How to run your gamma at a cool +3V - with a 'factory' mod - Page 3

**vbthanks** · 15-04-08, 11:20 AM

Happy to report the mod gave me a rock solid 3v and took all of 10 minutes.

I used a more stable ground pad off the edge connector. To confirm if it is ground - leave it plugged in to the mainboard and check resistance between the tuner case and the pad with a multimeter for 0 ohms.

After 1 hour - the card is not even warm.

Well done eyesee!

**under-** · 15-04-08, 02:41 PM

Picture of your completed mod would be good.

**under-** · 15-04-08, 02:49 PM

Check out the picture!

**eyesee** · 15-04-08, 03:57 PM

Quote:

Originally Posted by vbthanks

... I used a more stable ground pad off the edge connector...

.. there have been lots of questions about what is the BEST point to use for GND.

... from a stability point of view it would be a point that has no additional GND loops - for that its best to use the same point where the TDA device gets its GND - which is pin 14 (please check on the data sheet at beginning of this thread). I used an easy point of solder - rather than the pin 14 itself - this was the SMD capacitor - the 'end' which is closest to Pin 14 of the TDA device (see original photo) ...

Having said all of the above ... since the way the +3V operation mode of the card reader is governed by internal logic switching of the TDA device ... it’s probably not that critical AS LONG AS there is a healthy 0V (GND) on pin 3 to ensure that its enabled the +3V operating mode ...

If the pin is 'floating' (no connections to it - as could happen if it is cleanly broken) - then the internal circuitry of the TDA still stays at the +5V operating mode ... however this is not the correct mode of operation because the TDA device could 'latch-up' and not operate correctly ... NEVER leave pins floating - even though the internal 'pull-ups' do somehow clamp to Vcc - there is no guarantee and you are leaving the door open to potential transients...
... if you have indeed broken the pin and its ‘floating’ don’t worry too much its really unlikely that it will latch up – but it will stay in the +5V mode of operation ... unless you try to reconnect to pin 3 and bring it to a GND point (method described in a precious post in this thread) ...

Hope the above clarifies the questions of where is the best point to GND pin 3

**CoopsOz** · 15-04-08, 07:30 PM

Just a quick one, I was initially measuring across the "mystery" pin 5 and getting 1.8V, after doing some research I think I was measuring at the wrong spot, my pic will explain it easier. So, now I get 3.061v across C1 and C5. Is that correct? I used GW1's picture so disregard all the other components.....this is merely a 3rd leg lift mod.

**beer4life** · 15-04-08, 07:31 PM

Quote:

Originally Posted by eyesee

.. there have been lots of questions about what is the BEST point to use for GND.

... from a stability point of view it would be a point that has no additional GND loops - for that its best to use the same point where the TDA device gets its GND - which is pin 14 (please check on the data sheet at beginning of this thread). I used an easy point of solder - rather than the pin 14 itself - this was the SMD capacitor - the 'end' which is closest to Pin 14 of the TDA device (see original photo) ...

Having said all of the above ... since the way the +3V operation mode of the card reader is governed by internal logic switching of the TDA device ... it’s probably not that critical AS LONG AS there is a healthy 0V (GND) on pin 3 to ensure that its enabled the +3V operating mode ...

If the pin is 'floating' (no connections to it - as could happen if it is cleanly broken) - then the internal circuitry of the TDA still stays at the +5V operating mode ... however this is not the correct mode of operation because the TDA device could 'latch-up' and not operate correctly ... NEVER leave pins floating - even though the internal 'pull-ups' do somehow clamp to Vcc - there is no guarantee and you are leaving the door open to potential transients...
... if you have indeed broken the pin and its ‘floating’ don’t worry too much its really unlikely that it will latch up – but it will stay in the +5V mode of operation ... unless you try to reconnect to pin 3 and bring it to a GND point (method described in a precious post in this thread) ...

