Problems getting ZVS driver to work / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Problems getting ZVS driver to work

1 2 3 4 next last

Move Thread

LAN_403

Heisenberg

Mon Jan 11 2016, 12:51AM

Registered Member #58279 Joined: Sat Jan 09 2016, 05:39AM
Location: Houston, TX
Posts: 13

Hello guys, this is my first post as I'm a newly registered member but have been a lurker for some time. I'm relatively new to the electronics hobby and recently ordered parts from Digi-Key and built my own ZVS driver. I soldered it all together and when I power it up, nothing happens. I triple checked all the solder joints visually and used a multimeter to probe all over the board and everything is wired correctly and there is voltage flowing throughout the circuit as it should be, but nothing is happening when I connect it to a flyback. I have a ZVS driver that I purchased pre-built from Amazon that is the same design and almost the same components as the ones I'm using, so I know the flyback is good because it works with that ZVS driver like a champ, but I am unable to get mine working. The only thing weird that I noticed was when testing the voltages on the negative sides of the UF4007 diodes, one of them reads 12 volts...same as the input, the other one shows a lower value that steadily decreases as I keep the probes on it. I'm not sure if this is normal or not. Below is the schematic I used to build the circuit and the list of parts from DigiKey. I was hoping that someone may spot something I may have missed.

Any advise on troubleshooting, etc, would be awesome. If need be, I can upload a couple of pics of the actual circuit I soldered. Its ugly as heck and I'm prepared for the ridicule I will receive when I show it to you provided someone can help point me in the right direction.

DigiKey Parts Used
2x UF4007CT-ND - DIODE GEN PURP 1KV 1A DO41
2x A105901CT-ND - RES 470 OHM 3W 5% AXIAL
2x IRFP260MPBF-ND - MOSFET N-CH 200V 50A TO-247AC
2x 1% 10k OHM RESISTORS 1/4W
1x 732-7166-ND - FIXED IND 150UH 10A 21 MOHM
2x BZX85C18_T50R-ND - DIODE ZENER 18V 1W DO41
2x BC2592-ND - CAP FILM 1UF 20% 630VDC RADIAL (Series MKP339 X2)

haxor5354

Mon Jan 11 2016, 04:16PM

Registered Member #2063 Joined: Sat Apr 04 2009, 03:16PM
Location: Toronto
Posts: 352

check your diode polarities perhaps?

jdub1581hv

Mon Jan 11 2016, 07:46PM

Registered Member #55219 Joined: Tue Jun 09 2015, 11:21PM
Location:
Posts: 80

double check the zeners and UF's one may have blown or got too hot during solder..

jpsmith123

Tue Jan 12 2016, 01:28AM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

I once had a ZVS driver that would only start up if the DC power was abruptly applied to it.

If the power supply was connected to the circuit first and then the power supply was turned on, it wouldn't work, but if the power supply was turned on first and then the output was connected/switched to the ZVS driver, it would work.

Heisenberg

Tue Jan 12 2016, 03:04AM

Registered Member #58279 Joined: Sat Jan 09 2016, 05:39AM
Location: Houston, TX
Posts: 13

I did some more troubleshooting and I discovered that something unusual happens when I plug in a 24v power supply to the input and have the flyback connected, the power supply LED comes on and promptly off because of the short/overload protection I'm assuming. When I disconnect the center tap from the three-pin connector, the power supply turns on, but nothing happens with the coil. I have a meaty 150uH torroid going from positive to the center tap so I'm not sure what the problem could be.

Here's a couple of pics I snapped of the top and bottom of the board. Please be gentle, I know its grizzly. lol

jdub1581hv

Tue Jan 12 2016, 04:13AM

Registered Member #55219 Joined: Tue Jun 09 2015, 11:21PM
Location:
Posts: 80

Have you tried using a car battery for supply? Sounds like the one you're using doesn't have the amperage to run..
Most recommend 10a supply...

Can't see behind the caps, but are the 2 black "Drain" wires connected to the same side of the Caps? If so That's the problem.

Heisenberg

Tue Jan 12 2016, 04:23AM

Registered Member #58279 Joined: Sat Jan 09 2016, 05:39AM
Location: Houston, TX
Posts: 13

The power supply is definitely powerful enough...24v 15a. The black drain wires are going to opposite sides of the caps which are wired in parallel. The third black wire is going from the left side of the caps one of the screw terminals and the opposite side of the terminals is connected directly to the other side of the caps without wire.

edit:
So I discovered a few new observations. I started by desoldering and removing the inductor and tried using a smaller one with a yellow torroid ring, about 50-80 uH or so and the power supply didn't do a short shutdown like previously, but still, nothing happened. I removed one of the 1uF caps and left the other, same thing. I tried using two inductors in series and it didn't like that because a couple of the alligator clips I had attached to them started melting so I quickly turned it off. It was the alligator clip connected to the center tap of the flyback. So I removed the 1uF cap and replaced it with a 680 nF cap, power supply powers on without problem but still nothing happens. So the new observations I made is that the size of the inductor seems to have an effect on whether or not the PSU powers on and stays on or it shuts down immediately due to a short or overload.

