(which is good for 100W DC), and the project is KiCAD 5.x; interested?

[Vince]

Yes sounds good Zoli, let's do that !!! 

[Zoli]

Yes sounds good Zoli, let's do that !!! 

Mail sent.

[Vince]

What would I do without you ! 

[TERRA Operative]

My vote is for a load bank using wound nichrome wire and knife switches.

[Zoli]

My vote is for a load bank using wound nichrome wire and knife switches.

30  years ago the was my choice, too; 4Ω made of 4X0.5mm winded on 3mm FR4 dissipates 400W in a cup of water

[Zoli]

PCB's and built unit; Vince, can you handle the 0603's?

[Vince]

PCB's and built unit; Vince, can you handle the 0603's?

Uh....... I don't know but I guess I will soon find out !! 

Luckily there are only a few of them and I only need to make one board so... I will do the best I can !
I don't think I can make it look "pretty", but I think I can make it "work"/functional.... the electrons won't be able to tell the difference I think....

[Vince]

My vote is for a load bank using wound nichrome wire and knife switches.

30  years ago the was my choice, too; 4Ω made of 4X0.5mm winded on 3mm FR4 dissipates 400W in a cup of water

Well ideally I would just use a huge wire wound, antique-looking rheostat, but these things don't know where to buy and would cost a fortune. Copper and metal are expensive there days, whereas MOSFET and heat sink, I have already at home, zero expense.

Hell these rheostat even 30 years ago probably already cost a fortune. I now remember my physics teacher in high-school who told us about how silly expensive these were, and that the one he was using in front of us for an experiment that day, he got for dirt cheap from a flea market from a guy who had no clue what he was selling !

Would love to make my own wire wound beedy resistors like you did Zoli, but I guess the special wire that's required must cost a pretty penny per meter, and that I would need a selection of various gauges and resistivity to cater for all needs, which I guess would soon ad up to quite an expense.

I am interested though, because I think I may have blown the shunt from the ammeter / galvanometer in one of my old power supplies. I put the two floating outputs in // and I don't think it liked it... now the ammeters display garbage. I think they must have overheated briefly and maybe there resistance has changed now, so the galvanometer displays non sense.
This future repair of mine is also an excuse for me to get a decent bench meter with 4 wire measurements, so I can measure accurately the shunt resistors in this power supply, to confirm my diagnosis (and help me wire / trim my own replacement shunt if that's indeed the problem).

No, I won't be asking for schematics for a cheap milli-ohm meter, I will just splash some cash and get some decent bench meter. Which means it won't happen any time soon...

[TERRA Operative]

Pull the heating wire out of a few old toasters, or find some of those spiral elements from some old electric stovetops.
They are good for low ohm, high power resistors.

Then just stick your new resistor in an old paintcan full of oil and job done.

[Vince]

Sounds like a cool little week-end project ! 

Problem is to get hold of "old toaster  "... there is no such thing as getting old stuff for free here. People either sell them on leboncoin.fr, or dump them in their local junk yard, where all domestic appliances are stored in a secured dedicated container with CCTV watching, where it is strictly forbidden to steal/pick anything. You can only add to the container, not take anything from it... I tried, just once, and got caught on camera within 20 seconds, with the guy operating the plant screaming at me : " Don't you do that, put it back : You want me to lose my job or what ?! It's strictly forbidden don't you know ?! ".

OK OK..... no need to scream, I am not deaf...

But yeah, if I ever find an old toaster somewhere, I will think of bringing it back home with me !

I do have a toaster, which I use one a year at X-mas to eat some foie gras, but that's about it. With the little use it gets, I will probably die before it does... so it will probably not become a donour anytime soon I would think...

PS : what oil should I use to fill the can ? Does olive oil give better performance or is sun flower oil good enough ? Can I use waste engine oil ? I need to replace the engine oil in my car right now.
It's 10W40 semi-synthetic, is that good enough  or is olive oil better ? 

[Peter_O]

Shot these two guys last year, when a real load was needed, not an electonic one.   
10Ohm/8A and 100Ohm 2A


Paid some 20 Euros a piece.

the 10 Ohm ist heating away with 280°C at 8 Amp.   

Why messing around with wire and water when you can get fine TE.   

[Vince]

Yes that's what I had in mind.... though I would love to find some vintage looking ones, with a wooden base and nice metal work,  a cast iron wheel and worm screw to move the wiper along......I should start attending local flea markets but I am too lazy to go out on a week-end morning... only work can force me to wake up in the morning, because they actually pay me for that, so I feel obliged but.... other than that 

[Robert763]

I'm a bit late to this but for a load capable of what vince needs just 3 IRF603s on a decent heatsink and a psu. Tie the 3 drains (easy don't use insulators but remember heatsink is "hot") for load positive and the 3 sources together for common. Put a 100R at each gate (short lead to gate, this is to stop oscillation) tie the other end of 3 resistors together and put a 4k7 from there to the common sources (acts as pull down to turn MOSFETs off). Then you just put a low voltage PSU across the 4k7. Increasing the voltage increases the lood current.
All the action happens between about 3 and 5V on the gates so fine control is needed. Don't exceed 15V on the gate or you will blow it.
As an alternative to the variable supply you can use a 9V battery and a 10K 10 turn pot. Don't fit the 4k7, tie the wiper to the gate resistors. battery negative and pot CCW end to source connection, battery positive to pot CW end via on/off switch. 

Robert G8RPI.

[Vince]

Wow I asked for quick and dirty, it doesn't get any better than this, not even a schematic to talk about, just instructions ! 

Will try and play with that tomorrow (I have the day off), will keep me busy while Zoli's parcel is crossing the big pond with a cleaner and more durable solution 

Thanks Robert 

[tautech]

Wow I asked for quick and dirty, it doesn't get any better than this, not even a schematic to talk about, just instructions ! 

Will try and play with that tomorrow (I have the day off), will keep me busy while Zoli's parcel is crossing the big pond with a cleaner and more durable solution 

Thanks Robert 

You think.
40 years ago we purchased a new 180A welder from the manufacturer and one of the techs that lived out these ways delivered it with fresh paint barely hard.

In conversation he revealed he had been on the team testing these for NZS certification which for welders meant the transformer core temps had only 1 standard they would target, the best so they could be certified as 100% duty cycle. These welders, a 2 thyristor phase control design using all copper windings, primary 3mm, secondary 3 x 10mm copper strap unfortunately went 1C over max temp allowed for a 24hr test at full power.

The load, a large coil of heavy fencing wire in a 200L drum of water that required topping up for 24 hrs.