Analog multiplier still alive ?

Hello,

I am from Belgium (Europe) and have a problem in obtaining an analog multiplier for a basic design. I have been digging up a few types like AD532, AD534, AD633 or even MLT04 from Analog Devices

It seems to be pretty hard to find one of these. Multipliers have become pretty expensive (about 20 Euro, about 24 Dollar) or should be ordered in high volume (500 pieces). I only need 2 of them !!

Can anyone from the regio advise a simple common type (it shouldn't even be 4 quadrant) that is still common in Europe ?

Thanks a lot

Eng. Geert DB Belgium

Reply to
Pinchy
Loading thread data ...

I like the expensive AD734. The AD633 isn't too expensive. Sadly, the cheaper ones have become toast, unless you buy hundreds or more.

--
 Thanks,
    - Win
Reply to
Winfield Hill

Farnell list a bunch of Analogue Multipliers - their price for the Analog Devices AD633JN is 13.60 euro (plus tax) for one to nine (and they will sell just two).

They also list the New Japanese Radio Corporation NJM4200D for 7.52 euro in the same quantities - it isn't a multiplier that I've ever used, but you might want to take a look at it.

----------- Bill Sloman, Nijmegen

Reply to
bill.sloman

That's their version of the old Raytheon RC4200, now made by Fairchild. This two-quadrant multiplier has sinking current inputs and outputs.

--
 Thanks,
    - Win
Reply to
Winfield Hill

Is the LM13600 an in-production part? I thought it had been obsoleted by National, but maybe somebody else picked it up.

More expensive but similar is the Burr-Brown (now TI) MPY634.

If this is for audio/synth use (2-quadrant good enough) the "electronically programmable potentiometers" are somewhat available and useful for many things. Doing Google searches for "VCA" turns up a lot of circuits used in analog synths, some based around jellybean components. (Even though jellybeans like CA3046's have been discontinued in the past few months, too!)

Tim.

Reply to
Tim Shoppa

I'd consider using a different scheme, preferably one that uses common easy to obtain components. One possibility is a pulse width, pulse height multiplier. A simple one can be designed using a 555 in a pulse width modulator mode to convert the first analog voltage to a constant frequency, variable width signal on the output of the 555. This is clamped with a second analog voltage to a precise pulse height in an op-amp. A second op-amp connected as an integrator, integrates these pulses into an analog voltage output that is the function of the product of the two inputs. This scheme is relatively free of drift and temperature effects, has a good dynamic range and can operate at a reasonable bandwidth depending on oscillator frequency.

If these signals or their products are ever digitized, multiplication in the digital domain or multiplication by a variable ADC reference are also possibilities. I say, anything but an analog multiplier for cost and availability reasons if nothing else. Bob

Reply to
Bob Eldred

[snip]

AD633JN is about the cheapest of the 4-quadrant pre-trimmed multipliers. Very easy to use, few external components normally required.

The 4200 is even cheaper, but only easy to use when the input signals are restricted to being positive only.

Operational Transconductance Amplifiers (LM13600) are peanuts and can be made into multipliers, but require development work and the results may not be accurate enough. For just 2 pieces the AD633JN probably represents the best overall value for money.

--
Tony Williams.
Reply to
Tony Williams

Excellent for unipolar use, especially for XY/Z apps, but a total swine when used with AC signals..... you need a doctorate in precision resistor computation. The cost of an output opamp also has to be factored-in.

--
Tony Williams.
Reply to
Tony Williams

From relatively recent memory New Japan Radio Corporation - variously known as JRC in Europe or NJR in the states introduced a 'consumer quality' multiplier a few yrs back.

Sorry - don't have a part number. I was probably alerted to it by profusion - the audio semi specialist distributor though.

formatting link

I think it used the transconductance amp principle but wasn't as limited as the 13600/700 though.

Graham

Reply to
Pooh Bear

Curious. That may be the part I mentioned in another post to this thread.

Anyway I can't find it on the Farnell site and it doesn't seem to be on JRC/NJR's either.

Graham

Reply to
Pooh Bear

Talking of things being discontinued, I was astonished to see Philips remove the NE5532/4 ( the classic high performance audio op-amp ) from its current product line-up. Still available from TI & JRC though.

Graham

Reply to
Pooh Bear

I relabel 5534's toTDA1034 to give them that "pre CD" sound.

All in all, it is a stunning bit of silicon, there are better, (AD797?), but what a price differential

martin

After the first death, there is no other. (Dylan Thomas)

Reply to
martin griffith

LOL ! I actually have some original TDA1034s somewhere in the metal can package !

Collosal. The 5532 is typically around 20-30 cents in quantity. I guess Philips weren't making money from it anymore.

Graham

Reply to
Pooh Bear

What about MC1495? Still available from some shops, even in Europe (ex. fibra-brandt).

-- Andy

Reply to
Andy

Hello Pinchy,

If it doesn't have to be top notch precision look at the old workhorse MC1496. It's under a Dollar single qty and around 30c high qty in the US. Reichelt in Germany has it for 53 Eurocents. Don't know about Belgium, there I usually concentrate more on the good stuff (Duvel, Trappiste and so on...).

Groetjes, Joerg

formatting link

Reply to
Joerg

Wonder what they do with the old artwork/masks etc (limited knowledge..)

sell them off?

martin

After the first death, there is no other. (Dylan Thomas)

Reply to
martin griffith

Back in 1974 I worked on a pulse-width multiplier, built with LM301 op amps and saturated transistor switches. It was extremely accurate - about 0.1% - but as slow as a wet week.

A modern design with MOSFET switches (most likely CMOS) faster integrators, and decent comparators could be quite a lot quicker and equally precise. It wouldn't get anywhere near the MHz bandwidths you get with a Gilbert cell analog multiplier, but for some applications could be a cheap way to go, if you've got the board area available.

---------------- Bill Sloman, Nijmegen

Reply to
bill.sloman

Time for an ascii drawing, Bill.

Or two.

--
 Thanks,
    - Win
Reply to
Winfield Hill

Nowhere near the dynamic range? Massive overkill? :-)

I know, I know, flashing an LED requires a PIC these days, if not an ARM CPU...

Reply to
Joel Kolstad

Why not use a DSP? Digitise both inputs, multiply them, and output via a DAC?

8-)

Leon

--
Leon Heller, G1HSM
http://www.geocities.com/leon_heller
Reply to
Leon Heller

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.