Based on that info, any other suggestions? What are pros/cons of using more or less caps?

Kiwihvguy

Tue Jan 12 2016, 10:03AM

Registered Member #3395 Joined: Thu Nov 04 2010, 08:42AM
Location: Christchurch, New Zealand
Posts: 193

I can see that you've soldered the pins of the fets on the same breadboard traces, however from the picture you uploaded it appears the gates seem to be directly connected to the sources of the other mosfet - this should not occur. The gate of one mosfet should connect to the source of the same mosfet *only* via the 10k resistor and zener diode. Can you confirm whether or not there is an electrical seperation that may not be clearly visible in the picture?

Your UF diodes appear to go from the gate of each mosfet to the opposite mosfets' drain, which is good and I also see you have made a physical gap in the centre breadboard trace on which the mosfets drains are soldered to.

Do remember that the mosfet's metal thermal back pad is also an electrical connection to the drain of the mosfet. It appears the mosfets have no thermal insulating pad, so in this case the entire heatsink and its mounting pins are effectively connected to the drain. Looking at the bottom of the veroboard, the heatsinks' pins are soldered onto the same traces and on one of the mosfets' heatsink, the drain is also soldered onto the negative rail with no apparent electrical separation - again, this should not occur. Only the sources of the mosfets should be connected to the negative rail.

Ensure that all your solder connections do not overlap onto the copper traces adjacent to the ones they are soldered to.

I personally would recommend that you mount the capacitors separately closer to the connections to the flyback's primary so there is less chance of the mistakenly connecting it incorrectly to the rest of the circuit.. It's also a good idea to use an MMC (multi mini capacitor) instead of a couple of capacitors because the capacitors will see higher than the input voltage and higher currents due to the intrinsic resonant nature of the ZVS circuit. A capacitor or MMC with a lower capacitance will cause the operating frequency to increase, and a higher capacitance will cause the operating frequency to decrease.

Using more capacitors in series and parallel (to increase voltage rating and current capability) is a good idea as they will get heavily abused at higher powers. For my 1st ZVS I used a single, tiny MKP capacitor. It got very hot and cracked open!!

Try rectifying the above suggestions by cleaning up the solder joints, creating electrical separations by cutting the copper traces and using thermal insulating pads as well as thermal grease (messy, but improves heat dissipation - colder fets are happier fets).

Albeit exhaustive in detail, I hope this is of help to you if you take the time to read through and put some thought to it. Well done on making it this far - I know very well the frustrations of a circuit that doesn't work, but then it feels amazing when it does work! Let's hear back from you soon

jpsmith123

Tue Jan 12 2016, 01:55PM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Why don't you try what I impliedly suggested; i.e., turn the power supply on first, and then connect your circuit to it by closing a switch or by manually connecting a clip lead or something.

Heisenberg

Tue Jan 12 2016, 06:22PM

Registered Member #58279 Joined: Sat Jan 09 2016, 05:39AM
Location: Houston, TX
Posts: 13

jpsmith123: I did try what you suggested - no go. PSU shuts down as soon as I connect the lead when the flyback is attached.

Kiwihvguy: Awesome writeup, thank you for the time to do so. I also noticed after reading your post the rows with the fet gates not being severed, I thought for sure this would fix the problem so I went ahead and severed the gates of both fets so they wouldn't connect to the source of the other one, still no dice. Here's a pic of the solder traces after making the adjustments you suggested.

Also, the pictures doesn't show it, but both fets have a mica insulator between the drain and heatsinks as well as thermal paste. When I check for continuity between the drains and the sinks, there is none.

I also made sure none of the rows are bridged accidentally by doing continuity checks up and down the board, everything seems to be in order, yet when I have the flyback connected and flip the switch on the PSU, the fan starts and promptly stops immediately as though it was connected to a short. When I remove the flyback, the PSU fires up and stays on.I'm stumped, but I think we're heading in the right direction.

If this leads to a dead end, I'll be desoldering everything and starting over on a fresh board, but I'm hoping it was something simple that I overlooked such as the gates being bridged which I didn't notice due to project tunnel vision, this is why more sets of eyes help a lot.

Another weird thing is that when I recreated the same circuit on a breadboard using IRFP250s instead of 260s, same thing happens. I can't get it to work on a breadboad either so I'm thinking I overlooked something. Either that or the breadboard jumper wires aren't thick enough to handle the current demands.

1 2 3 4 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob