Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

14 to 40 or 50 volt boost circuit

Status
Not open for further replies.
hoooooo ray, my computer came back, I took it apart, the only think I found was one terminal on the RT battery hold was loose, unsoldered, so I fixed that and cleaned it all back, mopped the spilled drinks out of it and blew out the lint, and it lives again. Now to see if it all works correctly.
Kinarfi
 
yea, I'll say, what should I do with the extra screws?
 
update, looks like I destroyed the 1 to 4 usb hub and the DSO2150 pc oscilloscope ($60 repair)
 
I had a Lambda power supply with an ungrounded output powering the circuit, the computer was running on the AC power supply and I'm not sure if it was grounded or not, I think Yes, and the tip of the solder iron is definitely grounded and the PC oscilloscope had it's probes attached to the circuit board, can't remember what I was monitoring for certain, probably the the gate and drain of the FET. The cap was almost certainly above 100 volts and maybe about 200 volts.

As an update to the circuit that Ronv gave me, I built it and then modified it and built it again, and as an answer to another posted question about what best in electronic, I was absolutely ecstatic when I power the circuit up and it worked perfectly, that's a wonderful feeling!!!
I have a 66 LED string of 11 LEDs in 6 parallel set that were the back light out of a broken laptop screen, so I thought I'd just power them up and use them for a dome light in my buggy, turns out they needed around 30 volts and I decided that 1 ma per string was enough so I made a CRD to limit the current using an N-channel JFET, J310, to limit the current to about 6 ma.
On Ron's circuit, I analysed the transistor circuit below the FET keeps the current down for start up and then once the output cap is charged and the current drops, that transistor circuit basically falls out of the circuit. The transistor circuit below the load keeps the output voltage down, in this case, It watches the voltage of the current limiter and keeps it from getting too high which keeps the power losses down and keeps the voltage low enough to stay within the specs of the current limiting device.
Is that about correct????
Any way, It works awsomely, the .asc has measured information on it.
Thanks for the help, hope others can use this information.
Kinarfi
 

Attachments

  • Current Boost 6 ma.asc
    5.1 KB · Views: 89
  • post.jpg
    post.jpg
    655.3 KB · Views: 131
I was trying to design a higher wattage application for this design, 80 watts and 80 volts, but it would need a 5uH inductor and it failed to generate enough voltage, so I went back to the 40 watt 40 volt design similar to the one use the 5 lead IC mentioned earlier. Since I am working with current control, I monitor the current regulator voltage rather than the output voltage and it was suggested that the feed back resistor voltage divider could be replaced with a zener diode, the voltage of the zener would depend on how much head space the current control unit needs. How ever, Spice showed a sustained oscillation when using a zener where as the voltage divider seemed to divide the oscillation signal and the zener would not.
Here is another version of this design for a 40 watt 40 volt 1 amp supply for my 4 10 watt 10 volt LED that I will be building today and tomorrow for my buggy light bar.
**broken link removed** **broken link removed**

PS. The light bar uses 12 10watt LEDs and I need 3 of these circuits to power it, 4 lights per circuit. The 4 lights in the center are on a separate switch and the 8 lights on the ends are on another switch and the whole bunch are on a single switch.
 

Attachments

  • 40 Watt 40 Volt.asc
    5.2 KB · Views: 105
Last edited:
Here's something quirky, I had the circuit running again and I noticed the .2 ohm resistor getting warm and started checking voltages and found it was generating 70+ volts, 33 across the load and 40 across the current limiter, but with the current limiter working good, the lights survived, and if I turned the supply down 14 to about 5 or 6, the voltage remained high until the supply got low, then the output went to 36 and stayed there, even when I turned the supply up to 20, and if I disconnected the power and then reapplied it, the voltage went back up to 70+. I also notice the frequency was around 2 - 3 KHz when the voltage was high and 109 KHz when the voltage was 33.
Any one have any ideas on how to stabilize this??
Kinarfi
 

