Gate drivers - transformer vs bootstrap
akkudakku, Mon Aug 29 2016, 11:06AMHey guys.
After looking over a bunch of schematics that deal with H-bridges I have found out that you exclusively use GDTs. Why? Isn't the simple bootstrap capacitor approach simpler AND cheaper?
Given that most interrupters are run over fiber (and such are isolated) why even bother with a GTD?
Attaching a simple and cheap high-side/low-side driver schematic. NOT TESTED, but should work. Any tips on it?
Re: Gate drivers - transformer vs bootstrap
Sulaiman, Mon Aug 29 2016, 11:27AM
you need twice the number of drivers for a H-bridge as your circuit is for a 1/2-bridge driver
for experimental/diagnostic work it is good to have complete electrical isolation between control and power circuits
a GDT naturally prevents a 'continuous on' condition for the power transistors due to potential driver 'mis-calculations'
for a solid-state solution there are high&low-side driver ICs commonly available that incorporate 'boot-strapping'
Sulaiman, Mon Aug 29 2016, 11:27AM
you need twice the number of drivers for a H-bridge as your circuit is for a 1/2-bridge driver
for experimental/diagnostic work it is good to have complete electrical isolation between control and power circuits
a GDT naturally prevents a 'continuous on' condition for the power transistors due to potential driver 'mis-calculations'
for a solid-state solution there are high&low-side driver ICs commonly available that incorporate 'boot-strapping'
Re: Gate drivers - transformer vs bootstrap
akkudakku, Mon Aug 29 2016, 11:43AM
Yes I know there are IC drivers
it's just that they are unavailable in hobby shops where I live. And ordering a bunch of IC's is not very economical.
Also yes what I shown is half-bridge, but it's simple to just make 2 of them.
I agree about the points you made about diagnostics. Less chances to nuke your project and measuring equipment.
Also fair point on the GDT defaulting in the off state.
But with direct drive you dispense with the hassle of making the transformer.
akkudakku, Mon Aug 29 2016, 11:43AM
Sulaiman wrote ...
you need twice the number of drivers for a H-bridge as your circuit is for a 1/2-bridge driver
for experimental/diagnostic work it is good to have complete electrical isolation between control and power circuits
a GDT naturally prevents a 'continuous on' condition for the power transistors due to potential driver 'mis-calculations'
for a solid-state solution there are high&low-side driver ICs commonly available that incorporate 'boot-strapping'
you need twice the number of drivers for a H-bridge as your circuit is for a 1/2-bridge driver
for experimental/diagnostic work it is good to have complete electrical isolation between control and power circuits
a GDT naturally prevents a 'continuous on' condition for the power transistors due to potential driver 'mis-calculations'
for a solid-state solution there are high&low-side driver ICs commonly available that incorporate 'boot-strapping'
Yes I know there are IC drivers
it's just that they are unavailable in hobby shops where I live. And ordering a bunch of IC's is not very economical.Also yes what I shown is half-bridge, but it's simple to just make 2 of them.
I agree about the points you made about diagnostics. Less chances to nuke your project and measuring equipment.
Also fair point on the GDT defaulting in the off state.
But with direct drive you dispense with the hassle of making the transformer.
Re: Gate drivers - transformer vs bootstrap
Mads Barnkob, Mon Aug 29 2016, 06:50PM
If you think winding 15 turns of Cat5 network cable on a ring core and twisting the wires together is a hassle... that is one of the easiest parts to make :)
There are almost no cons against GDT in hobby use, the only reason industry moved away from it, its large, its heavy, its expensive to make, requires more manual/human handling and IC companies does not sell transformers (a actual punch line in a high side driver commercial ).
High side drivers require more complex or tighter control, they are not noise immune as some 4hv members have experienced and moved back to GDT.
Mads Barnkob, Mon Aug 29 2016, 06:50PM
akkudakku wrote ...
Sulaiman wrote ...
you need twice the number of drivers for a H-bridge as your circuit is for a 1/2-bridge driver
for experimental/diagnostic work it is good to have complete electrical isolation between control and power circuits
a GDT naturally prevents a 'continuous on' condition for the power transistors due to potential driver 'mis-calculations'
for a solid-state solution there are high&low-side driver ICs commonly available that incorporate 'boot-strapping'
But with direct drive you dispense with the hassle of making the transformer.you need twice the number of drivers for a H-bridge as your circuit is for a 1/2-bridge driver
for experimental/diagnostic work it is good to have complete electrical isolation between control and power circuits
a GDT naturally prevents a 'continuous on' condition for the power transistors due to potential driver 'mis-calculations'
for a solid-state solution there are high&low-side driver ICs commonly available that incorporate 'boot-strapping'
If you think winding 15 turns of Cat5 network cable on a ring core and twisting the wires together is a hassle... that is one of the easiest parts to make :)
There are almost no cons against GDT in hobby use, the only reason industry moved away from it, its large, its heavy, its expensive to make, requires more manual/human handling and IC companies does not sell transformers (a actual punch line in a high side driver commercial ).
High side drivers require more complex or tighter control, they are not noise immune as some 4hv members have experienced and moved back to GDT.
Re: Gate drivers - transformer vs bootstrap
Mads Barnkob, Mon Sept 12 2016, 12:18PM
I came by the sales presentation that I mentioned and thought I would add it here for your entertainment :)
Mads Barnkob, Mon Sept 12 2016, 12:18PM
I came by the sales presentation that I mentioned and thought I would add it here for your entertainment :)
Re: Gate drivers - transformer vs bootstrap
Sulaiman, Mon Sept 12 2016, 01:44PM
Even though biased, I like the honesty, and use of an emoticon.
akkudakku
"I agree about the points you made about diagnostics. Less chances to nuke your project and measuring equipment."
actually I am more concerned about 'nuking' myself !
galvanic isolation is good.
Sulaiman, Mon Sept 12 2016, 01:44PM
Even though biased, I like the honesty, and use of an emoticon.
akkudakku
"I agree about the points you made about diagnostics. Less chances to nuke your project and measuring equipment."
actually I am more concerned about 'nuking' myself !
galvanic isolation is good.
Re: Gate drivers - transformer vs bootstrap
Hazmatt_(The Underdog), Tue Sept 13 2016, 01:37AM
How do you expect the design to work when the high side source is at Vcc/2?
I could see this working for a low voltage application, but not mains.
If you don't want to use GDT's, and I have thought about this in the past, each driver requires its own isolated power supply, plus and minus, and optical isolation. You want a negative supply to slam the gate off, not just to zero volts. Unfortunately opto's can be slow, which brings us right back to a small iso transformer... so either way....
Hazmatt_(The Underdog), Tue Sept 13 2016, 01:37AM
How do you expect the design to work when the high side source is at Vcc/2?
I could see this working for a low voltage application, but not mains.
If you don't want to use GDT's, and I have thought about this in the past, each driver requires its own isolated power supply, plus and minus, and optical isolation. You want a negative supply to slam the gate off, not just to zero volts. Unfortunately opto's can be slow, which brings us right back to a small iso transformer... so either way....
Re: Gate drivers - transformer vs bootstrap
Patrick, Wed Sept 14 2016, 04:40PM
I like the emoticon too, its such a brazen argument. We dont sell it, so dont use it, or screw off.
Patrick, Wed Sept 14 2016, 04:40PM
I like the emoticon too, its such a brazen argument. We dont sell it, so dont use it, or screw off.
Print this page