F1A4 Mods anyone?

[downhalo]

Looking for some DIY F1A4 mods, anyone willing to share some tweaks/mods/etc?

- Dh

[Firey10_0]

hi downhalo
i have read on this and other forums that you modded your own F1A4. I have recently purchased one that has the vol & threshold mod but I was wanting to add the frequency, gain and timing mods to it. Would you be able to pass on any wiring diag or pics of these mods to help me achieve this. Any help would be greatly appreciated. Thanks in advance

[fo0bar]

I'm about to acquire an F1A4 and have also been searching high and low for some direction on how these upgrades are done and would appreciate any information anyone can offer.

I remember stumbling across a brief description somewhere on geotech but I forgot to bookmark it and maybe my search skills are lacking

Thanks

[nickel_n]

look at post #78
its talking about the modes
https://www.geotech1.com/forums/show...highlight=F1A4

[fo0bar]

The thread only goes up to #55 - is it possible you were looking at detectormods' post count rather than post #?

This has been the most helpful thread I've found so far. I feel you're referencing post #10? That's definitely the most helpful clue I've found in regards to changing the timings. My issue is that "the 2 very small smt resistors" is far too vague for me - there is a block of 6 tiny smt resistors directly near what appears to be the EEPROM, and I also don't understand how these modifications could result in the three levels of timings seen on modified models - as far as I would think, high and low would only correspond to two modes unless leaving the input 'floating' is another mode, which I highly doubt.

"Adding volume, tone and threshold are easy to do" - I'm sure it is but I'm not sure which circuit I should be loosely referencing to work it out now that I know it's "easy to do". Would the SD2000's circuit available here be similar enough to the F1A4 on the pulse logic to use it as a reference to modify the threshold?

Maybe I'm asking questions that just seem obvious to everyone else!

[nickel_n]

The thread only goes up to #55 - is it possible you were looking at detectormods' post count rather than post #?
Maybe I'm asking questions that just seem obvious to everyone else!

yes i was referring to detectormod's post count #78

[fo0bar]

I've now received the F1 and carefully removed the paint to expose some of the tracks and lands next to the EEPROM. I've also tested around the place with a meter to check continuities.

Looking at a photo from geotech of the eeprom, there's 6 little lands. I have a feeling on some revisions there may be a couple of 0R resistors there covering two traces, which could be what are being referred to in the post you reference above. I'm working on a REV3.2 board.

Firstly a little disclaimer - I'm sharing this information as there seems to be a lack of it available. It seems in the past there was a bunch of secrecy surrounding these mods but I've emailed detectormods asking for info and he's said he avoids working on them now, and I can't find anyone else still active on forums that's actively modding these to ask, so I'm assuming no one's really trying to rake a profit off these mods anymore. Nonetheless, if anyone loses their shirt over this info being public, just let me know and I'll be happy to remove it. I'm not out to rustle any feathers.

Anyway here's my findings:

- The top two are connected to VCC.
- The left middle goes up to the EEPROM. The right middle goes down under the IC below the EEPROM.
- Both the left and right middle are connected to VCC via traces to the lands above them.
- The bottom two are connected to GND.

Here's a super professional drawing based on the one I found here at geotech:


This is leading me to believe that I'll need to:
1. Cut the traces which are connecting the middle two to their respective top lands (shown in green)
2. Switch the middle two between their corresponding top/bottom (top == high & bottom == low)

If I'm correct, this will give 4 possible combinations of timings ....A level 3 modded F1A4 pictured at https://www.prospectingaustralia.com...c.php?id=18986 shows 4 levels of timing (big/gen/std/fine), but I'm unsure of what the other knob does. Most other modded machines show 3 levels of timings so maybe I've got something wrong here... At some point I'll see what my scope tells me but I might have to wait for my logic analyser to arrive before being able to see the big picture.

I'm probably going to wait a little while before testing this theory...Planning to take the machine out this weekend and would really like to get at least one use out of it before potentially releasing the magic smoke.

[fo0bar]

After doing more research it appears the gain, threshold and tone mods are going to be pretty simple. I'm referencing the SD2000 circuit as a very rough guide as some of the sections appear quite similar, comparing pin connections with datasheets of components.

I've also discovered the F1 has two-channel RS232 output which sounds logical considering the F1A4 is MCU controlled. It would be nice to connect this to a duino or raspie to log the serial data along with GPS location, and create a corresponding heatmap of the ground and target response (if those are the two channels supplied). Hopefully I'll find a not-too-hacky way to interface with it. Would also be very interesting to look at offloading the signal processing to another MCU or DSP to experiment with alternate filtering methods. I expect I'll be limited to the time domain but should be able to interpolate other data from it.

The F1 may be a far more upgradeable than I originally thought. This could be a deep rabbithole.

Will update here as I progress with research and modifications.

