Maker Pro
Maker Pro

questions about increasing voltage to power supply

bonedoc

Dec 21, 2011
122
Joined
Dec 21, 2011
Messages
122
Update- I have done some fine tuning. It turns out that I am now able to charge to 300V in under 1 second with a 12V NiMH battery pack. That will save the cost of 2 cells, and also allows for a slightly smaller voltage rating on other parts. The heat sink on the mosfet is slightly warm. The hottest it has gotten is 55C.

With several minutes of repetitive charging from 0 to 300V, there is some stuttering in the circuit. At this moment, the heatsink is fine, but the PTC that regulated the current is very hot. It was 80C+. This is pretty much the max temp for the PTC.

I have tested the current draw. For the most part, it stays in the 3-5A range for the total circuitry. This is what passes through the PTC. I looked at the datasheet, and it was one of those things where the website search was misleading.

The "max current" was 7.5A. I thought this was reasonable. But nope.... the operating current is actually .75A with a trip current of 1.5A. I think what was happening is that it was heating up from over use, which increases the resistance, which increases the heat....and then it starts the stutter process.

Can you tell me something?

If my circuit typically draws no more than 5A, what size of PTC should I get? Say I get a PTC with an operating current of 7A is is fine to have that very brief 20A spike?

If my circuit goes out of whack, I think it will go way over 7A, but I dont know what range causes a safe shut off, yet does not affect a normal surge. I am also concerned I may see some other issues once a higher rated PTC allows more current in ;-)
 

bonedoc

Dec 21, 2011
122
Joined
Dec 21, 2011
Messages
122
Ok, my worries were correct. When I went to a bigger PTC, things melted. So, here is my final questions:

-I am assuming that the new PTC is allowing more current in (and has lower resistance). I also assume that when I had success with the new frequency/duty on the new battery, it was optimized for the amount of current the old PTC was letting it. Do you think:

1- That the ON time just needs changed again with the new PTC mounted?

2- Should I try to optimize this circuit WITHOUT a PTC so that the duty cycle is optimal with unlimited current availability, and THEN add the PTC? I was just thinking I am going backwards a little because I set the ON time when the current was being limited too much.

You opinion is appreciated ;-)
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I recommend your circuit sense the current through the transformer and truncate the pulse at the point it hits a certain value.

Then all you need to do is determine the appropriate maximum current and the circuit will automatically adjust to the input voltage.

The problem with the PTC resistor is that it reacts to current integrated over time, so very brief high current pulses are not going to affect it the same way as a continuous high current. It may not provide the protection you desire.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Here is an example of a device which does all of this. Unfortunately, most of these devices are designed for rectified mains input. (90-250V) so they may not be directly applicable to your situation.

What you're looking for is a cycle by cycle current limit, because this will ensure that each pulse delivers exactly enough energy to "fill" the inductor to capacity.

You may be able to do a similar thing with discrete components. If you show us your circuit, we may be able to help.
 

bonedoc

Dec 21, 2011
122
Joined
Dec 21, 2011
Messages
122
Thanks for the link. I understand exactly what you mean in monitoring the current, and I do want to implement this in the next version of the pcb. I didnt think it was necessary at first, but as I push the limits, it seems more necessary. I will see if I can implement this.

I have the 10A Operating Current PTC on the board. As I was saying, things melted. So, I built a new PCB and lowered the duty cycle drastically. It does charge without overheating at the low duty....but it is very slow. i gradually raised the duty one bit at a time, and I got to charge to 300V in 6 seconds. After that, any higher duty cycle overheated it. That is waaay down from the 1 second charge I was getting from the small PTC.

Is this idea wrong:

My previous circuit was charging to 300V in 1 second with a 1.5A PTC. After so many minutes of use, the PTC overheated while the mosfet was not.

So, I increased to a 10A PTC to allow the circuit to pull the current but not overheat the PTC. The PTC is pretty much fine now, but the mosfet is going hot now. I thought this was because the bigger PTC was lower in resistance and could allow more current to pass. Only remedy to prevent melt down is to have a slow 6 second charge to 300V.

