Tip when building Jaycar Phoenix Mk II

[gw1]

Recent batches of Jaycar Smartcard Reader Mk II (KC5361) have extra holes drilled beneath regular REG1 and transistors Q1 & Q2. Jaycar have presumably done this to try and help because some parts have legs with different spacing.



However they've totally cocked it up! They've drilled holes which are virtually the same size as the tracks they're drilling into - cutting the tracks in half in some cases!



In the example shown above the middle leg of REG1 was cut off from ground, meaning it couldn't regulate at all. Consequently its output which should have been at 5V was actually floating up over 9V! The gods were smiling because the three ICs survived, but that was a matter of luck more than anything else. With the break in the track soldered over it all sprang to life as it should.

Someone at Jaycar ought to have their arse kicked over this, because it's actually fairly hard to solder over the holes without creating a short circuit - especially on REG1. If your soldering iron tip (or eyes) aren't up to the task you can work around it by soldering an extra link as shown below.



With Q1 and Q2 it's not quite so tricky. You just need to choose the right hole.

[hillbilly]

Good work gw1

there should be a Directory dedicated to the Jaycar programmer, seems there's a lot of good info spread out over the forum, and its a pretty popular item.

[fandtm666]

also they should take note of the extra link that is not present in previous versions

[bulbous]

@fandtm666 :-

What's the deal with the spare leg (?) on the uppermost transistor/regulator/device in your picture ?

Not denigrating your work, merely asking why the leg seems to go nowhere !

[fandtm666]

is one that i had to fix for some one , there were a couple things that were over looked when it was being built by some one new to kits.But apart from
those little things he did a pretty decent job of it.

[Frank Drebin]

another top post gw1 - loving your handiwork as always!

[sublib25]

@fandtm666 :-

What's the deal with the spare leg (?) on the uppermost transistor/regulator/device in your picture ?

Not denigrating your work, merely asking why the leg seems to go nowhere !

Tobacco??

(spelling nazi's check please)

[Duc998]

Wow what would we do without u guys.

Gonna check my recently built MKII when I get home to tongit.

Cheers

[OSIRUS]

Thanks, gw1 & fandtm666,

really usefull information,

You should be on the Jaycar payroll,

they owe you heaps providing first class support for their product,

But we thank you too

[Goose]

Thanks for that

[msay]

Hello guys, my first post here.

I want to build this MKII, since i don't have the correct board size and such i will do it from scratch using Eagle, so far i've read this two "hacks":



So to avoid do overs, can I do all this changes? including the mk2mod.jpg on th IC3?


In my country the 2n3906/2n3904 are not easy to find, can i use a standard BC557(BC558)/BC546(BC547) instead?

Last question, can I use a CD4053 to switch between PH/JDM?

Thanks.

[gw1]

Hi msay, welcome to the forum.

You should be able to order a kit from Jaycar (or their parent company Soanar - search their website for 'KC5361'). I'm sure they'll post overseas (if not I'm happy to buy one for you if you can use paypal). Buying a kit would be easier and probably cheaper than redoing it using Eagle, though not as educative.

If redoing it I would definitely add the inverter modification (perhaps add a jumper link to select between invert/non-invert). Also I would replace the 4K7 with a schottky diode (IN5819, BAT85, BAT54 or whatever is available with low forward voltage). Also link DB9 socket pins 6 & 9 to GND. And connect a bypass capacitor across the smartcard VCC/GND pins (preferably an X7R one but anything is better than nothing at all). There are a bunch of pages about these modifications - others might be able to dig up links, otherwise keep searching.

Yes you can certainly use BC548/BC558 or similar general purpose transistors, the circuit isn't fussy. Naturally you need to use a PNP for Q1 (2N3906) and an NPN for Q2 (2N3904). But the most important thing is to insert component leads in the correct position: the component overlay shows where the collector, base and emitter need to go. Looking onto front of the packages, the 2N390x transistors use E-B-C pin order but BC5x7/BC5x8 transistors use C-B-E. So basically you need to insert the BC5xx transistors backwards.

And finally, can you use a CD4053? Interesting question. I think the answer is yes. You would need to power it (Vdd pin) from the +13V point (zener cathode) but apart from that I think it should work: none of the currents are high and the switch impedance shouldn't be an issue. The wiring would be messy and you should try to keep the leads for S1a (clock) as short as possible. If you give it a try take some photos and let us know your experience.

good luck
-gw

[msay]

Hello gw1!