[eclipse]

Post some pictures of the board. Is it closer to SD2200, or SD2100 perhaps.

[fo0bar]

Post some pictures of the board. Is it closer to SD2200, or SD2100 perhaps.

Apparently these are indeed more closely related to the 2200, however the circuits do seem to follow some similar design principals to the pulse & frontend sections of the SD2000 circuit available here on GT - such as the use of blocking fets, dumping diodes, paralleled damping resistors, parallel fet pulse output etc. Of course there are a bunch of differences though. The issue here is that I just don't have a SD2200 circuit to follow

Images attached to this post somehow.. I'm a forum noob and don't know what I did there

[eclipse]

Yes it looks very similar to SD2200. The lack of 4th channel is obvious.
TX section and timings look the same but somewhat less component count on what on the SD2200 can be seen.
If you're up to modifications maybe look into reading and modifying the EEPROM chip on board 1, but I guess that's a bit hard (modification).
Other thing you could do, easier, is to find modern replacement with lower Rds and Coss on the pair of FETs, there must be a modern improved upgrade by now.

[fo0bar]

If you're up to modifications maybe look into reading and modifying the EEPROM chip on board 1, but I guess that's a bit hard (modification).
Other thing you could do, easier, is to find modern replacement with lower Rds and Coss on the pair of FETs, there must be a modern improved upgrade by now.

I'm up for all modifications
Yep, I've been reading into dumping the contents of the EEPROM - I've already ordered a couple of the EEPROMs, as well as these to adapt to my programmer:
https://www.ebay.com.au/itm/Plcc32-to-dip32-programmer-adapter-ic-socket-converter-module-O/283290765127

and then one of these will go back on the board instead of the EEPROM, allowing me to quickly interchange them without having to use hot air:
https://www.ebay.com.au/itm/10pcs-New-PLCC32-32-Pin-32Pin-SMD-IC-Socket-Adapter-PLCC-Converter-J-Hg/153358585640

I'll certainly be game to change the mosfets once I've tested out the mods I've mentioned above - perhaps something logic level would be ideal to make them react a bit quicker. IRLZ44N looks to be ideal as it's got a Id/Idm just exceeding that of what's on the SD2000 circuit i'm looking over, a Coss lower by around 150pF, Rds of about half and much lower Vgs. Of course if the circuit I've got uses differently specced mosfets to the F1A4 them my comparison is bunk but that won't stop me giving it a crack anyway!

Any other suggestions? I'm also looking into what I can do with the MCU. It appears to be a PIC, but I haven't spent much time examining that yet. Wondering if anyone can shed some light before I go reinventing the wheel.

[eclipse]

I used TL866II Plus (http://www.autoelectric.cn/en/Index.html) for reading the EEPROM, take a look if yours didn't work, it's real bargain for what it can do.
IRLZ44N is not suitable at all. The Coss parameter is way too high. Look for lowest Rds and Coss at no more than 25-35pF is my suggestion.
Are you sure about the make of the large MCU on board 2? I would bet it's Motorola, similar to what they used on the Sovereigns (MC68HC705).
There isn't too much stuff you can do with it anyway.

[fo0bar]

Ok great - will look at that programmer if mine doesn't want to play ball. Thanks for the heads up.

According to the datasheets I'm looking at, the IRLZ44n has a lower Coss than the original (400pF vs min 560pF)...? They're both MUCH higher than 25pF... I'm basing this off the assumption that it is, or is similar to the BUK455-60A as seen in the SD2000 circuit. Again, I know there will be differences, but this is all I've got.
http://www.e-ele.net/DataSheet/BUK455-60.pdf
http://www.irf.com/product-info/data...ta/irlz44n.pdf

Seems all the high power mosfets I've worked with or have stock of come with a Coss that FAR FAR FAR exceeds the 25pF you mention - any suggestions of a manufacturer or series I should be looking at? If i'm wrong and the SD2000 circuit i'm referencing uses completely different mosfets, do you know what these ones are so I can compare more accurately? I'm really trying to avoid scraping paint when I can get the information elsewhere

And you're right - the MCU is definitely Motorola. On a closer look I can see the telltale 'm' logo through the paint on one part of it. As I'm sure it wouldn't have any kind of super-special 16-bit ADC or other special capabilities, the MCU is one of the last things on my agenda.

[eclipse]

The TX section mosfets should be a pair of IRF640 or similar.
I'm talking about the blocking mosfets on the RX side. They should be ZVP4105 (maybe the bottom) or similar and VN2410L (maybe the top one) or similar P & N channel.
For example ZVP2106A could be upgrade - has lower Rds but higher Coss. Check out St.com, mouser and digikey for better alternatives and make sure those 2 are blocking mosfets. The SD2200 has ZVP4105 and 2SK216 - both old and obsolete.