The only change I have made is that I have a different mosfet:

http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00002071.pdf

I have another batch of what I was using coming tomorrow.

I guess I dont understand why I was charging so fast before and everything was great, and then having access to more current is causing a slower charge time. The only thing I can think of is that the above mosfet is not suitable or I am not giving enough time to empty the bucket. ;-(
 

bonedoc

Dec 21, 2011
122
Joined
Dec 21, 2011
Messages
122
It looks like everything is working for the most part. I noticed one strange thing and wanted to see if you had an explanation. When I set the duty cycle to 90% of the max on time, the mosfet overheats. If I start with a lower duty cycle and increase to this number, it works fine.

Does this make sense? What is strange is that if the duty at 90% of max on time and it gets hot for just a second, I can cut the power, turn it back on, and there are no issues. It is almost like the higher temperature is allowing for a higher on time.
 

bonedoc

Dec 21, 2011
122
Joined
Dec 21, 2011
Messages
122
Maybe this is anexplaination

"High peak switch currents during start-up may occur in
switching regulators. Since VOUT is far from its fi nal value,
the feedback loop is saturated and the regulator tries to
charge the output capacitor as quickly as possible, resulting
in large peak currents. A large surge current may cause
inductor saturation or power switch failure"
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Not if you are controlling the duty cycle. That would be for a converter that is free-running, siwtching the MOSFET off based on the voltage across the output capacitor.

Bob
 

bonedoc

Dec 21, 2011
122
Joined
Dec 21, 2011
Messages
122
I wanted to update everyone about my situation, in case anyone is following. Everything is working great now. My issues were:

-When I started making the converter, the mosfet and inductor did not have access to the optimal current. Because the on time and periods were set to be at the available lower current, things were burning up when the circuit had the proper current availability.

-When I calculated the period and on time, the real world values the microchip put out are not exact. When you consider the total errors, I was overheating.

-I have a safety feature that allows discharge of the circuit. I found that there was a split second where the converter was charging and the safety feature was discharging, which was causing some extra heat.

-At peak current, my inductor has an inductance of 9uH, but it is 10uH at 0A. I was using the 10uH figure when I started, and that did make some difference.

I tried to make a check list on making a DC/DC converter. It is for beginners:

1- Find the inductor max current. Also get the inductance at this max current. Calculate the max time your voltage can be pulsed to hit saturation with the formula

Seconds = Max Current x Inductance / Voltage

Make sure the voltage is the maximum you expect to use. If you unknowingly increase this, the time it takes to reach saturation will go down, and things will melt.

2. To be safe, double the above time to get the period. This will give an duty cycle of 50% at the max on time.

3. Select a mosfet that can handle the max current and voltage. Be aware that the mosfet may need a higher rating if there is a high voltage on the secondary side of the inductor and more windings on the secondary side. Any voltage/current induced from the secondary side to the primary side can cause damage.

4. After you pick your mosfet with the correct ratings, make sure that it can switch fast enough. Picking a mosfet with low gate capacitance and resistance will mean faster times.

5. Make sure that the voltage applied to the gate is enough to drive the mosfet at the desired current.

Anyway....I am new at this stuff. Feel free to add or correct me. I just wanted to help other people new to the area.

I am wanting to test my circuit in continuous mode. I suppose the limiting factor is if it is getting hotter than it can cool, so meltdown happens.
 

bonedoc

Dec 21, 2011
122
Joined
Dec 21, 2011
Messages
122
As far as continuous mode goes...

Is this performed by shortening the period? From what I understand, this is where we dont allow the current to fall to zero. The only way to achieve this in my mind is to shorten the period so that the switching begins before the current on the primary side falls to zero. Is this correct?

What is the benefit of continuous mode? Will it allow a faster/larger output on the secondary side? On one hand I can see that it is faster because the switch does not need to fall back to zero. On the other hand, it seems as though it would not be as effective because the purpose of switching it to create a "AC" current and it seems to me that the transitions would not be as large in continuous mode.

Thanks for the tutoring and being patient!
 
Top