1. Home
  2. Electronics Design
  3. Bandgap Design

Bandgap Design

I've finally boxed myself into an arena that I've studiously avoided for the better part of forty years. Heretofore I've had the luxury of at least a 12 volt system that I could easily regulate down to 6 volts for a stable reference.

I've got to now make a stable voltage source for four NiMH batteries (4.8 at full charge, 4.0 at cutoff). As with all things, I started off with AOE/Win and a google on "bandgap". In most things I consult Win for, both the theory and practice are fairly well laid out. However, in this instance there is some handwaving going on that I do not understand. Google gives me a lot of very well intentioned university dissertations with partial derivatives that show that the bandgap will, in fact, work well -- what my first engineering supervisor called mental masturbation.

Can anybody point me to a really good source for the design, care, and feeding of a bandgap reference made out of POST (plain old silicon transistors) and garden variety components?

The caution I should probably express is that this bandgap is going to be used to tell when those batteries should be charged, and the batteries themselves are the only reliable voltage source in the circuit. The charging source can't be used, as it is a silicon array that may (daylight) have a voltage greater than the batteries or (night) have a voltage less than the batteries. It should suck at most a milliamp or so.

I don't really care what the bandgap voltage itself is, as I'll reference that voltage to a divider - comparator to turn the charger on and off.

Thanks...

Jim

Reply to
RST Engineering (jw)
Loading thread data ...

[snip]

POST will work, but not all that well due to differential temperatures. Your best bet is to use a transistor array, such as is shown in....

"CA3046-BandGap.pdf" on the S.E.D/Schematics page of my website.

Of course, if you were making a chip, I have micro-power designs for bandgaps that will start at as little as +1.5V and vary less than a fraction of a mV over temperature and power supplies as high as +5.5V ;-)

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
I love to cook with wine.      Sometimes I even put it in the food.
Reply to
Jim Thompson

To both Jim and Rene...

Thanks for answering and suggestions...but...

This is for a magazine article where the builder isn't all that sophisticated AND he has to be able to get the parts in his local Rat Schlock on Sunday afternoon in East Undershirt Ohio.

I'm not looking for great temperature performance, just something better than the soft knee of a 3 volt zener.

Quick and dirty is just fine by me. If the regulated voltage is off by a percent or so (what's that, 10,000 ppm) over the normal outside temperature range, I can live with it. That means the charge/cutout voltage will vary

50 millivolts or so, and the NiMH can certainly handle that.

Given this further information, any suggestions?

Jim

Reply to
RST Engineering (jw)

"RST Engineering (jw)" wrote in message news: snipped-for-privacy@corp.supernews.com...

Jim's website link answers part of your question reasonably well, unless you mean to avoid doing any math whatsoever. In that case, you had better just give it up because the specs you mention are unreasonable without some careful design, especially if you insist on using discrete transistors.

AND he has to be able to get the parts in his local

That was once a "reasonable" constraint. Anymore, it's laughable. All the more so with mail/web ordering as it works now.

the soft knee of a 3 volt zener.

percent or so (what's that, 10,000 ppm) over the normal

voltage will vary 50 millivolts or so, and the NiMH

You might get it done if you were willing to adjust the result, bond the parts together thermally, and use more than one per transistor position to deal with gradients. Somehow, I doubt you are that interested.

If you're a troll, declare victory and go away.

--
--Larry Brasfield
email: donotspam_larry_brasfield@hotmail.com
Above views may belong only to me.
Reply to
Larry Brasfield

On just roll your own out of a few parts, as I did last week, see...

"BandGapExample.pdf" on the S.E.D/Schematics page of my website.

;-)

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
I love to cook with wine.      Sometimes I even put it in the food.
Reply to
Jim Thompson

Did you see my post with the suggestion to see...

"CA3046-BandGap.pdf" on the S.E.D/Schematics page of my website?

You can do that with discretes also, probably good enough for what you need.

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
I love to cook with wine.      Sometimes I even put it in the food.
Reply to
Jim Thompson

See also...

"DiscreteBandGap.pdf" on the S.E.D/Schematics page of my website.

Geez! This is fun, showing off how fast I can crank out BandGaps ;-)

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
I love to cook with wine.      Sometimes I even put it in the food.
Reply to
Jim Thompson

My business IS designing chips, so hardly a week goes by that I don't design a BandGap to fit a particular process and chip architecture requirement.

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
I love to cook with wine.      Sometimes I even put it in the food.
Reply to
Jim Thompson

Why don't you just use a part such as the Analog Devices ADR318, with a 1.8V output from 2.4 to 15V input at a TC of 5ppm/K ? At a few $ a bargain. And it only draws 120uA. The ready made references are hard to beat in terms of precision, power requirement and price.

