RX PRE AMP

[sinclairuser]

hi been reading the projects thread, and i have seen tinkerer's and moods approaches but has anyone here developed a simple pre amp that can be adapted to different projects?.
what i'm after is something that can be added to a detector as block between coil and rx input.
preferably this pre amp would not need a centre tapped coil and would use common components.
thanks ally

[Davor]

Your best bet is a differential amplifier built with BJT transistors that have low Rbb. You can achieve noise below 1nV/sqrt(Hz) with such preamp.

[Waikiki_Sweep]

As you can see for last 20 years NE5534 was used everywhere because of low noise.
This century you probably you can find some more advanced stuff.
Because of huge variety of offered ICs find something you really need is not too easy.

Probably key parameter to choose opamp is noise. We amplifying signal in range of microvolt.
Noise depends on frequency. Many opamps having good noise parameters in 1 kHz may have very bad noise at 10 Hz (like LM318 and LF357)
so you will hear low frequency oscillations. Check noise to frequency charts for opamp you checking. Also can see CA3130 in some solutions.
As I know it is not "low noise" opamp at all. So it has few digits more noise than "low noise" opamps.

Another parameter - slew rate - ability of opamp to react fast on input changes. Probably you need 10 V/μS or more.

If you going to use micro-controller it is good try to use one power supply 5V (or 3.3V) to have easy way
to connect your digital and analogue signals or you will need much more components to convert voltage levels from +5..-5 to 0..+5 volts and back.
Also you can skip -5V voltage convector ICl7660 making stuff simple and remove additional noise source.

Some of good low noise opams for one power 5V I found: ISL28191 and ISL28291
1.7nV/√Hz input voltage noise at 1kHz (2.5nV/Hz for 10Hz )
http://www.intersil.com/content/dam/...n61/fn6156.pdf

If you going to keep +5 and -5 volts power supply you can check a very low noise LME49990
(You can see it in QED detector - Patented High Tech PI detector from Bugwhiskers)
• 0.88nV/√Hz input voltage noise at 1kHz (1.3nV/Hz for 10Hz )
http://www.ti.com/lit/ds/symlink/lme49990.pdf

I did not spent much time to find those opamps. Probably digging more will give you more explorations.

[Davor]

These are surely very good amps, but for balanced input you need a balanced solution. THAT amplifiers are fine, but IMHO you can get there much cheaper, and surely more in spirit of a real DIY with plain old transistors. BC337 is cheap as chips and has Rbb well below 100ohm. If you buy a small quantity of these and match them, you can build a very fine balanced preamp for very cheap.

[Waikiki_Sweep]

I am sure you are right. I see big companies sometimes build preamps from transistors like Fisher CZ:

I see only one problem. Reliability of device depends on reliability of every component.
Adding 20-30 discrete components will reduce reliability of hand made home build detector.
One bad soldering and oops..

[sinclairuser]

thankyou, davor and waikiki, i have been reading datasheets lately to the point of blindness, i have read about some exotic opamps that sound great on paper but as waikiki eluded to the main problem is noise from components used to make the opamp work, i am analogue by nature and through hole by nesessity, i have only just started playing with pics,cplds, etc, all this lessens my opportunities.
i had intended to design and built this my self, with the intention of giving it to the forum, but i got so bogged down i decided to ask if others had done anything similar.
so i will look into davor's suggestion, solid state is my thing after all and at the end of the day if i build something and it works, maybe smarter more recently qualified people here can take that idea and build something better!.
one again thanks.

[Waikiki_Sweep]

I am disappointed too because it is impossible to find best new OpAmps in DIP (through hole) cases. Companies just do not want to made them in DIP.
And every day you will sey that migration to SMD until all through hole chips will die like dinosaurs.
But at least we have a solution to use SMD chips in our through hole PCBs.
For several cents you can buy SOP8 SOIC8 SO8 SMD to DIP8 Adapters on eBay:

[Davor]

I see only one problem. Reliability of device depends on reliability of every component.

That's more of a problem for mass production. True matched transistors are very exotic nowadays, and true low noise transistors are also not very easy to find. But as a real DIY-er you can make it work well with a proper choice of transistors. Heck - you are building it for yourself.

[sinclairuser]

thanks for the encouragement lads, those adapters are neat and the 7660 and ne5534 in my 2nd barra are smt versions with adapters, i can and have replaced smt components during repairs but i find them fiddley and i use a magnifying screen which would send you cross eyed if used for any length of time.
so in terms of complete builds i dont see the point of making it harder for yourself, and tend to stick to through hole and when smt chips must be used, due to availability or cost i use adapters, as a side note a guy on ebay even sells transistor adapters, they look like regular transistors but the tops are small triangular boards that take the smt versions but they are relatively expensive, and i have never had a problem finding the correct through hole transistors.

[Brian Deese]

I have been trying to study up on opamps myself lately,a question I have is why a current feedback amp would not be a good choice for a pi preamp. Maybe the gurus will shed some light on this. I swapped out a 5534 with an ad8055 and it was faster because It showed ringing where the 5534 wouldnt come out of saturation. But the output looked noisy and unstable. The noise spec isnt as good for the 8055 and maybe thats what I am seeing on the output. I also left the gain at 1000 for the comparison,I'm gathering that at that high of gain even the best amp isnt going to show much improvement. I have noticed that the current feedback amps are generally set up with lower gains to maintain their specs. I'm reading some of Walt Jung's stuff on amps trying to learn something.

[sinclairuser]

what im playing with at the moment has 48x gain, and allthough not for a detector, many of the principles are the same, i have two versions different in design but identical in output and input one based on the tle2072, the other is based on the lme 49999, and as soon as you crank up the gain the problems start.
with the pi preamp the first question must be, how much gain is needed?.

[Waikiki_Sweep]

lme49999

Equivalent input noise
f = 10 Hz 1.3 nV/√Hz
f = 1 kHz 0.88 nV/√Hz

tle2072

Equivalent input noise
f = 10 Hz 48 nV/√Hz
f = 10 kHz 12 nV/√Hz

[sinclairuser]

lme49999

Equivalent input noise
f = 10 Hz 1.3 nV/√Hz
f = 1 kHz 0.88 nV/√Hz

tle2072

Equivalent input noise
f = 10 Hz 48 nV/√Hz
f = 10 kHz 12 nV/√Hz

i did say it was not for a detector , its a commercial work project, one version built with my ideal components, the other built with cheaper more common components, in the application at hand neither version works as intended.

my post above only relates to this thread in terms of brians comment about opamp noise increasing with gain, do not try to reconcile my comments with any detector projects as its not relevant, when i said "the principles are the same" i only meant in the basic principle- input in, output out, no further discussion will be made by me on this issue..
as both designs belong to my employer, it was only a passing comment one i regret making, if i could i would remove it as i'm not supposed to discuss any work projects here.(i was very tired).

[Carl-NC]

I have been trying to study up on opamps myself lately,a question I have is why a current feedback amp would not be a good choice for a pi preamp.

They can certainly be made to work just fine, but don't have a lot of advantage in a PI preamp. The advantage of CFA over VFA is usually in slew rate and high gain BW. You'd think higher slew rate would help but in PI but the overvoltage recovery is more important. High gain BW is easily achieved in VFA with a 2-stage design. Drawbacks of CFA are higher noise, higher power consumption, and higher cost. CFA's can also be a little squirrely to properly compensate, so homebrewers might have trouble getting them to behave.

- Carl