LED reference current source

pads

That's helpful but not necessary. Just being cheerful and friendly would be a good start.

--

John Larkin                  Highland Technology Inc
www.highlandtechnology.com   jlarkin at highlandtechnology dot com   

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME  analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators

Thanks. I know Rick Karlquist slightly--smart guy.

Cheers

Phil Hobbs

-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203 Briarcliff Manor NY 10510 845-480-2058

hobbs at electrooptical dot net

than

that

uld

hunk

ld

her,

nt.

109... e

Oh, I am - unless you post nonsense.

-- Bill Sloman, Nijmegen

Or so the manufacturer tells you. Try to get it into your head that making the output of the bridge an AC signal with a controlled frequency - which you can distinguish from the 1/f noise - is better than having the output of the bridge as a DC signal which you can't distinguish from 1/f noise.

-- Bill Sloman, Nijmegen

pads

You're never going to change. Get used to it.

Today I think we'll destroy some Analog Devices laser driver chips, to see how far they can be pushed [1]. The big problem with telecom chips is that the people who sell them can't imagine they would be used for anything but telecom apps. The people who make RF parts have a similar mindset.

[1] This particular chip also draws Vcc current in the form of an 80 Hz square wave, which causes jitter elsewhere on our board. ADI doesn't know why. They have suggested that we are doing something wrong and causing a thermal oscillation of some sort, which is plain silly. We figure there is some undocumented laser-control servo loop inside, switching at 80 Hz, and the people that we can talk to don't know about it.

--

John Larkin                  Highland Technology Inc
www.highlandtechnology.com   jlarkin at highlandtechnology dot com   

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME  analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators

IC sensors tended to be packaged like other semiconductors, and should be about as fast as an RTD, if you take care to mount them in the right place, in the right way.

Thermistors are not well packaged, and do tend to be a bit slower in consequence, which doesn't stop them from being the most practical solution in a great many cases.

-- Bill Sloman, Nijmegen

Nobody's doubting that. It isn't like AC measurements are novel or anything--I've done them often, and I'd be very surprised if John hasn't as well.

It's just that good modern choppers, e.g. the OPA378, really don't have

1/f noise, at least not where my measurements are sensitive (down to the tens of microhertz), and that changes the game. Your average demodulator will have worse 1/f noise than that, not to mention its duty cycle sensitivity and DC offset drift. Residual thermal drift is more of a problem, but that's solved by putting the control circuitry (excluding the power driver) inside the controlled zone, and biasing everything at its maximum-power point to avoid thermals.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

Nope. Read the datasheets--typical TC for a TO-92 is 10 seconds in stirred oil. The other problem is that you basically read the temperature of the leads. SMT ones are probably faster, but also far more susceptible to drifts due to die stress.

SMT thermistors mounted on flex are much faster than the glass bead kind, and if you do it right, the thermal conduction through the copper is small. Wrapping the flex around an anodized aluminum thermal shield with the thermistors on the inside and attaching them to the anodized surface with indium solder paste makes them very competitive for speed, but of course it's a bit fiddly to avoid solder bridges. Maybe soldering one end only would be the best thing--I haven't optimized the process, for sure.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

Conservation of energy requires that an RTD not generate voltage 1/f noise. All it can have is Johnson noise.

If its *resistance* has a 1/f component (likely caused by minute temperature fluctuations) modulation just promotes that up into the working passband. Zero advantage for using AC.

A chopper opamp *uses* modulation/gain/demodulation to kill 1/f noise, same as your AC scheme, but all in one SOT23 package for $1 or so.

An AC based bridge needs, among other things, a very good, low noise excitation oscillator. That alone can be difficult.

Hmmm, one could make a bridge with two thinfilm platinum RTDs in opposite arms, to double sensitivity and do some physical sample-location averaging.

(You might note that the above is a group of technical statements with no personal insults embedded.)

--

John Larkin                  Highland Technology Inc
www.highlandtechnology.com   jlarkin at highlandtechnology dot com   

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME  analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators

Yup, AC power meters, inclinometers, FSK modems, LVDT acquisition, and, currently, an LVDT/synchro/resolver simulation/acquisition board. But those are inherently AC systems, where synchronous detection is mandatory.

(I have a PhD still working on advanced synchro tracking algorithms for that one. It's currently doing digital synchronous detection and dumb trig, good enough for static acquisition.)

With modern chopamps, I can't see any reason to use AC for RTD acquisition, and lots of reasons not to.

