CT-Scan on the 19th. ...Jim Thompson
CT-Scan on the 19th. ...Jim Thompson
-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | STV, Queen Creek, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I'm looking for work... see my website. Thinking outside the box... producing elegant solutions.
Exactly. That's why I also never subscribed to the PAL/GAL craze and tried to dissuade other engineers from that. Many used them for mundane stuff such as address decoders. They did not realize that it increases cost by a factor of 3-5 or more, that those things guzzle power and that the programming causes a slew of documentation overhead and thus even more cost.
-- Regards, Joerg http://www.analogconsultants.com/
If you have spare inverters/NAND gates/etc does the "trick" you often see of using them in parallel to drive a FET gate actually work well or is that just old wive'ss tale "engineering voodoo"
That can work fine.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
Yep. But you have to weigh the cost of a hex inverters chip against two BJT and at least for Asian production the BJTs always win. Unless you had 4-5 inverter vacant anyhow for some reason, which is rare.
-- Regards, Joerg http://www.analogconsultants.com/
I like the 74xx151! IDK if this is common knowledge but it can be used to synthesize any 4 input Boolean function. Like a "pico-PAL" as it were...
Gate drive can be built economically with both: use one gate to drive a complementary emitter follower. Or use two gates, one to drive the gate as such (holds the gate with zero saturation voltage), the other to drive the emitter follower.
The emitter follower usually "drools" a bit and overshoots the theoretical
0.7V-below-rail saturation point, which is fantastic to begin with. Just in case, the extra gate provides a modest holding force to keep it near the rail.Coincidentally, just the other day, I breadboarded this with a 74HC14 and
2N4401/3 at 3.0V. Into a 4nF equivalent gate, I get 50ns edges, not bad.Even better with low-Vce(sat) type transistors, like PBSS303NX/PX. The high hFE will give excellent results even with CD4000 logic, or a somewhat cheaper, still highish hFE BCxxx something or other, since you won't be able to utilize the full current capability of those.
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: http://seventransistorlabs.com
Here's a 100 volt pulse into a 50 ohm load, transformer coupled. The fet gate drivers are TinyLogic triple buffers, 13 cents each.
One trick is to run them at abs max Vcc.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
Send teh codez send schem
... and not pump out too many pulses per second. Otherwise the plastic package might start to bubble and smell funky.
-- Regards, Joerg http://www.analogconsultants.com/
You can also just do a PNP/NPN follower and run a resistor of several tens of ohms from the bases to the emitters. Zetex (now Diodes Inc) makes good pairs and they cost around 20c inqties if that's in the budget:
Way I use them is to tie the emitters together and then to the FET gate to be driven. VCC to V+, VEE to GND, IN to the driving gate. Resistor from IN to the FET gate.
None of ze codez needed :-)
If not in the budget I use two BJT de la maison, whichever is least expensive. The topper in that respect was a company in South Korea. They had long computer print-outs hanging all along one of the hallways. Those contained the daily pricing of a lot of such parts, like a really long stock ticker. The carpet in front of that wall was very worn and there were lots of burn holes in it (engineers were allowed to smoke there and unfortunately most did).
-- Regards, Joerg http://www.analogconsultants.com/
A college professor showed us that in a logic class just to show it was possible. He seemed to think it would be rather inefficient to construct chips that way, lol! I guess not many back then looked far enough ahead to see the day when we would have more transistors than we know what to do with.
-- Rick C
Had the guy ever participated in an industrial chip design? Sometimes professors know very little about real life out there. Even during their often mandatory "industry work experience" phase they are often cocooned in some plush ivory tower where only researchers work.
The topper at my university was a professor who blurted out that we only have to learn all this board level analog stuff for the exam. Once we graduate this would all be obsolete and it would all be ICs. I had to suppress a major ROFL outburst but the majority of students believed such nonsense. Oh well, for me that meant plenty of work opportunities.
-- Regards, Joerg http://www.analogconsultants.com/
..
You need an inverter as well if you don't have true and complement of at least one of inputs.
kevin
I didn't think that was possible, because it's not. It needs another chip.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
Fingers crosssed.
--sp
-- Best regards, Spehro Pefhany
I have a positive outlook. But I could be wrong. ...Jim Thompson
-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | STV, Queen Creek, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I'm looking for work... see my website. Thinking outside the box... producing elegant solutions.
You either need one input inverted, or you can use a mux twice the width with only ones and zeros on the muxed lines. The latter is actually how FPGAs use muxes for logic. Even the ones and zeros are supplied by a small block of RAM so they are completely programmable.
-- Rick C
I agree but the OP said any function of 4 inputs with a 74xxx151 which only has 8 inputs plus the selects and enable.
kevin
Well, if you must...
That saves a gate, and works reasonably well, outside of logic-level use.
The downside is, the transistors only boost when the resistor voltage drop exceeds Vbe. So, a CD4000 gate would never get any assistance with that low of a resistor (>= 220 would be fine though). 74HC would get a boost after a couple of volts, but that wouldn't be of much use against a logic-level MOSFET that's already gone through the Miller plateau. So, 100-220 would be better there as well.
The resistor will also delay boost, and clear stored charge more quickly, so you probably won't get the overshoot bonus, either. :(
So, for logic-level use, you might be better off without the resistor at all...
I find it amusing (or slightly infuriating) that they find it fit to call it a "gate driver", when it's painfully obvious from the datasheet that it's just two discrete transistors.
I wonder why they did that. Marketing? Does no one understand how to use BJTs anymore? (Well, probably. :( ) They don't give any tips on how to generate that nice big voltage swing, either (tens of volts in a couple nanoseconds, you say? That's /fantastic/!..), which is the bigger problem when making a driver. (On a similar rant: TI is mum about what magical gate driver they used to test their NexFETs. "R_G = 0 ohm". Yeah, sure.)
In any case, the transistors themselves are great, whether you buy singles, duals (ZXTDxxx I think) or these so-called "gate drivers". :)
Tim
-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website: http://seventransistorlabs.com
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.