MINE_LAB -F1A4 -A/OK- IMOP..

[Kev]

Hi All,
I just purchased one and intend applying a few mods, but need to wait until it arrives.
I was wondering what is the standard Xtal for these?
I see some say 1.9MHz some 2MHz others 4MHz.

Minelab have two versions of the F1A4, one with a small coil, high frequency and timings designed for Minimal Metal mines and an UXO version supplied with large coil, lower frequency and timings to suit deep unexploded ordinance.
I guess this would explain why some have the 4MHz Xtal and others the 1.9MHz (not sure about the 2MHz, maybe for close proximity to ones running 1.9MHz)

Woody I wonder if the firmware is the same for each type the MMD and UXO and they select appropriate timings with PROM strappings?

Independent Mine detector comparisons and testing showed that the UXO version had the deepest reliable detection ranges in hot ground of all the detectors then available.
This is an air test, but in an in ground test the A1F4s were the only ones besides the Vallons that didn't loose appreciable depth. This is an aluminium target one that is a good gold analog.

(Source Systematic Test & Evaluation of Metal Detectors (STEMD) Interim Report Laboratory Tests Italy EUR 22536 EN 2006)

Swapping the coils between the MMD and UXO may not have produced the same results since Minelab confirm that the MMD and UXO versions have different timings.

Cheers
Kev.

[detectormods]

Funny about Aluminium being an analogue for Gold, yes that is true to a large degree, I think either next months Gold Gem and Treasure magazine or the month after has an article i wrote about people using the wrong materials for testing in mineralized ground.
As for the F1A4 ,I would not muck around with the crystal frequency, it just does general housekeeping tasks. If you are game, copy the code (with the permission of Minelab of course) from the EEprom in a SD2200 unit and burn a new chip and stick in a F1A4...It would be good if they used sockets. These detectors use standard coils of half an ohm , 300 uh , just play with it, but be careful as all the ic numbers are ground off and it is a ***** to repair if something goes pop.

[Kev]

VLFs don't register pull tab bits as gold for nothing. Perhaps in Oz you may have a point aluminum being slightly magnetic, but as for medium ground it is far closer than lead or brass in my experience.

Mario did an excellent job applying DDS injection to his 2100 I suspect it will work on this unit as well, no harm in trying, unless the processor is an early 68hcx and I go above 4MHz.

I don't have access to any 2200 code and have no idea which resistors you are talking about, I also hope Barry doesn't follow your suggestion above, you have me wondering immensely.

[hoadlies]

Good to see you got the goods Barry, some (most)of the work on ML detectors is tricky, one mistake and it can bring tears to your eyes..if something goes wrong it can be a real pain to rectify. I just remembered that posting how to do's with electronics entices some people to blow up their detector and thus you can get married to other peoples problems..I have spent many countless hours of repair by email and its not a good idea. Im sure that you could do some of the changes but maybe you can send it here and i can do all of the difficult parts of the operation for you.

The forum is going back up, the website has been upgraded (modded) to a new design... just have to implement all the new changes.

HI, DETECTORMODS I fully understand your concern with posting pics & tips for mods mentioned above as to the delicacy of the Mine-Lab circuits & the kick-back to you if owners kill there detector through bad work etc etc..
I myself would take the risk &also cop in on the chin if I nuked mine , But some may not !! any way you have to limit complications in your line of work because its your living so NO-WORRIES -MATE Do as you see Fit..

A question could you explain the Battery set-up in the F1A4 ,as it seems there is a module [in Batt -case] that keeps the voltage at working level [adjusts as batteries run-down ] to get Max life from the 4xD'cells..

The reason I ask this is I was hoping to use 2x 3.2v Lion cells ??? I have some spare of these New ...Much lighter @6ooo,Mah -EA..

CHEERS,,,,,,,,,,,,,,,,,,,,BARRY..

[sinclairuser]

Funny about Aluminium being an analogue for Gold, yes that is true to a large degree, I think either next months Gold Gem and Treasure magazine or the month after has an article i wrote about people using the wrong materials for testing in mineralized ground.
As for the F1A4 ,I would not muck around with the crystal frequency, it just does general housekeeping tasks. If you are game, copy the code (with the permission of Minelab of course) from the EEprom in a SD2200 unit and burn a new chip and stick in a F1A4...It would be good if they used sockets. These detectors use standard coils of half an ohm , 300 uh , just play with it, but be careful as all the ic numbers are ground off and it is a ***** to repair if something goes pop.