--

John Larkin                  Highland Technology Inc
www.highlandtechnology.com   jlarkin at highlandtechnology dot com   

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom timing and laser controllers
Photonics and fiberoptic TTL data links
VME  analog, thermocouple, LVDT, synchro, tachometer
Multichannel arbitrary waveform generators

There are probably still situations where it's better, e.g. when the controlled volume is wet or corrosive or hostile in some other way, or else very small, so that it's really nice not to have to put the control circuitry inside it. Then you do have to worry about thermocouple offsets, so AC would make sense even though it's much more complicated.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net

If you put two in series, and put a high-value pot in parallel (wiper to the midpoint of the RTDs), you can adjust where the neutral line is, i.e. just where the local temperature tracks the average of the two RTD temperatures. That's useful when there's some inescapable asymmetry in the object, e.g. it has to be mounted on one side only.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net

better than,

than urethane.

inner

around the

hour would

that -

which

microdegrees. A

to AC-

as good

ovens.

even that

inside a

Actually, drilling for oil and gas is a decent-volume market.

--

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot com
http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation

h e

uld

ch

s.

.

ure

re

e,

or

lel

ill

of

es

're

y a

e a

it

idn't

xcept

e

Thanks, I'm mostly thinking about a diode connected transsitor (SMD)

I'm not understanding this idea. I've got a diode mounted on flex, the other side is solid copper and I solder that side to the heat sink. You use indium because of the low melting point. (the rest of the circuit won't de-solder.) Will indium wet anodized Al?

y r

Hmm how bad is the stress sensitivity. I don't anticipate needing the ~milli K stability of say a diode laser.

Well we've sold several diode connected transitors as temp sensors. (the price is right :^) I saw no change in the LN2 voltage reading with repeated 'stress test' dunkings. I have no idea if there is any long term drift. The best part is I can do a single measurment (at room temp against a known sensor) and get calibration from 77 to 400 K that's good to

0.5mV (and much better than that over most of the range.)

(That graph's a bit criptic, it's the difference in voltage for transistors from the same batch.)

George H.

t -

Yes, that's the trick. You get the low thermal resistance of solder without shorting out the thermistors and heater resistors.

Not sure. Try a voltmeter and some pliers!

Cute. The wonders of centre-to-edge uniformity on the wafer.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

n Lark>

000 n

But how often do they want micro-kelvin temperature control at the bottom of the bore-hole?

-- Bill Sloman, Nijmegen

Fine. Now tell me how conservation of energy stops the thermocouple voltages and all the other low level sources of low frequency noise that screw you up in real life.

Your imagination is working overtime. Pity about the rest of your brain.

Wrong. If the bridge is AC-excited, the signal coming out of the bridge is AC at the excitation frequency, and everything else is noise. Your $1 SOT23 package doesn't have that particular advantage - as you'd realise if you engaged your brain for thirty seconds.

Really? The excitation oscillator I'm trying to put together would be close to ideal, but would have been much too expensive for any of the real world jobs where we used AC excitation. In many applications "reversing DC" is perfectly adequate.

Nor - sadly - any evidence of technical insight. How on earth your can think that a chopper stabilised amplifier is in any way equivalent to AC excitation and synchronous demodulation quite escapes me. Phil Hobbs may be correct in asserting that such an amplifier may make a DC- excited bridge good enough, but at least he's sufficiently well- informed to understand that it isn't the same thing.

-- Bill Sloman, Nijmegen

Lark>>

Sorry, I'm not allowed to talk about that.

--

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot com
http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation

The advantage of using DC, and a chopamp, is that you don't need many parts, and the parts are really, really tiny. So the RTDs and all of the critical circuitry can be snugged right near the object to be controlled, deep inside the oven. Thermoelectrics will be minimal and mostly just a gain shift. Other sources of noise can be prevented with sensible design. An AC system has lots more hazards, especially if the oscillator, amplifiers, and PSD are outside of the oven.

If you won't address the factual issue here, your brain isn't working at all. You're just recycling lame insults.

Obviously. There's nothing I can do about that.

--

John Larkin         Highland Technology, Inc

jlarkin at highlandtechnology dot com
http://www.highlandtechnology.com

Precision electronic instrumentation
Picosecond-resolution Digital Delay and Pulse generators
Custom laser drivers and controllers
Photonics and fiberoptic TTL data links
VME thermocouple, LVDT, synchro   acquisition and simulation