If a tranistor has absolute max ratings of:
Collector current 20mA and Vce 35V
What will happen if I run it at Vce 7V and collector current 35mA ( 0.25W )
Will this damage the device ?
If so can anyone explain to me why ?
The absolute max power is not stated
is 0.25w too much for the average i/p tr ?
I realise that the beta may fall off but that is easy to check.
Collector current 20mA and Vce 35V
What will happen if I run it at Vce 7V and collector current 35mA ( 0.25W )
Will this damage the device ?
If so can anyone explain to me why ?
The absolute max power is not stated
is 0.25w too much for the average i/p tr ?
I realise that the beta may fall off but that is easy to check.
mikelm said:
Yes, all the maximum ratings specified in the data sheet,
including max. Ic, which you asked about first.
As for the reason for the current limit, I am not sure. I seem
to remember some data sheets for power transistors
mentioning the bonding wires as the limiting factor. However,
I doubt that is the case for small-signal devices, as in your
case. High injection certainly comes into play somehow and
it limits the beta at high currents, but I don't think that per
se gives you any maximum rating. After a quick glance in my
semiconductor physics book it seems that the high injection
current is limited by the emitter series resistance. A guess then
would be that exceeding the max rating might cause too much
heat developing due to this resistance, but that's just a guess.
Somebody will probably provide an answer to this.
Hi Christer
Thanks for the reply
Yes I would agree that you would think that absolute 20mA was there for a reason. I still don't understand what that reason is. Could it be in this case the absolute max's are 20mA and 35V together ?
The datasheet (ssm-2220) quotes a 'a very low bulk resistance' of 0.3ohms.
For my proposed 7V and 35mA I will be asking the transistor to be 200ohms c.e.
In worst possible case ( using a socket ) the device rises 100 deg C / watt.
So with 20 deg C ambient + my .25watt the junction should be about 45 deg C.
both these factors seem reasonable to me.
are there any other ideas as to how this could do damage ?
Thanks for the reply
Yes I would agree that you would think that absolute 20mA was there for a reason. I still don't understand what that reason is. Could it be in this case the absolute max's are 20mA and 35V together ?
The datasheet (ssm-2220) quotes a 'a very low bulk resistance' of 0.3ohms.
For my proposed 7V and 35mA I will be asking the transistor to be 200ohms c.e.
In worst possible case ( using a socket ) the device rises 100 deg C / watt.
So with 20 deg C ambient + my .25watt the junction should be about 45 deg C.
both these factors seem reasonable to me.
are there any other ideas as to how this could do damage ?
Well, remember that the high voltage is always over collector-
base junction, the base-emitter junction stays around 0.6 to
0.7 V. Hence, most of the power is disspiated in the cb junction.
If you look at the physical design of a transistor you will see
that the collector is very large compared to the emitter and
the collector is the one in contact with the metal tab/case if
there is one. That means the collector can more easily dissipate
power. So maybe it doesn't even have to do with the resistance
as I previously guessed,
perhaps it is just that the max. rated current times the be voltage
drop is about as much power as the tiny emitter can dissipate.
base junction, the base-emitter junction stays around 0.6 to
0.7 V. Hence, most of the power is disspiated in the cb junction.
If you look at the physical design of a transistor you will see
that the collector is very large compared to the emitter and
the collector is the one in contact with the metal tab/case if
there is one. That means the collector can more easily dissipate
power. So maybe it doesn't even have to do with the resistance
as I previously guessed,
perhaps it is just that the max. rated current times the be voltage
drop is about as much power as the tiny emitter can dissipate.
I think that I will have to do some tests to see what happens.
I'll try running them in spec for a while and then turn up the current and see what happens. If they don't self destruct I will listen for any subtle sound degradation. The alternative is two devices in paralell which I can use as a reference for comparison.
It might be a while but I will post my results when I have finnished.
I would be interested if anyone else has some input on this.
I'll try running them in spec for a while and then turn up the current and see what happens. If they don't self destruct I will listen for any subtle sound degradation. The alternative is two devices in paralell which I can use as a reference for comparison.
It might be a while but I will post my results when I have finnished.
I would be interested if anyone else has some input on this.
I don't think they will go up in smoke if you exceed the current
rating at Vce where you stay within the power handling capacity.
They may stop working, but more likely you will just degrade
reliability and life time of the devices.