Rene

--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net
Reply to
Rene Tschaggelar

Not surprised. Anything out of M$ is bound to be used hay.

Go to

formatting link
and enter LM285 into the search box. It will lead you to spec sheets, application notes and prices. In particular, the LM285M-1.2-ND is $1.16 each in oneses. That part is SOIC package. If you'd rather To-92, LM285BYZ-1.2-ND at $2.57.

Ted

Reply to
Ted Edwards

Been there, done that about 40 years ago. Great for learning but for using, I'd rather go with NS. :-)

*Very* nice site!!

Ted

Reply to
Ted Edwards

Funny- but no one slow charges NiMH by threshold voltage-" Slow-charging a NiMH battery is difficult, if not impossible, because the voltage and temperature profiles associated with a C-rate of 0.1C to 0.3C don't provide a sufficiently accurate and unambiguous indication of the full-charge state." Any higher C-rate is "fast charging" and termination on temeperature gradient is the most reliable and practiced method- voltage measurement only comes into play for determination of necessity of a pre-conditioning trickle cycle before proceeding to fast-charge phase for an over-discharged battery.

Reply to
Fred Bloggs

Thank you, Jim, that was EXACTLY what I was looking for. I absolutely detest "design by cookbook"; if I can't get a feel for why the design works as it does, I know for damn sure the thing is going to leave teethmarks in my hiney some day when I least expect it. Lord knows, even if we DO know why we think something works, we get bit from time to time.

I hadn't ever had the need or the pleasure of visiting your website before. Really nice and well done. I'll peruse it at leisure some day when I don't have a deadline slapping me smartly upside the head.

Jim

Reply to
RST Engineering (jw)

I don't mind "doing the math" if I have some basis for WHY the math works as it does. If you are a cookbook engineer, then I guess plugging into the formula is fine, but I need an explanation of what happens when the formula doesn't produce the expected results.

My editor pays me moderately well for publishing designs that Joe Sixpack can kludge together and make work. If I *absolutely* have to have a special part, I'll use one. If I can get by using floorsweep spec parts, I'll do so. There *is* an art, you know, to designing for the 3 sigma limits of the component distributions.

Somehow I doubt that I need that level of sophistication. Jim gives very detailed equations for what happens when things aren't perfect and I can live with the expected variations.

You know, I've hired assholes like you who think they are the second coming of Tesla. They last with me about as long as you are lasting with me in this thread.

PLONK.

Jim

Reply to
RST Engineering (jw)

Funny, but almost every manufacturer of NiMH batteries gives similar instructions for how to slow-charge their batteries and the equations/voltages to calculate when a battery is fully charged. If you don't have the luxury of assuring enough current to do fast charging, you do what you have to do to make the sucker work.

Jim

Reply to
RST Engineering (jw)

Jim, Google for the sed post below.

Message-ID: From: Win Hill Newsgroups: sci.electronics.design Subject: Re: High Power Zener?? - zero tempco version Date: Fri, 10 Aug 2001 12:42:48 GMT

--
Tony Williams.
Reply to
Tony Williams

I have no idea if Panasonic knows what they are talking about. Try these:

formatting link

formatting link

or just google on

nimh battery charging design trickle

and use any of the 11,000 hits you feel appropriate.

Jim

Reply to
RST Engineering (jw)

Really? Would Panasonic know what they're talking about? They say "Trickle charge is not recommended for NiMH batteries." And the only method they describe is rapid charge.

Reply to
Fred Bloggs

My SPICE runs on that circuit show that it does not exhibit particularly good regulation against input voltage variation, and it relies on Icbo multiplication for start-up which may not be such a good thing especially at low temperatures. Radio Shack still carries the LM317 and the Temperature Stability graph shows under 2mV change in reference over

0-70oC. Maybe you can use that- it does require a minimum current that is a bit high, 10mA worst case over all operating conditions, but if that is a major consideration, there is such a thing as pulsing it on to make your battery comparison which can be done every 30 seconds or so for a millisecond time frame- no law says it has to be done continuously. Line regulation on that part is outstandingly good, as low as 0.005%.
Reply to
Fred Bloggs

I read somewhere that "fast charge" current spikes is a better way to go than a reduced steady current. NiMH and nicad batteries have a problem with growing little shorts in the cell if you don't charge them right.

--
--
kensmith@rahul.net   forging knowledge
Reply to
Ken Smith

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.