hi all, before i packed it in i used to do firmware mods to game consoles, and i was wondering if some of the tried and tested methods could be used here, if this has a chip that is hard soldered there are female clip on chip converters, some use identical ones some change the chip to a different package.
you could use one of these if it had a small breakout circuit on it with the appropriate crystal it could be simply clipped on, all you do then is cut the power pin to the original chip and using a switch or if you are really clever a relay, switch between the two, the standard setup is unaffected too since unclipping the chip clip and re-bridging the power pin the board is as standard, if you search for dreamcast bios mod you will get the idea.
the benefit of this is the original setup runs on the onboard chip and at the flick of switch the sd2200 firmware can be run, indeed any custom firmware can be loaded on this chip, these kind of adapter boards are widely available on the net in fact a lot are made and sold from australia.
this method is perfectly suited to this application since the only "mod" to the host board is the power pin cut, woody look into it mate, or indeed any one all the bits needed are out there, this method or versions of are used on loads of different hardware, you will find a chip clip that fits and either an alternative programmable chip or even the same chip if you look.
the hardest part is programming the chip.
its worth looking into, especially for woody as he could sell the chip clip and breakout board un programmed in kit form making chip swap easy,with no infringement problems, and because you dont have to use a switch, you can just cut power pin to host chip and connect it to the new bios, leaving the old one untouched.
i have never had any problems with this method its simple and easy, granted in a different hardware environment but it may just work here.
p.s. of course you would have to clean the paint off the bios pins, but that's easier than cleaning more and removing the chip.

[Kev]

Good idea sinclairuser if you have access to the code as Woody obviously has.

When I get my unit I will need to do some probing and see if my intentions are viable.

If I can entrain the system heartbeat I plan on digitally synthesizing the clock which will enable all sorts of tricks to be employed, such as clock suspension during the TX pulse, or samples, up frequency shifts during sample delays etc., My fear is that the logic generates its own clock, I have access to a LA but I don't really want to re-engineer the whole kit and kaboodle, injecting a clock would be so simple.

Are you having a good laugh there Peter? You know how it functions but I don't yet, I'm determined though to beat this sows ear into a cheap super gold mop, by pushing Bruce's stability limits.

To wind up the gain it's imperative that we have fine control over the main frequency so that any beating and intermodulation distortion can be folded away from the audio circuits.
The same micro could intercept the signal post mixing and subject it to filtering such as Kalman and then generate a separate audio indication.

I'm super busy and don't have time to build up a whole new detector from scratch, this unit will make an excellent experimentation platform. At least until the GPX-6000 is released, and then I'll be too busy with other things.

If I find this is workable I'll start a new thread.

Cheers
Kev.

[Kev]

A question could you explain the Battery set-up in the F1A4 ,as it seems there is a module [in Batt -case] that keeps the voltage at working level [adjusts as batteries run-down ] to get Max life from the 4xD'cells..

The reason I ask this is I was hoping to use 2x 3.2v Lion cells ??? I have some spare of these New ...Much lighter @6ooo,Mah -EA..

CHEERS,,,,,,,,,,,,,,,,,,,,BARRY..

Hi Barry,
It's good the ticker is behaving itself, metal detecting is damn good exercise for the heart, especially if you're up and down a lot digging targets.
Also doing your own homebrew detector at 75 is really inspiring as well.

Since Peter hasn't replied I thought I'd add what I know, which is very little as far as the F1A4 is concerned, but I do know a little about batteries, and I'll be looking into this for myself anyway.

The rechargables recommended for the F1A4 are NiMH and they need to have a cutoff point in their discharge, with NiCads it was 1.1 Volts per cell, I suspect NiMH will be similar.
So as the batteries run down the detector will start the low battery warning once the pack approaches 4.4 Volts, and then stops you from discharging them any further.

I suspect the batteries you have are Li-Polymer being 3.2V, if they are, then the absolute lowest discharge voltage is 2V, so it would appear they are suitable in this respect.
Fully charged they will be about 3.65V. Making it 7.3V when combined. A fully charged NiMH set will be in the region of 5.8V, and the 6V lead acid will be 6.6V
I don't know what the maximum input voltage is, but the Lithium cells are three quarters of a volt higher than the lead acid battery.

It maybe OK I'm not totally sure but if you used a diode (1N4004 etc, not Schottky) in series with the battery lead it will drop that 0.7 volts, and make it look like a lead acid is connected
You'll not need to worry about getting full capacity out of your lithiums, since the discharge curve is fairly flat with lithiums until it reaches 3V, some most of your juice is gone by 3V anyway.
It could be even better to use the diode, as it gives a little bit more discharge protection.

Of course if they are actually Li-ion cells you have (usually 3.7 Volts) the above is invalid.

Sorry Barry I can't be more specific, if I had my unit I'd be able to look at the input circuit and be more helpful.

