Hi John, as far as I remember, higher Zener currents alone (i.e. excluding temperature rise effects) let the pn structure age / drift faster compared to operation at lower currents. This might probably be an exponential effect, so I would not increase the Zener beyond the given limit of 5mA at all. Remember, a Zener pn junction is operated in reverse mode, and charge is highly accelerated inside the silicon bulk material by the Zener potential. Higher currents will therefore interact stronger with the silicon crystal structure, than lower ones, that means it will create more permament defects on the long run, which gives rise to stronger aging. Also, this is not a discrete Zener component, but a very delicate sub-surface structure, optimized for lowest drift, but presumably not for high currents. 10, 20 or even 30mA will for sure damage the Zener. Frank

Hi Frank, That is an interesting idea and would explain why 5mA is used. There is an "electromigration" effect that people blame for long term damage caused by high current density I think, Taken to its logical conclusion, one could turn on a low-current, high-noise, low-drift reference once an hour and use it to adjust a lower-noise, higher drift type. Which could be another LTZ1000 run at higher current! :) But.... I don't think the datasheet does give an actual upper limit, unless I missed it. Just two or three example current values. The newer(?) LTZ1000A would have a significant temperature rise caused by the extra current (~15C per extra 5mA) so I can see why 5mA would be the practical maximum for this part. Perhaps I should ask them. The LM399 is another buried zener, the ancestor of the LTZ1000 as I understand it. It's datasheet does state a 0.5-10mA operating range and 20mA absolute maximum. John Frank Stellmach <frank.stellmach@freenet.de> writes: > Hi John, > > as far as I remember, higher Zener currents alone (i.e. excluding > temperature rise effects) let the pn structure age / drift faster > compared to operation at lower currents. > This might probably be an exponential effect, so I would not increase > the Zener beyond the given limit of 5mA at all. > > Remember, a Zener pn junction is operated in reverse mode, and charge > is highly accelerated inside the silicon bulk material by the Zener > potential. Higher currents will therefore interact stronger with the > silicon crystal structure, than lower ones, that means it will create > more permament defects on the long run, which gives rise to stronger > aging. > > Also, this is not a discrete Zener component, but a very delicate > sub-surface structure, optimized for lowest drift, but presumably not > for high currents. 10, 20 or even 30mA will for sure damage the Zener. > > Frank -- John Devereux

Some years ago the company I was working for at that time used a 6.2 V zener of about 1cm square, that had steadily decreasing dynamic resistance up to quite high currents...one thing you could explore is the dynamic resistance as the current is slowly increased.If there is a sudden increase in dynamic resistance or there is a change in slope of the zener volts with current then you are starting to use the shoulder of the zener rather than the flat base region for a buried zener. A few days ago I found a website with very clear drawings of how a buried zener is created and doped to keep the zener region deep in the silicon but I cannot find it again now I need it. I believe that keeping the temp down is the main concern for lower currents , if you can keep it cool you can run it harder. I am going to try 30mA on an LTZ100a, I believe that with a 20C room temp at that current the zener region will be at about 120C. M K > From: john@devereux.me.uk > To: volt-nuts@febo.com > Date: Fri, 21 Oct 2011 19:33:33 +0100 > Subject: Re: [volt-nuts] LTZ1000 at higher currents > > > > Hi Frank, > > That is an interesting idea and would explain why 5mA is used. There is > an "electromigration" effect that people blame for long term damage > caused by high current density I think, > > Taken to its logical conclusion, one could turn on a low-current, > high-noise, low-drift reference once an hour and use it to adjust a > lower-noise, higher drift type. Which could be another LTZ1000 run at > higher current! :) > > But.... I don't think the datasheet does give an actual upper limit, > unless I missed it. Just two or three example current values. > > The newer(?) LTZ1000A would have a significant temperature rise caused > by the extra current (~15C per extra 5mA) so I can see why 5mA would be > the practical maximum for this part. > > Perhaps I should ask them. > > The LM399 is another buried zener, the ancestor of the LTZ1000 as I > understand it. It's datasheet does state a 0.5-10mA operating range and > 20mA absolute maximum. > > John > > Frank Stellmach <frank.stellmach@freenet.de> writes: > > > Hi John, > > > > as far as I remember, higher Zener currents alone (i.e. excluding > > temperature rise effects) let the pn structure age / drift faster > > compared to operation at lower currents. > > This might probably be an exponential effect, so I would not increase > > the Zener beyond the given limit of 5mA at all. > > > > Remember, a Zener pn junction is operated in reverse mode, and charge > > is highly accelerated inside the silicon bulk material by the Zener > > potential. Higher currents will therefore interact stronger with the > > silicon crystal structure, than lower ones, that means it will create > > more permament defects on the long run, which gives rise to stronger > > aging. > > > > Also, this is not a discrete Zener component, but a very delicate > > sub-surface structure, optimized for lowest drift, but presumably not > > for high currents. 10, 20 or even 30mA will for sure damage the Zener. > > > > Frank > > -- > > John Devereux > > _______________________________________________ > volt-nuts mailing list -- volt-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts > and follow the instructions there.

Hi, This link shows in two of the links that noise is inversely proportional to the square root of the current, and one of them gives a threshold below which the noise increases dramatically. http://www.ciphersbyritter.com/RES/NOISE.HTM Regards, M K > From: john@devereux.me.uk > To: volt-nuts@febo.com > Date: Fri, 21 Oct 2011 19:33:33 +0100 > Subject: Re: [volt-nuts] LTZ1000 at higher currents > > > > Hi Frank, > > That is an interesting idea and would explain why 5mA is used. There is > an "electromigration" effect that people blame for long term damage > caused by high current density I think, > > Taken to its logical conclusion, one could turn on a low-current, > high-noise, low-drift reference once an hour and use it to adjust a > lower-noise, higher drift type. Which could be another LTZ1000 run at > higher current! :) > > But.... I don't think the datasheet does give an actual upper limit, > unless I missed it. Just two or three example current values. > > The newer(?) LTZ1000A would have a significant temperature rise caused > by the extra current (~15C per extra 5mA) so I can see why 5mA would be > the practical maximum for this part. > > Perhaps I should ask them. > > The LM399 is another buried zener, the ancestor of the LTZ1000 as I > understand it. It's datasheet does state a 0.5-10mA operating range and > 20mA absolute maximum. > > John > > Frank Stellmach <frank.stellmach@freenet.de> writes: > > > Hi John, > > > > as far as I remember, higher Zener currents alone (i.e. excluding > > temperature rise effects) let the pn structure age / drift faster > > compared to operation at lower currents. > > This might probably be an exponential effect, so I would not increase > > the Zener beyond the given limit of 5mA at all. > > > > Remember, a Zener pn junction is operated in reverse mode, and charge > > is highly accelerated inside the silicon bulk material by the Zener > > potential. Higher currents will therefore interact stronger with the > > silicon crystal structure, than lower ones, that means it will create > > more permament defects on the long run, which gives rise to stronger > > aging. > > > > Also, this is not a discrete Zener component, but a very delicate > > sub-surface structure, optimized for lowest drift, but presumably not > > for high currents. 10, 20 or even 30mA will for sure damage the Zener. > > > > Frank > > -- > > John Devereux > > _______________________________________________ > volt-nuts mailing list -- volt-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts > and follow the instructions there.