Hope the above clarifies the questions of where is the best point to GND pin 3

G'Day Cobbers,
I may be barking up the wrong tree, but, to my mind, this re-enforces my contention that it is not necessary to lift pin #3 of the IC, merely to ground it. It is apparently held high with with a pull-up resistor.
This would then make the mod switchable or easily reversed, rather than a "Brute Force" solution.
Apart from simplifying this mod, it removes the hazard of isolating or breaking the IC leg.

Very Kindest Regards, " The Druid ".

**Aido68** · 15-04-08, 08:45 PM

I wonder if the winner gets the link to this thread..........

mod your strong 4658x 4653x 4654x NO GAMMA overheating - eBay, Other Information Products, Information Products, Lots More.... (end time 19-Apr-08 20:34:33 AEST)

Admin - you should ask for a percentage........

Aido

**eyesee** · 15-04-08, 08:56 PM

Quote:

Originally Posted by beer4life

G'Day Cobbers,
I may be barking up the wrong tree, but, to my mind, this re-enforces my contention that it is not necessary to lift pin #3 of the IC, merely to ground it. It is apparently held high with with a pull-up resistor.
This would then make the mod switchable or easily reversed, rather than a "Brute Force" solution.
Apart from simplifying this mod, it removes the hazard of isolating or breaking the IC leg.

Very Kindest Regards, " The Druid ".

... NO NO NO ... do not GND pin 3 - if you do you will be taking Vcc rail to GND and possibly blow your power supply if there is no protection circuitry ...

... on the existing 'factory' setup ... pin 3 is connected to +5V or the Vcc rail ... there is no pull-up on this pin on the EXTERNAL board.

... I referred to 'pull-up' on pin 3 as part of the internal logic of the TDA ...

... most IC's when having floating pins can latch-up - because in effect if the pin is truly 'floating' it can change state from high to low and oscillate between this at a quite high rate ... all bad stuff for the internal circuitry ...

... so most IC's try to tie a 'floating' pin via a VERY week pull-up - so the pin will be not truly 'floating' ...

... this internal pull-up has an extremely high impedance - so when the pin is taken to a logic low (GND) the pull-up has no further effect ...

If we connect GND to pin 3 directly - without FIRST unsoldering the pin and then lifting it so its 'free' from any contact - then we will 'short out' the Vcc to GND - a BIG NO NO!!

I hope the above explanation will remove any confusion about the internal chip pull-up on floating pins

**gw1** · 15-04-08, 09:56 PM

eyesee is absolutely correct - you need to lift (or cut) pin 3 before connecting it to ground, otherwise you short out the power supply.

The corner screw holes are suitable ground points. Actually they're even better than the ground pin (C5) on the smartcard socket. If you're fussy and can find pin 22 of the TDA80x4 then that's the ideal ground point to use. On the Humax smartcard board that corresponds to either side of jumper link J701.

By the way, the point marked "mystery pin 5" on CoopsOz's photo earlier is one of the contacts of the socket's card detection switch. It's not the same as the smartcard's C5 ground pin (marked "Pin 5? = EARTH").

**beer4life** · 15-04-08, 10:01 PM

Please permit a geriatric to disagree.
Seems to me that Vcc is on #17, not #3.
As I previously asked, where does the track on #3 go to?
If, as you previously explained, it is a logic input, and not a 5 V rail, then why not just switch it to ground.
After all, this is what the multi-voltage logic card input does.
I dips me lid to those who are wiser in this regard, but need to be put in my place if I'm wrong.

Very Kindest Regards, " The Druid ".

**under-** · 16-04-08, 11:13 AM

Argh All i want is pictures haha!

**gw1** · 16-04-08, 12:12 PM

Quote:

Originally Posted by beer4life

Please permit a geriatric to disagree. Seems to me that Vcc is on #17, not #3. As I previously asked, where does the track on #3 go to? If, as you previously explained, it is a logic input, and not a 5 V rail, then why not just switch it to ground. After all, this is what the multi-voltage logic card input does.