Attachments

  • Trooper Dome Light.asc
    5.1 KB · Views: 100
  • Trooper Dome Light.png
    Trooper Dome Light.png
    35.6 KB · Views: 115
I'm not sure how J1 and R5 work. With 220 ohms in series with the LEDs there can't be much current. Not sure an nfet can carry the load, but I don't know the one you used. Having said that it seems like R7, R8 are in the wrong place to make it either current limit or voltage adjust. Which did you intend? The only other thing I see is that R6 is .2 ohms this would reset the 555 with about 3 amps. This may not be enough to get the cap to charge so it might want to be more like .1 or .15 ohms. This might explain why it won't start but will run. I suspect you have a different current limit scheme than the one shown?
Is the diode a schottky type?
 
Last edited:
The J310 & 220 resistor make a current limiting device, the J310 model was saved from an older installation of spice. Run the attached .asc , the J310 model is very accurate and limits the current to 10 ma, so by making the 555 run in voltage mode via R7 & R8 to keep the voltage across the current limiter down, but allows it to be high enough to let it do it's limiting job, the 555 regulates at what ever voltage is needed to power the LED string at the preset current plus some overhead needed by the current limiter, just like putting a resistor in the output leg of an LM317 to make it a current limit or current source.
I simulated several values of R6 and as R6 got lower, the time from start up to where the voltage started rising got less and the current got higher, .2 seemed to work well and I had some on the shelf. I tried several different approaches to the run away voltage and the best results was to apply power to the inductor directly and apply power the 555 via a 10 ohm resistor with a 10uf cap in parallel with the 555, that brought the tolerated voltage up to about 11 where it was 6. Another idea I'm playing with, but may not need, is to replace Q1 with a 4N35, see attached .asc
I'm starting over though, I messed around with the circuit board and eventually let the smoke out, should have a delivery of a bunch of 555s and other items tomorrow, I'm also working on a 40 volt 50 watt set up too.
Kinarfi
Sure miss my oscilloscope!
 

Attachments

  • Draft1.asc
    574 bytes · Views: 87
  • Trooper Dome Light -- optos.asc
    5.5 KB · Views: 102
Oh, I see the dilima. 10 ma. I thought it was the 1 amp circuit. The circuit values are for 1 amp so all kinds of bad things happen with the very low load. The inductor value and frequency are set to provide 1 amp at about 75 Khz. So with only 10 ma to voltage will go way high. To try to compensate and keep the voltage down the circuit will increase the frequency. But it can only go so far. In the simulation I think it was over 200Khz. So it either wants a much larger inductor (so the current doesn't rise as fast) Or a very high frequency (so the current doesn't have time to go very high) and a short on time.
Having said all that I think what is happening in your situation where it won't start is that the inductor is saturating on power on. The switch gets flipped and it has to charge the cap thru the inductor to get started. The power on reset is to prevent the 555 from starting until after the cap has charged. That is the big current spike you see if you start the simulation at time = 0. So if the timer starts while the inductor is saturated the inductor just looks kind of like a small resistor and the current sense shuts off the timer (reset pin)and it tries to start again. It probably stays in this loop. You can prevent this by making power on reset longer (say R4 47k and R3 6.8K)This is similar to what you are doing delaying the power to the 555.
Hope this helps.
But bottom line is it isn't set up for the low current. We could probably make a real simple one for 10ma.
 
That definitely helps, I was thinking of increasing C4, but I didn't realize that the purpose of it was to shut down the timer until the C2 cap was charged, another possibility would be to use a smaller C2. I will take this all into account as I rebuild the driver for the Dome Light. I'm presently working on a driver for 30 volt @ 1 amp. Can you suggest a good source for information that would relate to this project, specifically, Inductor calculations and 555 usage.
Some suggestions for the 10 ma 36 volt supply would be appreciated.
Thanks
Kinarfi
 
Last edited:
**broken link removed**

I got it running, 40+ volts out, current limited to 965ma, Lights glow bright white, Perfect. I wasn't getting enough current and thought I would have to modify the voltage divider for the LM317 feed back, but I accidentally modified R4, the C4-Reset voltage divider and got the increased current I needed.
On the scope screen shots, Green trace is the gate and yellow is L1-Drain. it's been educational and fun.
Kinarfi
 
