Help with PSU (Temp control fan, load bank, & PWM circuit)

[()blivion]

Good times! Things are looking good with this.

@PCB.
There is a free program called PCB123 that lets you make a PCB from a schematic, view what that kind of board actually turns into in 3D, then when your ready, you can order it right in the software. That is the 1 2 and 3 part. They also allow paneling multiple circuit on one single board.

There is also **broken link removed** cheep low density board service. You have to find the right configuration to get max $ value though. But you get 5 boards, at whatever size you choose. PLUS you can open source for 2 more boards free of charge. So you can get 7 PCB boards, and you can panelize. You can make them any color you want too for just +$10.

$60 for seven 10cmx10cm plain boards. Much room to expand.

Only problem is you have to supply them with multiple gerber files. It's a standard PCB CAD file though.

Edit: I have never used the above service, so I can not speak for their quality.

 

[jocanon]

Thanks ()blivion...I'll check those out.

 

[jocanon]

I have a question to pose. I think I already know the answer but would like to get some expert input. Going back to the temperature control circuit we made for the PSU...there is also a current share pin on these units that voltage goes up on it when current increases and down when current goes down...point is it seems there would be a simple way to tie in pin 4 (fan control) with this current share pin and control the fan speed strictly based on load current draw. Question is, would this be a superior way to control the fan speed or not? I still think the temp control circuit we made would be superior for a few reasons. One, there are other factors beside current that will effect the temp if the PSU like outside ambient temperature, I know in Tucson it gets well above 100F many summer days for example, it's possible these could be operated outside with a generator. Another reason is that as the PSU current goes down the internal temperature lags and will not be lower until a time after current level decreases and he PSU might like having the fan run faster to get it cooled down (on the flip side, the speed up will lag current increase because the temperature will not instantaneously rise as the current level rises, but I think this lag is fine as heat is the real problem we are trying to control so even though current level goes up, we don't really need the fan speed to increase until the increase in current causes the temperature in the PSU to rise). In short, I think linking the fan speed control directly to the internal temperature of the PSU is a superior method versus linking it directly to the current level draw of the load connected to the PSU. Any concurring/differing opinions? All input welcome.

Edit:
Another argument in favor of temperature control, it seems that most other devices control fan speed based on temperature, in fact, there is a PSU that is very similar to the DPS-600PB that uses a temperature controlled fan, not current controlled.

 

[ronv]

I'm with you. The fans are there to control the temperature so it is best to measure that rather than a secondary effect. I can hear my other PC's fan going up and down now and it is not doing anything.

Have you built the 24 volt controller yet?

 

[jocanon]

Thanks ronv...actually, I was just going back through the posts to find the schematic for the 24v controller, it's on post #481. I have the day off work, so I am going to build it now.

 

[jocanon]

I am having some trouble with the 24v fan controller. At first I couldn't get the fans to turn on at all, but then I realized I had 75K ohms between the op-amp output and the gate instead of 75 ohms! I don't have a 75 ohmer laying around so I used 68 ohms here for now. As it is, I can adjust the speed of the fans between min and max with the 10K pot but only in like the first 1/10th turn of the pot, so when the pot is turned all the way to one side the fans are going at min speed, but then when I just turn it about 1/10th of a turn the fans are on full. Also, the thermistor seems to have no effect on anything. I tried shorting the thermistor and it didn't change fan speed at all. I am sure I have something wired wrong (other than the 75 ohm resistor), but I have checked it over and over and don't see what I am missing.