Don't give up looking for the yellow, it gets easier with persistence.

Cheers
Kev.

P.S.
Thinking about it again, I doubt highly that Minelab would not have accommodated for that extra 0.7 V and that diodes going to burn quite a lot of your juice.
I think you should be OK without the diode and just connecting her up, BUT hell I wish someone whose actually fiddled with a F1A4 would respond, I don't want to be responsible for charring your machine, maybe that's why no one else has responded?

[Kev]

cut the tracks under them to allow access to the address lines of the eeprom,

Would that be A10 and A11 by any chance?

[sinclairuser]

you will need to apply power to the chip, unless you a using board power, but i have had more success in the past when powering chips independently, sometimes you dont want the host board to know you are reflashing its chip, however since this unit is ex military i would expect this particular chip to be flashable, since repairs and servicing is done by the military usually, in fact i'm surprised this model has the white paint if i'm honest.

[Kev]

The level 2 service manual Lists complete PCBA Swap outs for minor Malfunctions,
Seems Peter is correct about processor being passive with regard to logic functions.
Question is how will it respond to the PROM being molested?

[hoadlies]

HI ALL ,, THANKS KEV & S/user,for the above info,,,,, S/User I am sorry to say I am a bit lost when it comes to programming chips etc the Digital world is beyond my knowledge ..

KEV ,YES these lion batteries are 3,2v 6000-Mah BUT I will not apply them until I am sure of the f1A4's battery set-up ,But I take onboard your explanations Thanks..

I have bought as well ,,4xdcell- Nimi rechargables & will use those untill I find-out more info..I have a Bungy-harness to relieve the weight works well..
IF you look in the service manual it mentions programming port I think its the power connector plug combination ,obviously M/L can alter Data with special programming equippment ????????

CHEERS ,,,,BARRY..

[sinclairuser]

actually barry with the right sourced parts and a decent tutorial anyone can do this kind of mod, after all pre-pubescent teens with no electrical experience do these type of mods every day.
the hard part is the first couple, working it all out and writing a step by step guide, you could do this in a way that left the original chip untouched, the danger is bricking the chip, thats why i favour designing a snap-in board that is programmed out of the unit, then added in a way that leaves the original bios/flash/eeprom intact.
p.s. consoles have allsorts of encrypted keys, and such, but we dont need to see the code, we dont need to inject modified boot loaders, all we need is to dump the eeprom/flash/ whatever from a 2xxx version, and put it on another chip then connect that new chip so the hardware boots it and not the on board one.
the reading and reflashing is the easy bit, its playing or decrypting the code thats hard, which we should not need.

[Kev]

My unit hasn't arrived yet but the more I search here and elsewhere for info regarding these machines the trickier this all becomes.
I didn't believe it was necessary to mess with the PROM other than manipulating the clock that feeds the counter which drives the PROM sequencer.
However I don't yet understand how the auto-tune and ground balance fit into this, they obviously are controlled by the MPU. I thought that the MPU would've controlled the sequencer, but I read that changing it's frequency doesn't affect timings. I guess I must patiently wait until I can get inside it.

Woody a question, now I realise this is going to be rhetorical since you hardly ever return and qualify the statements you have made, above is a case in point.
Does your F1A4 level 2 mod munt the auto-tune?

Another couple Woody.
You saythe F1A4 is more akin to the SDC than the SD/GPs, is this because of better EMI immunity in the F1A4? It's certainly not sensitivity as a 0.1 gram had to be coil scrubbed to get a response in one vid.
If so, is the MPU signal processing responsible for this better immunity, or is it due to other factors?

In Jack's F1A4/3500 comparison vid as far as I could tell the F1A4 with level 2 mod had far better stability and better range on targets.

Chow

[sinclairuser]

hi kev, i mentioned the firmware angle since it was implied(here and elsewhere) that it may run 2xxx code, and since at one time pulling firmware and re flashing was my thing, i thought i would explain that its not as hard as some think.
i do agree however it should not be needed, some of these units may need a crystal change(though i have a feeling that will be done before resale), and extra controls may be helpfull.
it may be the case that these units are altered so at the end of the day no tweaking can be done unless you have a working knowledge of minelab stuff, and crucially correct schematics.

[Kev]

Thanks sinclairuser I do appreciate your input and it may be a route to take if simpler ways fail.

There are supposedly units out there with frequency and timing mods, perhaps the PROM needs to be flashed with 2200 code to get the hidden timings?

Woody so often drops clues and hints on a subject, an ego thing I think because he very rarely colours in the outline, just leaves it with no substance.

My machine should arrive next week, I'd like to see if fragments of the MPU part # remain, things will be much clearer then.

Cheers