Glad to hear from you, i would like to take this opportunity to thank you for all your efforts.

No need to worry about the kit, but i thank you for that also.

I'm not an expert but i'm doing good by myself for the past 20 and some years, i can redesign the circuit, add the changes and build the board at home, last night i've just finished the library design for the smart card adapter that i've got from , looks like all other adapters follow the same design and measurements.


My concern about the CD4053 is that it can work with 13V(15V max) but i want to save that voltage and current load for the SC, so i intent to feed the CD4053 with 5V, in this condition the impedance of his internal switches reach 270 Ohms according to his datasheet, in this case i'm thinking on reduce the 470 Ohm resistors attached to this switch (PH/JDM) to about 200 or 220 Ohms so we can have a match and play safe.

When i finish the design i will post it here, to see if it meet with the changes and then design the board.

One thing i know for sure is that i don't know nothing about programming a smart card, i have two STB at home for tree years now that the company doesn't want it to take it back and told me to trash then. In that case why not try to do something with it, one thing that i know bout this STB is that it uses a gray/silver nagra cards (i think is nagra 2) with MDS antenna.

Well, i will start the design, you will hear from me soon.

Thank you gw1.

[msay]

Ok gw1, schematic is up, anyone can download it from here:
File removed.

Please let me know if there is any important "hacks" to consider before i start to design the board.

I've added:
- GAMMA & DELTA modifications.
- One Tactile Switch to control PH/JDM function.
- Led's to indicate PH/JDM function.
- Inverter mod that can be activated by jumper.
- D-SUB9 pins 6, 9 and 5 connected to ground.
- Added suggested capacitors to Smart Card VCC.
- Added shottky mod with 1N5819.
- Every IC has now a 100nF bypass/decoupling capacitor for better stability.

Thanks.

[gw1]

A few suggestions.

• You absolutely must power IC4 (4053) from the +13V point! If you only use +5V then the high voltage your circuit applies through Q1 to 4053 pin 3 will destroy the device.
• Consequently you'll need to ensure IC4's SWCLK input has full rail swing. One way to do this is to power the 555 from +13V too (which means you should consider a 7555 to reduce drain on that point). Alternatively you could feed 555 pin 3 into a common-emitter NPN (with say 10K base resistor) and use the collector to drive SWCLK with a 10K pullup to +13V. The latter approach involves more components but avoids loading the +13V rail.
• Assuming you have SWCLK swinging up to +13V you will also need to power IC5 from +13V. (There are alternatives of course but you get my point.)
• Although I previously suggested a tantalum for C9 (across the smartcard) I would now suggest an X7R or X5R multilayer ceramic around 470nF as its characteristics are better suited. A tantalum will still work though and is better than nothing. If you can get an X7R ceramic for C10 that's preferable too.
• If you can locate 74HCU04 then use it in preference to HC04 for IC1 as it gives better oscillator stability. The HC04 will be fine though if you can't. (Don't use 7404.)
• You have an error on IC1 pin 6, it needs to be left unconnected (obviously)
• Your INV-MOD jumper link option needs to come after IC2f, not before. The way you have drawn the circuit is wrong because when link is in non-invert position the input of IC2f floats dangerously, and even worse its output is short circuited to ground by the card switch. The jumper arrangement used for phoenix/smartmouse selection shows the correct way.
• You should tie pins 1,2,15 of IC5 to either Vdd or GND. Don't leave them floating.
• The Jaycar Mk II has a bunch of 470R resistors (six of them) protecting logic inputs from static discharge which may come in on the smartcard socket pins. The resistors don't provide bulletproof protection but they're a good idea. Just don't make them too large, especially not on the I/O line. Your circuit has retained some of them but they do no good behind the 4053 switches. Where you need them is between the card socket lines and the 4053s, to protect the 40532. Once you've put them there you don't need them anywhere else. (Put R10 on the 4053 side of the current limiting resistor, not the C7 side.)
• Your 220uF on the 7805 output is unusually large, causing unnecessary high surge at powerup and stressing the regulator's internal protection diodes after power is removed. No doubt the regulator would cope, especially if you use 7805 rather than 78L05, but it's overkill. A good quality 22-47uF should be plenty.
• Beware the pinout of 2.5mm connector footprint. Some manufacturers have different numbering (especially for pins 2 & 3) so check your library footprint is what you expect. Also, double check its hole sizes as it's frequently misspecified.

There are probably a couple of other things I've missed but that'll do for now.

