Once again some simple gdt/h-bridge questions/verifications
Thomas, Mon Sept 18 2006, 10:33PMAll right I'm beginning to understand that driving a chip at 60°C isn't that bad. I'm honsetly new to power electronics, though I have 2 sstcs and one drsstc. However, I'm just asking for some basic questions that need opinoins on (just to give me some security in my designs). I mean, they work, nothing blows. Well...except for my flyback gdt :(, I making the toroids after homework tonight.
H-Bridge Questions:
1. To check how my gdt is working; I check the resistance of my gdt secondary. So in order to do so I connect my ohm probes to the gate and the other to the source. I should see a little less then 10ohms right, which is actually the Rgate (the resistor before the gate or in series).
2. I noticed that when I run my gdt *without powering up my bridge my MOSFETs get a little warm (like 30°C) nothing that you can't touch, just a tad warm. This should be fine right?
3. Should I use a switching diode like the 1N4148 in reverse bias with Rgate? I do do this, i don't see much of a difference in the waveform, but I would like to think it's increasing my gate shut off time.
4. EVR why do you series two rectifying diodes? On you SSTC1 and SSTC2 (oh yeah an audio mod) H-Bridges? Is there a reason other then the addition ≈.7V drop? Unless your using it like some sort off double bufferish thing? But shoudn't they be paralleled? Idk, I'm only on my second year of college.
GDT question:
1. How hot is hot for the chip drivers to be running? I can get mine on my SSTC to run at 50°C this is good right? (before ferrite 77 some blew) (also i think my TC442X drivers are dying see q3) I also heatsink my mosfet drivers (temp measurement without heatsink)
2. I seen some several pcb boards of sstcs and howcome no one has think traces in they're drivers. Isn't it ideal for signals and stray inductance? Is think b/c they tend to work as antennas? What about huge ground plains? Would this act as a ground loop of some sort or shield my board?
3. Oddity, if my TC4422/TC4421 are running without a load they heat up they're probably dying?
Re: Once again some simple gdt/h-bridge questions/verifications
EDY19, Tue Sept 19 2006, 12:10AM
The heating of most of the components you describe depends a lot on the frequency you are driving them at causing switching losses etc. What is the frequency you are driving them at?
EDY19, Tue Sept 19 2006, 12:10AM
The heating of most of the components you describe depends a lot on the frequency you are driving them at causing switching losses etc. What is the frequency you are driving them at?
Re: Once again some simple gdt/h-bridge questions/verifications
Thomas, Tue Sept 19 2006, 01:18AM
When I ran my driver and the FETs heated up that was around 200kHz. My flyback driver problems I'm not in concern with, since I have to make the right transformer. That runs around 10kHz - 30kHz. My flyback tends to be best at 25kHz. Also for my coils the waveform is clean all around, with load and without. It just runs hot. I'm not sure why the unloaded fets heat up though. Also why my TC442X heat up on my board unloaded (with out gdt connected). Oh to add my UCC3732X don't heat up unloaded. Hmm, maybe I should try some new TCs? I think mine might be damaged since they did get hot when I tested them all :( oh my flyback driver issue, before realizing my gdt sucks.
Add: yeah for some reason my TC's are heating up. I just stuck some UCC's into my GDT and they run fine, I guess my TC's are no good. I'm gonna replace them and I'lll get the next model above the EPA package. It's V something and its rated at I think 125°C.
Thomas, Tue Sept 19 2006, 01:18AM
When I ran my driver and the FETs heated up that was around 200kHz. My flyback driver problems I'm not in concern with, since I have to make the right transformer. That runs around 10kHz - 30kHz. My flyback tends to be best at 25kHz. Also for my coils the waveform is clean all around, with load and without. It just runs hot. I'm not sure why the unloaded fets heat up though. Also why my TC442X heat up on my board unloaded (with out gdt connected). Oh to add my UCC3732X don't heat up unloaded. Hmm, maybe I should try some new TCs? I think mine might be damaged since they did get hot when I tested them all :( oh my flyback driver issue, before realizing my gdt sucks.
Add: yeah for some reason my TC's are heating up. I just stuck some UCC's into my GDT and they run fine, I guess my TC's are no good. I'm gonna replace them and I'lll get the next model above the EPA package. It's V something and its rated at I think 125°C.
Re: Once again some simple gdt/h-bridge questions/verifications
Steve Conner, Tue Sept 19 2006, 10:39AM
attempts to answer H-bridge questions.....
1. This partly shows that you wired it up right (but not fully as you could still have some of the secondaries phased wrong)
2. The on-chip connection to the MOSFET gate structure is made of crappy polysilicon with a high resistance. So the gate drive power can heat it up. I've heard of people experimenting with real high-frequency Class-E drive (like 10MHz) who have managed to destroy the MOSFET with just gate losses.
3. I've never used these diodes in SSTCs myself. But when designing a small H-bridge converter at work, I found that using the diodes lowered idle current and improved efficiency. My converter didn't use a GDT with bipolar drive though. I think the bipolar drive is what causes the diodes not to make much difference.
gdt answers...
1. If you can stand to hold your finger on the chip, it'll probably live.
2. I used thick traces where needed, and a groundplane, on my Mk2 DRSSTC drivers.
3. Depending on your GDT design, it may pull enough magnetizing current to make the driver chips heat. In a situation like this, adding those Schottky clamp diodes to the output may help the chips run cooler.
Steve Conner, Tue Sept 19 2006, 10:39AM
attempts to answer H-bridge questions.....
1. This partly shows that you wired it up right (but not fully as you could still have some of the secondaries phased wrong)
2. The on-chip connection to the MOSFET gate structure is made of crappy polysilicon with a high resistance. So the gate drive power can heat it up. I've heard of people experimenting with real high-frequency Class-E drive (like 10MHz) who have managed to destroy the MOSFET with just gate losses.
3. I've never used these diodes in SSTCs myself. But when designing a small H-bridge converter at work, I found that using the diodes lowered idle current and improved efficiency. My converter didn't use a GDT with bipolar drive though. I think the bipolar drive is what causes the diodes not to make much difference.
gdt answers...
1. If you can stand to hold your finger on the chip, it'll probably live.
2. I used thick traces where needed, and a groundplane, on my Mk2 DRSSTC drivers.
3. Depending on your GDT design, it may pull enough magnetizing current to make the driver chips heat. In a situation like this, adding those Schottky clamp diodes to the output may help the chips run cooler.
Re: Once again some simple gdt/h-bridge questions/verifications
Thomas, Tue Sept 19 2006, 02:08PM
oh i know they're wired correctly. This bridge has run off mains 120Vac. Plus after a one minutes of run time my MOSFETs don't heat up. I just think it's odd that the GDT heats them up, even when there is no power on the h-bridge or load. I'm using the IRFP260N. Oh also just to add i have schottky clamp diodes on my gdt output to gate transformer.
Thanks though. Also the IRFP260's do have a lot of gate capacitance.
Thomas, Tue Sept 19 2006, 02:08PM
oh i know they're wired correctly. This bridge has run off mains 120Vac. Plus after a one minutes of run time my MOSFETs don't heat up. I just think it's odd that the GDT heats them up, even when there is no power on the h-bridge or load. I'm using the IRFP260N. Oh also just to add i have schottky clamp diodes on my gdt output to gate transformer.
Thanks though. Also the IRFP260's do have a lot of gate capacitance.
Print this page