Thanks for your contributions, kind Druid. Here's my understanding - I welcome correction if I've got the wrong end of the stick somewhere.

Like many non-volatile devices of their day, early smartcards ran only at 5V but needed up to 21V for programming. Philips' early TDA8000 interface catered for them and was seminal for designs that followed by Philips/NXP and others. It included an on-chip 5V regulator for the card (and perhaps also the controller) plus a voltage supervisor which generated a reset on brownout, with pins to monitor card voltage and control the reset period. Two bidirectional I/O lines were supported plus reset and clock.

Over time cards that required only 5V for programming appeared and 3V MCU hosts became more common. The TDA8004(T) was launched to cater for these changes and quickly became popular and cheap. (The T option denotes smaller SOIC package.) Though architecturally similar to its predecessor there were enough changes to warrant a new pinout. A step-up 5V generator was added since the host system might now only need 3V. A third bidirectional line was added (AUX1 and AUX2 as well as I/O) and the card clock generator was enhanced to support higher frequencies and four division ratios instead of two. The voltage supervisor matured, no longer requiring monitoring or control pins. Two pins previously dedicated by the TDA8000 to alarm signalling were omitted, that feature being folded into the /OFF indicator. In the end the TDA8004(T)'s package had two spare pins: pin 18 which was unused and unconnected, and pin 13 specified "reserved for future use" with advice that it be tied high for the time being - ideally via spare microcontroller I/O if available.

Subsequently Philips brought out the enhanced TDA8004A(T) to add support for 3V cards. Such cards must be 5V tolerant for hosts that only support it, though they may indicate their intended operational voltage in the ATR. The previously reserved pin 3 was utilised by the 'A' interface variant to control card voltage. High logic level (ie Vdd) generates Vcc=5V, but if pulled low a Vcc of 3V would be produced and the level detection & brownout thresholds would allow for that. In that way Philips retained backward compatibility with PCBs and firmware designed for the non-A part. Pin 18 remains unconnected.

There may be an internal pullup on TDA8004A(T) pin 3, but I wouldn't assume so. The datasheet isn't clear but implies there may not be one since it groups pin 3 with other logic inputs which have none; the only internal pullups are on the I/O, AUX1 and AUX2 signals (which need them because of the way the level shifters work).

I hope that clarifies the situation with pin 3. Pin 17 is the card Vcc, generated by the interface IC, fixed at 5V output for original TDA8004(T) or switchable in TDA8004A(T) using a low level on pin 3. The non-'A' TDA8004(T) requires pin 3 to be either driven by a microntroller or tied to Vdd, which is why if you're modding a PCB you can't simply solder a wire from pin 3 to ground - you need to separate it from the PCB track as well.

Doing that mod for products using the 'A' part or a second-source equivalent will give you 3V as intended. Doing the mod on a non-A part won't necessarily give you 3V. It may but the datasheet doesn't commit to that; it just says that pin is reserved. By all means try it, measure the result with a multimeter and let us know your findings. It may be an undocumented feature but I've not seen that acknowledged anywhere official.

By second-source I mean companies other than Philips/NXP who make drop-in compatible parts. An example mentioned earlier in the thread is SGS-Thomson's ST8004, which is a low cost equivalent for the TDA8004A. It supports both 3V and 5V cards, selected using pin 3 as per the 'A' variant. As a bonus they support an enhanced voltage supervision method (using a window around 3.0V together with a time-based noise filter), enabled by strapping pin 18 low; if high or unconnected the default method is used (2.2V deactivation threshold with hysteresis).

Several newer smartcard controllers have since been introduced, including several that support 1.8V cards as well as 3V and 5V.

**gw1** · 16-04-08, 12:19 PM

Quote:

Originally Posted by under-

Argh All i want is pictures haha!