Last edited:
Smps

I don't know if there are others out there who are as interested in voltage boost circuits as I am, at the moment, but being as I posted some circuits, as I find improvements to my circuits, I feel obliged to post them also, hope no body disagrees.
I've been working on powering up 4 10 v 10 watt LEDs in series with a current limiting supply and this is my best so far using easy to find components, namely, the 555 timer. Earlier version pulsed several times then 'rested', this one has an even continuous pulse width and appears to stabilize on my scope in just a few seconds as it charges up at power on.
Out put to the LEDs is 40 V and 956 ma, the input voltage can be varied from 11 - 15 volts and I forgot to get a good current reading, but it appeared to be about 6A at 11V and under 4A at 15V on the inaccurate meter on the power supply.
 
Another variety

I came up with another variety of current control for boosting voltage, I needed to power up 2 10 watt LEDs, 12.5 volt, 800 ma, from 12.6 to 14.2 automotive power which didn't lend it self to the use of LM317s to control the current, the overhead was to high, so I boosted the voltage and ran the LEDs in series and controlled current via feed back with a pot. As the temperature rose, the current decreased some, but not enough to create problems, Yet. I have also become a fan of length wise bussed circuit boards.
**broken link removed**
Kinarfi
 
Here's something quirky, I had the circuit running again and I noticed the .2 ohm resistor getting warm and started checking voltages and found it was generating 70+ volts, 33 across the load and 40 across the current limiter, but with the current limiter working good, the lights survived, and if I turned the supply down 14 to about 5 or 6, the voltage remained high until the supply got low, then the output went to 36 and stayed there, even when I turned the supply up to 20, and if I disconnected the power and then reapplied it, the voltage went back up to 70+. I also notice the frequency was around 2 - 3 KHz when the voltage was high and 109 KHz when the voltage was 33.
Any one have any ideas on how to stabilize this??
Kinarfi

well for starters there is no way a 60mA FET can handle a 1A diode string...
 
Turned out nice Jeff. Really small, must have been a challenge.

Ron V.
 
Here's something quirky, I had the circuit running again and I noticed the .2 ohm resistor getting warm and started checking voltages and found it was generating 70+ volts, 33 across the load and 40 across the current limiter, but with the current limiter working good, the lights survived, and if I turned the supply down 14 to about 5 or 6, the voltage remained high until the supply got low, then the output went to 36 and stayed there, even when I turned the supply up to 20, and if I disconnected the power and then reapplied it, the voltage went back up to 70+. I also notice the frequency was around 2 - 3 KHz when the voltage was high and 109 KHz when the voltage was 33.
Any one have any ideas on how to stabilize this??
Kinarfi

well for starters there is no way a 60mA FET can handle a 1A diode string...

I'm sorry you misunderstood the drawing, the string of LEDs represents the 66 string of LEDs I have from a broken laptop computer screen and are used to represent the voltage drop that I have, not the current, the current for this little part of my project of experimenting with current control for LEDs is 9.5ma and is regulated by the J310 FET and the 220 ohm resistor as explained in post #31, I built the circuit and it works very nicely. I have also made several other varieties using the 555 timer as the base component.
Kinarfi
 
I'm sorry you misunderstood the drawing, the string of LEDs represents the 66 string of LEDs I have from a broken laptop computer screen and are used to represent the voltage drop that I have, not the current, the current for this little part of my project of experimenting with current control for LEDs is 9.5ma and is regulated by the J310 FET and the 220 ohm resistor as explained in post #31, I built the circuit and it works very nicely. I have also made several other varieties using the 555 timer as the base component.
Kinarfi

my bad... I only looked at the FET and LED part numbers

a TPS61040 with a dual inductor ( output switch to the center tap) will easily drive 20mA into a 50V LED string if you put the current sense to the FB pin.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top