Your 555 flip flop circuit is nice, by the way. I'm surprised I haven't seen that configuration before. You may not need R24 & R25 (the 555's internal resistances are only around 5K to each rail if I remember correctly) and I doubt you need C23 in this case either.

[msay]

Hello gw1, thank you for the suggestions.
I hope i got it right this time, here is the rev.02 of the schematic:


I've removed the other one because is wrong anyway.

I did not understand the issue about the connector footprint, i don't have anything in my library to compare to or to take reference, the only reference i had was the available images of the board to identify the pins, and the product datasheet.

I can see the connectors for the card and they are all straight, there are no swaped pins, the drill size is set to 1.016mm and the pad is 1.9304mm, all the adapter body size, spacing between pads etc, was followed precisely, here is the datasheet used for .

And to compare.

That 555 circuit is a very old one, no my idea by the way, from 70's books about 555, and you can see some variation of the same circuit on some magazines today, the internet etc.
Here is one with details:


Search for "Relay Toggle Circuit Using a 555 Timer ", it works the same but the design used here was modified to work in real situations, the original circuit might "buzz" the relay every time you press the switch and you doesn't know if the relay will be on or off, that also depends on the application that you are using it, it might fail or not.

The C23 define the initial state when the power is on, since is grounded the circuit start with Q low, if you connect C23 to VCC it will start high. I'v used this circuit on power amps, power supply, on/off switch for other devices and works like a charm.

Thanks.

[gw1]

I did not understand the issue about the connector footprint

If you compare datasheets for versus (both 2.5mm DC jacks) you'll see slight difference in required hole dimensions, and their pin numbering differs too (1-2-3 vs 1-3-2). So before manufacturing your design it pays to double check
- you've routed to the tip for positive and to the lead underneath the package (not the one on the side) for negative
- that your slot holes are big enough (be sure to observe the plated hole sizes supported by your PCB manufacturer to avoid surprises)

The C23 define the initial state when the power is on, since is grounded the circuit start with Q low, if you connect C23 to VCC it will start high.

I understand, nice trick. You should definitely arrange for it to power up in Phoenix mode (either by C23 or 4053 x/y pin selection) to avoid subjecting non-PIC smartcards to +13V.

I was wrong about the 10K resistors on pins 2 & 6 - they definitely are required. (The IC's internal ladder is on its comparators, not pins 2 & 6.)

Other suggestions

• For C9 I meant 470nF X7R ceramic, not 470uF.
• It's only IC1 that benefits from being HCU04 (if available), for the crystal oscillator amplifier. I would keep using HC04 for IC2.
• You no longer need R10.
• Since you're using transistor T2 there's no need to power 555 from +13V: use +5V instead.
• Don't forget to wire card socket VCC to +5V.

With a couple of extra transistors you can eliminate IC2 if you wish.


The above circuit should work but I can't guarantee it. Although it is fun to make your own designs, if you have never programmed smartcards at all before, you will reduce learning curve and frustration if you use trusted hardware initially (eg a kit or pre-built product). Attempting too much at once makes it harder to diagnose teething problems. You won't know whether your problem is error in hardware design or construction, software tool settings or usage, card program or data files, operating system compatibility problem, etc.

good luck!

UPDATE:
There's a bug when the 'reversed' jumper link position is selected. Can you see it? The base junction of the NPN transistor (whose collector drives the 'normal' jumper pin) clamps the smartcard detection switch to 0.7V. The current through the transistor is enough to switch it on, meaning the 'normal' option will work, but the clamped voltage in the 'reversed' position will be insufficient to drive the MAX232 input high. Replacing the NPN with a MOSFET should be enough to fix the problem. (Oversights like this occur frequently, hence my warning earlier about using trusted hardware.)

[msay]

Hi gw1!

I'm using this 2.5mm dc jack:


Don't worry about the PCB manufacturer because i can't use any on my country, there are a lot of PCB manufacturer here, but for mass production only. I do PCB boards myself, i use photographic(UV lights) and toner transfer process, the photographic process i only use for final designs, two side PCB's and when the design i made is for SMD. The toner process is done for fast and quick prototyping.

Thanks for the schematic, and i will follow you advise and stick with a interface that we know it works, this is my very first circuit to deal with communication, this is really not my area, i've never develop a device to interface with the computer.

So i will stick with this schematic:


Just to avoid any frustration.

About R10, i will put IC2D aside and use a transistor instead, some years ago i've add a led to my PS2 memory card with the information from . It was a disaster, some times saves got corrupted, other times memory card access error with no reason, etc.

