Replies to Minipulse Plus

[Davor]

Thanks for the final version B. It is very nicely described, and I'll order my PCB(s) in a week.
I wonder about possible improvements in form of replacing the NE5532 with NJM2068, my favourite op amp of late. It is less noisy (with 1k towards the coil it does not make much difference though), it is faster, and it has better THD. I have no idea how it would do in a PI front end.
BTW, schematics look immensely better in documents when they are placed there in just about any format other than jpeg. Only jpeg makes that ugly artefacts around lines and text. For me gif works the best.

[Qiaozhi]

Thanks for the final version B. It is very nicely described, and I'll order my PCB(s) in a week.
I wonder about possible improvements in form of replacing the NE5532 with NJM2068, my favourite op amp of late. It is less noisy (with 1k towards the coil it does not make much difference though), it is faster, and it has better THD. I have no idea how it would do in a PI front end.
BTW, schematics look immensely better in documents when they are placed there in just about any format other than jpeg. Only jpeg makes that ugly artefacts around lines and text. For me gif works the best.

The Minipulse Plus is a design that we (Geotech members) can take further if desired. Please feel free to experiment and make improvements. You might also want to try even higher TX pulse rates, and longer pulse lengths, whatever. In general though, I would like to keep this design simple to build, without lots of extra options and adjustments. I think we should leave the design as it is for a while (at least until the current batch of PCBs is exhausted), with possible modifications scheduled on an annual basis. That would give everyone time to develop new ideas ... such as adding ground balancing, for example.

Regarding the images, I agree. The 3D PCB image on the title page is a GIF, whereas all the others are JPG; except the TX oscillator waveform, which is a BMP (generated by the scope). I'll change these to GIFs when the document next gets updated.

[6666]

Thanks for Ver B, very interesting that you have managed to do away with offset adjustment on preamp.

PCBs are available from www.silverdog.co.uk

When they are listed I will order a PCB.

[bernte_one]

so andy please list them in the shop

[Davor]

Thanks for Ver B, very interesting that you have managed to do away with offset adjustment on preamp.

That's the very reason I'm up to making this particular rig. Differential integrator is the key feature of this gem that is very forgiving. It acts as a balanced detector so even the FETs charge injections cancel each other. Trouble with most detectors in amateur stratum are half wave VLFs and unbalanced PI detectors. This one is a step beyond

I believe this is a nice platform for further improvements, however, I don't expect many of those.

GB will be a nice challenge.

[silverdog]

OK, I have a limited number of REV-B boards left from the prototype run - I have put 10 in the shop for those awaiting them
I will be making a very small number of kits later in the week - then arranging a further batch of boards

Please thank Qiaozhi for all his hard work on this project, without this kind of commitment we would not have projects to build

[6666]

Thanks Andy and Qiaozhi, have ordered a PCB.

[eogeo]

Thanks Qiaozhi for all the effort you have put into this project and a first class build manual which will make assembly and testing so much easier. PCB ordered. Also thanks to Silverdog for providing a great service.

There was a mention somewhere else on Geotech, Baracuda thread perhaps, that matching the J113 FET's may improve performance a bit. Is that also the case with this design?

[bernte_one]

ok i will take a two kits as soon as available

[Qiaozhi]

There was a mention somewhere else on Geotech, Baracuda thread perhaps, that matching the J113 FET's may improve performance a bit. Is that also the case with this design?

Personally I don't think you would notice any difference.There are two samples taken: the main sample and the secondary sample. The secondary sample is taken much later after the eddy currents from any metal targets have more or less died away. However, both samples will contain signals that are received from an external source (in this case, the Earth field), so subtracting the second sample from the main sample will eliminate the unwanted external signal. There would be virtually no benefit from matching the FETs, as they are either turned on or off, and do not operate in the linear region.

[Davor]

Often neglected is a gate charge. In effect all half wave detectors with switches, and even more so with FETs suffer from gate charge injection. In balanced configurations, such as this one, FET gate charge injection is cancelled together with any offset and earth field.

[eogeo]

Thanks for the FET matching explanations. Good to know it is not a factor in this design.

[6666]

I have ordered some real J113 fets, in the mean time is there a suitable replacement ?

I have also ordered some LT1054 chips, in the mean time will just use a 9V battery for U3.

[sinclairuser]

before any jerk starts saying i'm sore that the kits are not up yet(its not that).
but i wondered about some of the component choices, some of these bits are not that common, i'm also wondering about the cost of kits, some of the prices i have been quoted for some parts as singles have made me give up(i never normally buy singles always multiples), i intend to wait for the kit, may even wait until the next revision if i'm honest, when no doubt some of them will have been replaced

[Qiaozhi]

... i wondered about some of the component choices, some of these bits are not that common ...

Which parts do you consider "not that common"?

The bipolar transistors have been replaced by general purpose types (2N3904 and 2N3906), and all the other parts are pretty standard. The only thing you may not have in stock is the LT1054, but even so it's inexpensive and readily available.