If you're able to get a nice clear closeup photo I'll draw you something for your board (I don't have one of those myself). Most modern digital cameras have a macro option - see if yours has it, and experiment with lighting and camera distance. I need to be able to see the part numbers clearly.

**CoopsOz** · 16-04-08, 12:50 PM

Further to my earlier post, my card at 3.061v is not exactly cool....it's warm at best. A goldie retains no heat whatsoever when inserted in the same box, is a little heat allowable?

**Izzy** · 16-04-08, 12:55 PM

The 3 volts (and the current for the card to work) still comes through the TDA8024 IC. Therefore the IC will still heat up. Maybe not as much as prior to your mod.

I may be corrected but I don't think this mod will be as successful
as taking the supply away from going through the IC

Just my opinion

**beer4life** · 16-04-08, 01:08 PM

Quote:

Originally Posted by gw1

Thanks for your contributions, kind Druid. Here's my understanding - I welcome correction if I've got the wrong end of the stick somewhere.

Like many non-volatile devices of their day, early smartcards ran only at 5V but needed up to 21V for programming. Philips' early TDA8000 interface catered for them and was seminal for designs that followed by Philips/NXP and others. It included an on-chip 5V regulator for the card (and perhaps also the controller) plus a voltage supervisor which generated a reset on brownout, with pins to monitor card voltage and control the reset period. Two bidirectional I/O lines were supported plus reset and clock.

Over time cards that required only 5V for programming appeared and 3V MCU hosts became more common. The TDA8004(T) was launched to cater for these changes and quickly became popular and cheap. (The T option denotes smaller SOIC package.) Though architecturally similar to its predecessor there were enough changes to warrant a new pinout. A step-up 5V generator was added since the host system might now only need 3V. A third bidirectional line was added (AUX1 and AUX2 as well as I/O) and the card clock generator was enhanced to support higher frequencies and four division ratios instead of two. The voltage supervisor matured, no longer requiring monitoring or control pins. Two pins previously dedicated by the TDA8000 to alarm signalling were omitted, that feature being folded into the /OFF indicator. In the end the TDA8004(T)'s package had two spare pins: pin 18 which was unused and unconnected, and pin 13 specified "reserved for future use" with advice that it be tied high for the time being - ideally via spare microcontroller I/O if available.

Subsequently Philips brought out the enhanced TDA8004A(T) to add support for 3V cards. Such cards must be 5V tolerant for hosts that only support it, though they may indicate their intended operational voltage in the ATR. The previously reserved pin 3 was utilised by the 'A' interface variant to control card voltage. High logic level (ie Vdd) generates Vcc=5V, but if pulled low a Vcc of 3V would be produced and the level detection & brownout thresholds would allow for that. In that way Philips retained backward compatibility with PCBs and firmware designed for the non-A part. Pin 18 remains unconnected.

There may be an internal pullup on TDA8004A(T) pin 3, but I wouldn't assume so. The datasheet isn't clear but implies there may not be one since it groups pin 3 with other logic inputs which have none; the only internal pullups are on the I/O, AUX1 and AUX2 signals (which need them because of the way the level shifters work).

I hope that clarifies the situation with pin 3. Pin 17 is the card Vcc, generated by the interface IC, fixed at 5V output for original TDA8004(T) or switchable in TDA8004A(T) using a low level on pin 3. The non-'A' TDA8004(T) requires pin 3 to be either driven by a microntroller or tied to Vdd, which is why if you're modding a PCB you can't simply solder a wire from pin 3 to ground - you need to separate it from the PCB track as well.

Doing that mod for products using the 'A' part or a second-source equivalent will give you 3V as intended. Doing the mod on a non-A part won't necessarily give you 3V. It may but the datasheet doesn't commit to that; it just says that pin is reserved. By all means try it, measure the result with a multimeter and let us know your findings. It may be an undocumented feature but I've not seen that acknowledged anywhere official.