And from that experience i knew that messing with I/O lines can cause problems, until i found that show the same mod but, it never gave me problems with the cards. Even if the levels are high i don't feel comfortable to feed a led with it.

About the X7R capacitor, i've searched digikey and they returned only electrolytic ones, in my design i use 5.8mm components for the capacitors, including that 10nF, X7R uses the same size or bigger? I've been looking for vishay, panasonic datasheets and they give a huge variation, what component size should i been aiming for?

In any case, since everything looks OK, i will start the board design.
Here is the rev.03 schematic with inor changes:


Thanks!

PS: I know that sound crazy but, since we are re-engineering this interface and if this works, why not add a function to deal with atmel cards also. I've a elektor magazine here that show a fun/jupter/purple/pink card reader that is even smaller than the pic based Silver/Gold reader.

[gw1]

Atmel/fun cards are much simpler to program than PIC, but by the time your circuit supports three programming conventions the component count and amount of wiring is silly. That's why virtually all multi-card / multi-mode programmers use microcontrollers rather than discrete logic. Doing the work in software doesn't mean the technical issues go away though, it just means they're discovered later in the development process where rework can be costly. I'll wager that every multi-programmer manufacturer underestimated wildly the amount of work needed to complete their firmware.

Whether you're designing hardware or software it usually pays to sleep on your blueprints for several days before getting out the acid or the compiler. Design faults are cheaper to fix if found early, and the majority of them can be found more efficiently using eyes, brain and imagination than by trial and error with a multimeter and CRO or debugger. Rapid prototyping tools are great during research to help uncover issues quickly, but there's a good reason you don't see many commercial products built using VB.

Wandering off the subject a bit, I'm amazed when people say "I let the compiler find my bugs for me". It's just so naive and careless. I sure hope such people never get jobs with airlines - I don't want my aeroplane revealing its bugs at runtime, thanks very much. That attitude (along with unrealistic schedules, bad design choices and lack of component-level contracts & testing) is responsible for the atrociously buggy software you see around the place. Google-copy-paste is not the way to build anything of quality. You can easily spot essays and newspaper articles written that way: they're poorly organised, don't address the subject properly and are hard to follow. Software written that way turns out the same.

As much as I love microprocessors, any time you can manage without them you should try to do so. The easiest way to make something flakey and unreliable is to give it needless power or complexity. There's a reason why most species have small brains, or no brains at all. It's a matter of taste as to what point you abandon discrete logic and use a processor. Most electronics magazine publishers (except for Elektor and Silicon Chip) have always tried to use discrete as much as possible, even at cost of size and hardware complexity, as it's more instructive and accessible to the average hobbyist. In my opinion the Jaycar Mk II is about as far as you want to go before introducing a microcontroller of some kind.

Re X7R capacitors, they're not critical - don't make a special purchase order just to get them. Other capacitor types will be fine for hobby use. It's just that X7R ceramics, where available to you, are a good choice as their are well suited to the noisy demands of smart cards. Smartcard controller IC application notes often recommend them. Stockists like Farnell have them: if you search au.farnell.com for "0.47uF X7R" you'll see a bunch of them from different manufacturers such as .

[msay]

Yes, i understand what you are saying, i wish i could program and work with microprocessor that would make my life a lot easier, the thing i like about 4053's is that it can make your life easy or a complete hell when routing the board. I always try to avoid doing double side PCB, one because not everyone can deal with it and they cost a little fortune over here, plus we doesn't know how this will end up, i hope for the best.

This is the board so far:



I will do it all over again, too much dead ends, is a good challenge for the mind to solve that mess, the good part is that i don't see the day go by.

About your off topic issue, don't get mad about it, it happens here too, we are in the process to move to DTV (Digital TV), we already switch to digital but some states are applying to the new system (Brazil, ISDB with H.264), there is this "" a software created 15 years ago to deal with interactive TV that is not ready yet, we don't develop anything but card box and plastic bag (to resume the issue), there is no big electronic company here, all electronics we buy overseas or build as a OEM kit on some factory over Manaus, all parts come some place else and we just build then.

Our Set Top Box for TVD are a joke, the people who make the firmware of this devices might do exactly what you said, they set the compiler on "Auto Pilot" and let it go because of the lack of stability, functions that doesn't work, system crash and the list is very long.

Believe me, I understand your frustration.

Thank you for the reply about the Atmel/fun cards.