It is difficult to see what's going on in the pics attached so I labled what each wire going out of the picture is in the first pic and I will describe here how I have everything connected so hopefully you can help me see where I am going wrong:

-I have one side of the pot to +24v, center wiper to +op-amp input, the other side of pot to ground.
-the 27k resistor to +24v and the thermistor/one side of the 47k resistor
-the other side of the thermistor to ground
-the other side of the 47k resistor to the negative input on the op-amp
-the 330k resistor from negative input on op-amp to op-amp output
-75 ohm (68 ohm right now) from op-amp output to gate of the transistor
-transistor drain to +24v
-transistor source to the + wire on the fans
-negative wire of the fans to ground

 

[ronv]

It looks like the problem is with the transistor. This one is a BJT transistor so the pin out is different:

https://www.electro-tech-online.com/custompdfs/2012/10/CD00000942.pdf

So pin one is ok - base of transistor.
Pin 2 should go to +24
And pin 3 to the fan

 

[jocanon]

Actually, I think it was a bad thermistor. I am pretty sure I had it hooked up the way you described above, pin 3 to fans pin 2 to +24, I think I just had the terminology wrong. I switched out the thermistor and the temperature control works great now. The pot is still very touchy, but I can live with that as most of the time I will not need to be adjusting it.

 

[ronv]

Cool, I guess it is pretty sensitive because it has gain. Didn't think about that. Is the range okay?

 

[jocanon]

Yes, I think the 24v fan controller is good to go. It would be nice for th pot to have a little more room for adjustment, but like I said above its not a big deal so I think I am going to build this circuit now.

On another note, I was just testing out the placement of the thermistor for the 24v fan controller and while doing so I loaded up the dummy load to max with 47 amps at 24 volts and after about 1 min at that current the TVS cracked again. Everything still seems to work, it didn't short out this time like it did last time, but with the crack in the casing of the TVS I shut it down. So apparently the TVS is not quite up to the task. I think I better get one that can handle more current.

 

[jocanon]

Looking at the specs on the TVS:

https://www.mouser.com/Search/ProductDetail.aspx?R=5KP26Avirtualkey57610000virtualkey576-5KP26A

It seems to be well within spec, except maybe the operating temperature. Since I have it connected directly to the current sense resistor (the ground side) I wonder if the heat is traveling up the TVS and that is what's cracking it. The max operating temp is +175C, remember when I took readings of the temperatures and it was close to that or above...so maybe I just need to connect to ground a little ways away from the current sense resistors to get it away from all that heat.

 

[ronv]

That could be it. It would make more sense. For the TVS to do any work at all the voltage has to be to high. Maybe you could move it back to where the PSU comes into the board?

 

[jocanon]

I am soldering the actual 24v fan control circuit now. I remember when we did the dummy load ()blivion said it is good practice to tie up all the loose ends on the op-amp circuit that we are not using, but if I recall he said it may not be necessary on this op-amp. So, do I need to do that, or can I just leave the unused pins on the LM-358 open?

 

[ronv]

You can probably leave them open.

 

[jocanon]

Works great! I got everything all buttoned up now on the dummy load. Wish I had time to take some pics and video but it is just too late, I will have to do it tomorrow.

 

[jocanon]

I took some video of the dummy load in action! I cracked another TVS diode though...I wonder why these things keep cracking? This time it cracked again when I was demonstrating the over current shut off protection of the PSU (I thought connecting it to a cooler spot on ground would solve the problem so I thought it would be save to over load the PSU again...guess not). But the last time it cracked I wasn't even over loading the PSU, just running it at 47 amps, so the over current protectiond didn't shut off the PSU last time and it still cracked the TVS??? Do I just need a better TVS? What would be the risk of just taking it off all togather, bad idea?

After I upload the video to YouTube and I will post a link to it.

 

[ronv]

I'm a litttle concerned that the supply is going flakey at high current. Are you running it as a 24 volt supply or 12 volt when it blows?
Maybe you could measure the AC and DC voltage of the supply at various loads.

Maybe someone has some other ideas.

 

[jocanon]

Yes I was running it at 24v.

 

[jocanon]

Assuming you mean to measure the AC & DC voltage of the DUT not the logic supply, right?

 

[ronv]

Right. For the TVS to blow the voltage has to go above 26 volts, but I am thinking back to the ripple that gave the fan controller a bad time. If it is high frequency the voltmeter may not measure it, but it is worth a try.