By second-source I mean companies other than Philips/NXP who make drop-in compatible parts. An example mentioned earlier in the thread is SGS-Thomson's ST8004, which is a low cost equivalent for the TDA8004A. It supports both 3V and 5V cards, selected using pin 3 as per the 'A' variant. As a bonus they support an enhanced voltage supervision method (using a window around 3.0V together with a time-based noise filter), enabled by strapping pin 18 low; if high or unconnected the default method is used (2.2V deactivation threshold with hysteresis).

Several newer smartcard controllers have since been introduced, including several that support 1.8V cards as well as 3V and 5V.

G'Day "eyesee" and "GW1",
Very many thanks for the time and effort that you both have given for this explanation.
This is what caused the confusion in my mind:

Quote:

... on the existing 'factory' setup ... pin 3 is connected to +5V or the Vcc rail ... there is no pull-up on this pin on the EXTERNAL board.

I've since borrowed one of the card readers, and ascertained that #3 is tied, via SR12, to the Vdd (5V supply for the IC), and not Vcc as stated.
Careful examination of the tracks around #3 and #1, suggest that while a mod would be possible, that it is not easily made.
I therefor withdraw my suggestion.

Very Kindest Regards and in all humility, "The Druid".

**Decapper** · 20-04-08, 07:54 PM

Lets say if I was stupid enough (not the first time I have done or said) by accidentally leaving the 3rd leg connected. And once I undone what I did to the box and it still did not work. Would there be a fix or should I bin it?

**best4less** · 20-04-08, 08:00 PM

Quote:

Originally Posted by Decapper

Lets say if I was stupid enough (not the first time I have done or said) by accidentally leaving the 3rd leg connected. And once I undone what I did to the box and it still did not work. Would there be a fix or should I bin it?

Are you still getting 5 volts to the chip and the card reader ?

**Decapper** · 20-04-08, 08:13 PM

Don't know. Do not have a meter to test. I know when I enter a card it flashes and then the box resets

**best4less** · 20-04-08, 08:20 PM

Don't use one of your special cards

hate for that
to stop working
You haven't got and solder splash shorting anything else

15-04-08, 07:30 PM	#45 (permalink)
CoopsOz Premium Member iTrader: (4) Join Date: Mar 2008 Posts: 337 Spent time on board: 3 Weeks, 3 Days and 8:38:28	Just a quick one, I was initially measuring across the "mystery" pin 5 and getting 1.8V, after doing some research I think I was measuring at the wrong spot, my pic will explain it easier. So, now I get 3.061v across C1 and C5. Is that correct? I used GW1's picture so disregard all the other components.....this is merely a 3rd leg lift mod. Last edited by CoopsOz : 15-04-08 at 07:32 PM. Reason: Forgot picture.

15-04-08, 08:45 PM	#47 (permalink)
Aido68 Junior Member iTrader: (1) Join Date: Apr 2008 Posts: 24 Spent time on board: -	I wonder..........? I wonder if the winner gets the link to this thread.......... mod your strong 4658x 4653x 4654x NO GAMMA overheating - eBay, Other Information Products, Information Products, Lots More.... (end time 19-Apr-08 20:34:33 AEST) Admin - you should ask for a percentage........ Aido

15-04-08, 10:01 PM	#50 (permalink)
beer4life The Druid. iTrader: (2) Join Date: Feb 2008 Location: The Nether World. Posts: 655 Spent time on board: 1 Month, 3 Weeks, 5 Days and 21:08:22	Please permit a geriatric to disagree. Seems to me that Vcc is on #17, not #3. As I previously asked, where does the track on #3 go to? If, as you previously explained, it is a logic input, and not a 5 V rail, then why not just switch it to ground. After all, this is what the multi-voltage logic card input does. I dips me lid to those who are wiser in this regard, but need to be put in my place if I'm wrong. Very Kindest Regards, " The Druid ". __________________

