New high voltage opamp - ADHV4702-1

[splin]

Quite impressive - +/-110V, Ibias 2pA max @25C, 8nV/rt(Hz), 170dB AVol, 160dB CMRR!

$21 each or $13.43 @ 1K (Mouser)
 
What are you going to use yours for?

[EDIT] 0.1 - 10Hz noise not specified but Vn <12nV/rt(Hz) @ 10Hz, approx 42nV @ 1Hz (reducing to approx 30nV if operating current reduced, which also reduces bandwidth, using Radj).



https://www.analog.com/media/en/technical-documentation/data-sheets/ADHV4702-1.pdf

[Zero999]

There are various high voltage op-amps available. They're not cheap. If you're doing a large volume it starts to make more sense to use a low voltage op-amp with a discrete high voltage gain stage.

[splin]

There are various high voltage op-amps available.

True, but it's much better than other TI and AD/LTC HV devices wrt Ibias, voltage noise and CMRR (at least) specs. TI have a new 180V OPA462 with 2pA typical @ 25C compared to 300fA for the ADHV4702, 100pA max ct. 2pA and much higher noise. Maybe better parts are available elsewhere. I don't normally post about new product releases but this one caught my eye as the specs seemed particularly impressive.

They're not cheap. If you're doing a large volume it starts to make more sense to use a low voltage op-amp with a discrete high voltage gain stage.

Or bootstrap the supply voltage of a low voltage amp to that of the incoming signal. Either way, you may have to make other significant compromises such as additional, higher leakage input protection (not easy to match the amp's 2pA over temperature) or higher input offset drift TC.

HV applications isn't something I know much about, hence my question as to what applications might this amp be particulary suited that could justify the high cost? SMUs might be one?

[Marco]

With bootstrapping you are also adding significant source resistance which the ADHV4702-1 can do without.

[NiHaoMike]

High voltage, high bandwidth opamps are very common in high resolution CRT monitors, as a hint for the hobbyists out there. (Low resolution monitors generally use discrete amplifiers instead of opamp chips.)

[Kleinstein]

The video amplifiers for high resolution CRTs are faster - this OP would be hardly OP for low quality NTSC TV.

The DS lists a few applications. These are usually more like low volume cases where the BOM costs are not that important.
A discrete of bootstrapped amplifier may be possible, but tends to use more power and needs extra testing. So in low volume products it's easier to use a relatively expensive OP if available.

One usually does not need all the nice specs at the same time. So one could well use such an OP for far less demanding jobs - like in a case where you don't care much about bias current.

[Doctorandus_P]

High voltage opamps would be nice to have in your toolbox, but those prices makes them prohibitively expensive for almost all applications.
So post again if you've found one that costs 20ct or so.

[001]

  (Low resolution monitors generally use discrete amplifiers instead of opamp chips.)

It is interesting for me
What discret circuit You can reccomend?

[Cliff Matthews]

..post again if you've found one that costs 20ct or so.

20 cents would be nice but ~$1 US hackers can go from 36v to 45v with something R2R like MCP6V51 (no dip8 though..)
https://www.digikey.com/product-detail/en/microchip-technology/MCP6V51T-E-MS/MCP6V51T-E-MSCT-ND/9826439

[Kleinstein]

The discrete video amplifiers usually used a cascode circuit with a high voltage NPN transistor. They often ran quite hot, e.g. as a class A amplifier.

Old video amplifiers can still be found as NOS - even down to 20 cents:
https://www.pollin.de/p/tea5101b-st-microelectronics-rgb-enstufentreiber-multiwatt-15-100733   
The datasheet shows a reasonable circuit for the gate drive that could be used similar in discrete circuit.

The amplifiers tend to be not OPs, but special, fixed gain amplifiers.

[dzseki]

The problem with the CRT type amplifiers is that the good transistors for those are long gone...
For high bandwidth sanyo had a few hybrids VPAxx line where xx corresponds to speed 13 is 130MHz. These could do 40-50Vpp (single sided) at the specified bandwidth. Alternatively you could build discretely the cascode circuit, great high BW high voltage transistors wer MRF548-549, BFQ232-252, but again these are long gone...

[Zero999]

If there's no DC to worry about and the bandwidth isn't massive, then one solution is to build a low voltage amplifier and connect a transformer to the output. Negative feedback can still be used to reduce the output impedance, but a lot of care needs to be taken to avoid oscillation, as the transformer will have some phase shift.

There's an example in the Art of Electronics.

[Doctorandus_P]

The good old OP07 goes toe 44V (+/- 22V) and +/-30V diferential input voltage, and this is not too uncommon for opamps.

Power opamps such as for examle for audio amplifiers (LM3886) also go a lot higher.

But I suspect low power high voltage opamps are rare because there is not much use for it. Adding some transistors to the output and/or a voltage divider on the input takes care of most situations.

Stuff is also mostly going down in voltages and power. From the 200+V in those obsolete glass bulbs to less than 900mV core logic in FPGA's and fast processors.

A 30V 3A power supply used to be a lot more used 20 years ago than now (but still usefull of course) Having a few good quality adjustable 15V 1A supplies will be sufficient or even preferable nowaday's.

[awallin]

Quite impressive - +/-110V, Ibias 2pA max @25C, 8nV/rt(Hz), 170dB AVol, 160dB CMRR!
$21 each or $13.43 @ 1K (Mouser)
What are you going to use yours for?

I think we used Apex PA340 for driving slowly changing DC-voltages +/-100V to electrodes...
not sure from the specs if the AD part has lower noise or not..
https://www.apexanalog.com/resources/products/pa340u.pdf

[splin]

Quite impressive - +/-110V, Ibias 2pA max @25C, 8nV/rt(Hz), 170dB AVol, 160dB CMRR!
$21 each or $13.43 @ 1K (Mouser)
What are you going to use yours for?

I think we used Apex PA340 for driving slowly changing DC-voltages +/-100V to electrodes...
not sure from the specs if the AD part has lower noise or not..
https://www.apexanalog.com/resources/products/pa340u.pdf

Much higher voltage, 350 v 220V but the noise is in a different league - 337uV rms in 10kHz BW v 0.8uV!

Vos max drift 250uV/K v 2uV

Ibias max @ 25C 200pA v 2pA

CMMR min 84dB v 140dB