Amplifier noise

[Anonymous]

Hi Mr. Bill
Happy New Year! I'm only just recovering after a "heavy" night - we are high up on a hill in SE England - it was a cold clear night and we had at least 25 "free" fireworks displays all around. Good party...
Anyway, I am involved in manufacturing electronic products, mostly DSP-based. Perhaps I should say I design them, and fabricators manufacture them. So the economics of the process are well understood which is why I asked the question about what is gained (if anything) from using it. Time moves on, new components appear. If they are cost-effective, use them. If they are effective, but expensive, maybe just use them in private projects. Experimentation moves the process along... I could put a stonking great DSP in my PI detector (based on what I do for bio-acoustic ultrasound stuff), but I have yet to be convinced about the merits of that approach - which is why (for me at least) why Robert's work is so interesting. Good stuff, that...
Which reminds me... The DSPs I mainly use (ADSP-2191Ms) are about USD 16, which when you consider what you are getting, is extraordinarily cheap. However, to use one I'd need a 4-layer board, a clean 2.5V as well as 3.3V supply, and really really carefull attention to layout...
Regards,
Nicko

 

[Anonymous]

Hi Mr. Bill and Happy New Year,
I didn't mean that you aren't already using the best that was available at the time of the design of the Goldquest or the Aquastar. As time goes on newer and better IC's are developed. I just meant that if a newer and better IC comes along that can actually enhance the performance of the detector that is shouldn't be passed by because it costs a bit more than the ones currently being used.
As you can see by those that have been doing the experimenting with different OP Amps and posting thier results here there is a difference in what is written in the spec sheets on these IC's and what happens when they are used in the front end of a PI. Only continuing experimentation will tell the true tale of whether or not a particular IC will or will not work well when used as part of a PI.
HH
Beachcomber

 

[Anonymous]

Happy New Year Nicko,
Since most detector run on batteries I don't think coming up with a very clean supply voltages would be a problem, however the 4 layer PC board might be. I believe and I could be wrong on this that most detectors currently being made at the most use only 2 layer PC boards. I don't know much about the cost of manufacturing multilayer PC boards. Roughly how much of a price difference is there in manufacturing a 4-layer vs 2-layer PC board?
Thanks
HH
Beachcomber

 

[Anonymous]

Happy New Year Beachcomber,
I have tried several different opamps also and have also returned to the NE 5534. Yes, opamps like the AD 8055 can generate noise that will affet the overall sensitivity so one has to be careful in the selection. However, I have still found that the NE 5534 is still one of the quietest and seems to work very well.
I have to agree with Eric's statement he made earlier which was; "Usually it is external electromagnetic interference which dominates, and amplifier noise is of secondary importance."
Having operated my pseudo Goldquest in both populated areas and out on the desert miles from the common electrical congestion, I can say the opamp used is not the main source of the typical noise one will encounter with a PI. On the desert, my PI is extremely quiet.
I would also like to add that there are other internal noise factors involved that, in my opinion, generate more noise than the preamp opamp. This became quite obvious to me when I build and installed a DD coil.
Part of the noise reduction I experience with the DD coil could very well be because of the smaller diameter of the individual coils. Eric pointed this out to me some time back. I haven't compared noise differences an 8" coil to the 11" coil but I am sure there are differences.
In other words, one just might see an increase in depth capabilities just by using a smaller coil, simply because it may not pick up as much externally generated noise. This may sound strange but I think it is very possible. It is also very possible that such an increase would far exceed any possible increase that may be obtained by using the best possible IC for the preamp.
Reg

 

[Anonymous]

[No message]

 

[Anonymous]

Not a good day - just as I was typing my reply to this, my dev PC uttered an ominous "phut" and died... I ripped the PSU out of another chasis, and it still (seems to) work... sigh...
Anyway, 4-layer boards cost (not surprisingly) twice what 2-layer boards do. Howvever, prototyping charges are much higher, and my usual prototyping services (e.g. Olimex) don't do 4-layer and I can't make them myself...
So, its a pain to develop, and turn around for new boards is slow and expensive... but if you are working with BGA or LQFP-144 packages (as I do) then 2 layers is not enough....
Nicko

 

[Anonymous]

I think the graph below should show the effect that amplifier noise has on total detector noise. And I think that the 5534 must put you close to the knee of the curve. Using a lower noise amplifier would move you left along the curve, but that does not do much good because the curve is flat in that direction. Using a higher noise amplifier moves you to the right along the curve which moves you up to higher detector noise.
Of course the Other Noise depends on the environment and in a very noisy environment the amplifier noise becomes even less important.
Robert