no...it will not run at 8 no matter what you do.
it will destroy the card if you persist.
Hello Fellow Overclockers,
Would be interested to know before attacking the jaycar if it is possible to run a blue in the jaycar with newcs at 8MHz. I have heard the other programmers can do but have not found any info that a jaycar can . What worries me is the ability of newcs (in linux) to force a non standard baud rate into the serial port . All information welcome .
Best Regards
RD
Look Here -> |
no...it will not run at 8 no matter what you do.
it will destroy the card if you persist.
<type>phoenix</type>
<mhz>800</mhz>
[ 2022 ] [ 090B ] Card ECM took: 0s 192902us
[ 2022 ] [ 090B ] Card ECM took: 0s 194732us
[ 2022 ] [ 090B ] Card ECM took: 0s 191731us
jaycar reader with 8mhz xtal on a sunix pci serial card
been working great for nearly 12 months
Thanks Guys,
Much appreciate your input . So it looks like one vote for yes and one vote for no . I Guess from what it looks like I have 50-50 chance of getting it working .
Best Regards
RD
There is virtually no chance of killing the card by overclocking it. Increasing the voltage above it's nominal level will kill it, but merely increasing the clock rate will just reduce the pulse duration to the point where the peak high and low voltages are insufficient for the processor to function and it will either halt or reset until power is removed again and a lower clock is applied.
At 8MHz the card wouldn't be 100% reliable, you may have an occasional glitch that would cause a reset, but if you really must have an enormous number of clients polling the card then that is a risk you may find acceptable.....
At higher clock frequencies the inherent capacitance in the wiring and circuit board track layout from the oscillator can distort the clock pulses so one card reader may perform better than another.
Hi davmel,
I understand what you are saying . My motivation was more to do with lowering the latency , than catering for more users . I notice that the largest delays are to be seen while the data is going in and out of the card . So if we can lower that then it gives more margin to the ip path , which sometimes is too close to the bone .
So I was wandering just how much a blue can be pushed (reliably), I do understand that spongebobs run happily at even 10 MHz .
Much appreciate your response.
RD
Spongebobs are designed to run at 6MHz and tend to work fine at 8MHz or even 10MHz. NDS cards however are designed for 3.58MHz, which is quite a bit slower. I don't have much experience with NDS so can't advise on specifics, but Spongebob and NDS Blue are different beasts, like comparing apples and oranges.
I haven't tested NDS blue card myself but others will be able to tell you whether NDS blue works at 8MHz or not. Try it first at 6MHz, if it doesn't work overclocked at 6MHz there's no way it will cope with 8.
In practice you'd be unlucky to destroy a card by overclocking, and if it did die the reason was probably excess heat. In general, when you overclock, temperature starts becoming an issue and reliability degrades as a consequence, which is why overclockers are into cooling techniques.
There are no issues with Jaycar Mk I or Mk II at 8MHz. You can just desolder either crystal, replace it with an 8MHz one, and off you go. Provided your soldering is correct, any problems you have will be due to the smartcard or the PC serial port or its driver - not the interface.
Orange cards wouldn't even start at 8MHz. Even at 6MHz they reset every few minutes etc so I wouldn't go above 5MHz if you want reliability.
Those speeds are at the nominal 3.3V that you're meant to supply the card. If you want to overclock and make it work reliably at higher speeds you need to increase the supply voltage (just like PC processors or any other processor), however to do that you need to consider the thermal issues so you would need some way of adding a heatsink by scraping away the plastic or active cooling with a fan if you push up the voltage significantly.
The Irdeto cards can clock higher since they have a higher nominal 5V operating voltage. The higher the voltage the easier it is to detect the trailing and leading edge of the pulses at higher frequencies.
The spec requires cards to support 1-5MHz, to tolerate 5V at least for startup, and to specify in their ATR if they require a lower voltage. Apart from mobile phones 98% of the smartcard hosts on the planet are 5V-only so I wouldn't get hung up on card voltage. I would definitely NOT be increasing the card voltage above 5V, no matter how fast you're overclocking. There are very few microprocessors in existence that tolerate supply voltages above 6.5V for any length of time, and that's really pushing it.
That's the beauty for N*S in that they can make their cards with any spec they like as they only allow their technology to work in integrated solutions and don't have to worry about 3rd party compatibility with industry standards.
If you're using an orange card you should be very hung up about card voltage unless you like asking your provider to keep replacing your cards.
That's why I said you need some way of dissipating the heat. If the junction temp of the silcon goes higher than it's maximum limit (anywhere from 90-150degC) then it will be irreversibly fried. Increasing the voltage above spec is fine if you can remove the heat so the temp doesn't exceed maximum levels.I would definitely NOT be increasing the card voltage above 5V, no matter how fast you're overclocking. There are very few microprocessors in existence that tolerate supply voltages above 6V for any length of time.
Of course that is very difficult to achieve with a processor surrounded by plastic with very poor thermal conduction properties.
I haven't tried putting the card in a freezer or liquid nitrogen but if you have money to burn you can see just how high it will overclock ;-)
I concur.
They fry very easily, annoyingly so....
Testing the orange on a jaycar 6mhz didn't work, it kept resetting the card.
3.58Mhz works perfect with ecm times around 0s 330000us.
Once saw a message when trying to overclock it saying not to overclock it over 5Mhz.
Anyone know if you can replace the 6Mhz crystal with a 5Mhz one on the jaycar and have the orange work?
I've also tried it in a smargo smartreader+, but that will only work in phoenix mode using newcs, even with the setting at 780Mhz doesn't do anything, card still responds at around 430000us.
The problem with newcs and smartreader is that newcs has only 2 frequencies in common with the smargo smartreader, 6Mhz and 8Mhz which don't work in smaratreader mode with the orange. Trying 5Mhz is ignored and it tries to use something like 3.58 in newcs but the smargo doesn't recognise 3.58 only 3.69 or 3.43 which don't work with the orange. Wonder if the newcs mob will ever fix this???
In the scsi reader in the DM500 the orange runs at 7.8Mhz with ecm's around 220000us, not sure if that is true 7.8 or its running ar around 5.
Anyone have better luck with apples and oranges?
As you discovered the smartreader will only run in phoenix mode with NewCS. I tested a blue, and in phoenix mode it ran ok at both 6mhz and 8mhz although nothing too spectacular.
According to newcs dev site, these frequencies are supported by newcs.
Supported frequencies for phoenix readers:
1600
1575
1431
1430
1230
1200
1071
1070
1050
1000
820
800
787
720
715
686
630
600
536
525
450
425
400
393
368
358
350
320
315
153
The real problem with NewCS and the smargo is it can't run in smartmode as it can with CCcam. The good news is that if all goes well with testing the smargos will work in smartmode with the next NewCS release.
Bookmarks