Nice work on the things, Sig.
I have a Fuji 6MBI 8L-120 here from some kind of old motor controller. What do you think
about using one or two of the half bridges out of these (or maybe paralleling all 3)? They're
only rated 8A, but it is a big module, so good thermal properties I'm guessing.
Yes it would work fine, but you'd (or I'd, haha) need to modify the code running on the ATTiny85 to have a far lower pulse width cap. It is easily done, but finding the longest safe operating pulse width involves deep math. If you knew the max rated PW that oneTesla recommends it would be easy to implement.
I should be getting a batch of ATTiny85's in the coming weeks, and I'm already expecting my Arduino UNO, when I have them both and some spare time I can offer the service of installing my code to the ATTiny85 and selling the ready to use MCUs. Then you just have to hook up a LM7805, two 10k pots, a switch, and your choice of fiber optic transmitter (I like the FB123/FB129 pair).
Nice work on the things, Sig.
I have a Fuji 6MBI 8L-120 here from some kind of old motor controller. What do you think
about using one or two of the half bridges out of these (or maybe paralleling all 3)? They're
only rated 8A, but it is a big module, so good thermal properties I'm guessing.
I wouldn't parallel them, Steve Ward has done extensive (pioneering, really) work on paralleling half/full bridges and there really is an art to balancing all the impedances. It is bulky, expensive, and a nightmare to debug. Leave it to the experts, imo. You also have to parallel gate drives when you parallel gates!
If their switching characteristics meet the requirements for your chosen secondary resonance frequency then by all means use them. Stick to a single half or full bridge though.
If you can pull up their datasheets I can extract an approximate max frequency for you.
450KHz, 300W, both ISSTC and CWSSTC modes shown. 3-4.5" streamers. Overall dimensions of the resonator are 14" x 6". If I had wood working tools, skill, and supplies I'd build a base for it. The entire assembly fits nicely on even a small desk. Unlimited runtime with no active cooling too.
I could look it up. The onetesla doesn't go that low. Its still playing an audible note at its lowest setting. I like seeing the high pulse width and the frequency in the single hertz. Or the single button fire option I have seen on the EVR one. Sadly I have no idea how to code. Its all Chinese to me.
You can run a DRSSTC at any interrupter frequency you like, it is the pulse width that has to be kept below a certain value. Too many RF cycles in a row (too long of an ON pulse) and your bridge explodes.
Yeah, The lowest frequency on the OT interupter is still an audible note though. I can look up the max PW that the interupter uses and use that but I wouldnt know where or even how to change the code for that.
I can change the code for that no problem. I can't verify pulse length below about 60Hz though, I don't have a digital storage scope, so discontinuous and non repetitive effects are beyond my analog scope.
So I can program it in with headroom but can't guarantee it.
I'd need to know the longest pulse length allowable in milliseconds or microseconds, and the average duty cycle range (in percentage).
That's good enough then, I can work out duty cycles from that. What "it's based off of frequency" translates to is "I used a shitty interrupter design".
Now then, I don't know if the ATTiny85 has enough resolution to work with such short pulses. There are clock issues with the ATTiny. When I get my UNO and 85's in I'll let you know. Write down the 150uS mark, cause I'll forget.
Haha, I am pretty choosy when it comes to FETs and IGBTs. I'd rather buy good ones that last than go through a pile of cheapies. These FDL100N50Fs have been really good to me.
You know, a 60Hz sine inverter has always been something I'd love to do, but I'd really want to do it analog, without a sinusoidal-PWM generator, and I just don't have the knowledge to do that yet. Plus I'd want it to be good for like a kilowatt...
Haha, I am pretty choosy when it comes to FETs and IGBTs. I'd rather buy good ones that last than go through a pile of cheapies. These FDL100N50Fs have been really good to me.
You know, a 60Hz sine inverter has always been something I'd love to do, but I'd really want to do it analog, without a sinusoidal-PWM generator, and I just don't have the knowledge to do that yet. Plus I'd want it to be good for like a kilowatt...
Ah, yep! I love those and I got a good handful of those FETs from when Fairchild did samples.
A 60 hz sine wave is actually *relatively* simple to do analog style.
You'd need a sine wave generator (a function generator IC, I'm planning on using the XR2206 myself), and then after some sort of Class D amplifier with current/voltage feedback. Obviously that's a huge oversimplification but that's how I'd do it.
You know, a 60Hz sine inverter has always been something I'd love to do, but I'd really want
to do it analog, without a sinusoidal-PWM generator, and I just don't have the knowledge to
do that yet. Plus I'd want it to be good for like a kilowatt...
What do you mean by analog? Linear? That has been done, but it takes a bunch of high
power transistors. It makes a really good electric heater.
I had an idea for a regulated true sine wave inverter a while back, but it was SMPS. There is
a flow chart of the design up on my site. It never did get built, though since modified sine
wave inverters are so cheap and work so well with inductive loads. Harbor Freight sells them
so cheap that it would be difficult to get into the market, even if it did work a tiny bit better.
Those are some pretty wimpy IGBT modules. They're only rated for 8amps and have a breakdown voltage of 120V. Oddly enough down farther in the datasheet they quote 1200Vce and 600Vce, but the maximum ratings clearly state 120Vces. I'm guessing there is a misprint and it is really 1200Vces. In addition, they are slow as hell. Maximum switching speed would be 105KHz, and that is for a 50% duty cycle, if you wanted to run them at a shorter duty cycle you'd have to lower the frequency.
To get any use of them in a TC setting you'd have to have a VERY low frequency. Assuming it really is 1200Vces, and you ran it off of doubled 240V, you'd have ~680V DC bus, running at close to 7A for a safety margin you'd get 4.75kW PEP, but unless you had some insane capacitance (I'm not going to run the figures now) your average power would be much lower. It would be possible, but not economical really. The extra cost of making a secondary that large would offset any gain from them being free.
edit: oh and by analog I mean analog; no digital MCU or such. The problem I see with going the Class-D route is you'll need a beefy filter network to remove the carrier wave, and then there is the issue that with class-D the load determines the filter network parameters, and a power inverter has to cope with dynamic loads.