16-04-08, 12:55 PM	#55 (permalink)
Izzy Registered User iTrader: (1) Join Date: Apr 2008 Posts: 31 Spent time on board: -	how-run-your-gamma-cool-3v-factory The 3 volts (and the current for the card to work) still comes through the TDA8024 IC. Therefore the IC will still heat up. Maybe not as much as prior to your mod. I may be corrected but I don't think this mod will be as successful as taking the supply away from going through the IC Just my opinion

20-04-08, 08:20 PM	#60 (permalink)
best4less I'am Not a Bloody Joke iTrader: (10) Join Date: Jan 2008 Location: Australia Posts: 2,616 Spent time on board: 1 Month, 4 Weeks, 0 Days and 9:43:15	Don't use one of your special cards hate for that to stop working You haven't got and solder splash shorting anything else __________________ Please wipe your feet before walking all over me

15-04-08, 11:20 AM	#41 (permalink)
vbthanks Member iTrader: (0) Join Date: Jan 2008 Posts: 30 Spent time on board: 5:45:09	Happy to report the mod gave me a rock solid 3v and took all of 10 minutes. I used a more stable ground pad off the edge connector. To confirm if it is ground - leave it plugged in to the mainboard and check resistance between the tuner case and the pad with a multimeter for 0 ohms. After 1 hour - the card is not even warm. Well done eyesee!

15-04-08, 02:41 PM	#42 (permalink)
under- Member iTrader: (0) Join Date: Apr 2008 Posts: 79 Spent time on board: 4:24:02	Picture of your completed mod would be good.

15-04-08, 02:49 PM	#43 (permalink)
under- Member iTrader: (0) Join Date: Apr 2008 Posts: 79 Spent time on board: 4:24:02	Check out the picture!

15-04-08, 09:56 PM	#49 (permalink)
gw1 Member iTrader: (1) Join Date: Jan 2008 Location: Hobart Posts: 69 Spent time on board: 1 Day and 14:35:48	eyesee is absolutely correct - you need to lift (or cut) pin 3 before connecting it to ground, otherwise you short out the power supply. The corner screw holes are suitable ground points. Actually they're even better than the ground pin (C5) on the smartcard socket. If you're fussy and can find pin 22 of the TDA80x4 then that's the ideal ground point to use. On the Humax smartcard board that corresponds to either side of jumper link J701. By the way, the point marked "mystery pin 5" on CoopsOz's photo earlier is one of the contacts of the socket's card detection switch. It's not the same as the smartcard's C5 ground pin (marked "Pin 5? = EARTH").

16-04-08, 11:13 AM	#51 (permalink)
under- Member iTrader: (0) Join Date: Apr 2008 Posts: 79 Spent time on board: 4:24:02	Argh All i want is pictures haha!

16-04-08, 12:50 PM	#54 (permalink)
CoopsOz Premium Member iTrader: (4) Join Date: Mar 2008 Posts: 337 Spent time on board: 3 Weeks, 3 Days and 8:38:28	Further to my earlier post, my card at 3.061v is not exactly cool....it's warm at best. A goldie retains no heat whatsoever when inserted in the same box, is a little heat allowable?

20-04-08, 07:54 PM	#57 (permalink)
Decapper pricelessweddings.com.au iTrader: (2) Join Date: Jan 2008 Location: Your mothers house Posts: 627 Spent time on board: 3 Days and 14:50:27	Lets say if I was stupid enough (not the first time I have done or said) by accidentally leaving the 3rd leg connected. And once I undone what I did to the box and it still did not work. Would there be a fix or should I bin it?

20-04-08, 08:13 PM	#59 (permalink)
Decapper pricelessweddings.com.au iTrader: (2) Join Date: Jan 2008 Location: Your mothers house Posts: 627 Spent time on board: 3 Days and 14:50:27	Don't know. Do not have a meter to test. I know when I enter a card it flashes and then the box resets

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Sponsored Links

Sponsored Links

Sponsored Links

Sponsored Links