If you attempt going beyond the power rating, on the other hand,
then don't forget to wear protective glasses. These things can
explode!
rating at Vce where you stay within the power handling capacity.
They may stop working, but more likely you will just degrade
reliability and life time of the devices.
If you attempt going beyond the power rating, on the other hand,
then don't forget to wear protective glasses. These things can
explode!
...
...
in this particular case, the maximum Ic rating is probably restricted by the junction area (or die size as a proxy).
I don't think it will explode if you ran 35ma through a 20ma small power transistor. It will probably just fail or work (I have ran both 2n5401 and 5551 over their rated Ic capacities) but not as reliably.
most, if not all, datasheets usually warn you of staying within the absolute electric ratings AND SOA. unfortunately, they usually don't state SOA for small signal transistors.
I don't think it will explode if you ran 35ma through a 20ma small power transistor. It will probably just fail or work (I have ran both 2n5401 and 5551 over their rated Ic capacities) but not as reliably.
most, if not all, datasheets usually warn you of staying within the absolute electric ratings AND SOA. unfortunately, they usually don't state SOA for small signal transistors.
Don't waste your time
The 20mA rating is an absolute maximum and is indepemdent of dupply voltage. This is a do not exceed under any circumstances rating. Lower supply voltage does not allow you to run more current through it. Period.
Besides, not many transistors have such a low current rating. If you need 35mA just buy a heavier transistor. The 2N4401 is like 10 cents and will handle several hyndred mA.
The 20mA rating is an absolute maximum and is indepemdent of dupply voltage. This is a do not exceed under any circumstances rating. Lower supply voltage does not allow you to run more current through it. Period.
Besides, not many transistors have such a low current rating. If you need 35mA just buy a heavier transistor. The 2N4401 is like 10 cents and will handle several hyndred mA.
Re: Don't waste your time
Would you by change happen to know the reason also? I agree this
must reasonably a max rating independent of Vce and speculated
earlier in the thread about what might be the reason for this. Do you
know if that speculation is reasonable, or if there is some other
reason for this limit?
dmfraser said:
Would you by change happen to know the reason also? I agree this
must reasonably a max rating independent of Vce and speculated
earlier in the thread about what might be the reason for this. Do you
know if that speculation is reasonable, or if there is some other
reason for this limit?
Reason
Because that's all the current capacity the designer of the transistor cared to make it good for. That's all the silicon cross section area they allowed for in the design of the device. Usually transistors with such tiny current ratings are pre-amp input transistors that are very low noise or very high bandwidth and are intended to operate in the sub mA range.
For reliable design, you NEVER run a transistor at more than 1/3 of its rated current.
If you need 35mA you are not building a low noise circuit. For that load current, you need a minimum 100mA current rating. Otherwise, you will likely have a failure sooner than you like.
100mA is not a big current rating. Just use a different transistor and be done with it. There is no worse time waste that overdriving a part when a part than meets the spec is worth like 10 cents. Heck, we buy 600mA rated transistors for 5 cents each because we buy 1000 at a time.
We engineers put these rating on for a reason.
Because that's all the current capacity the designer of the transistor cared to make it good for. That's all the silicon cross section area they allowed for in the design of the device. Usually transistors with such tiny current ratings are pre-amp input transistors that are very low noise or very high bandwidth and are intended to operate in the sub mA range.
For reliable design, you NEVER run a transistor at more than 1/3 of its rated current.
If you need 35mA you are not building a low noise circuit. For that load current, you need a minimum 100mA current rating. Otherwise, you will likely have a failure sooner than you like.
100mA is not a big current rating. Just use a different transistor and be done with it. There is no worse time waste that overdriving a part when a part than meets the spec is worth like 10 cents. Heck, we buy 600mA rated transistors for 5 cents each because we buy 1000 at a time.
We engineers put these rating on for a reason.
Re: Don't waste your time
This device is a high quality matched dual transistor especially designed for audio i/p's.
I am hoping that in my design, the tight thermal matching will, to a large extent, erradicate memory distortion. So using a two single transistors is not a valid option for me.
If anyone knows of a dual matched device that allows higher currents please let me know.
I am now building this cct so I will find out ( possibly the hard way ) if these devices can take more that 20mA. In reality though I will probably run these devices at a current that gives a highest current gain. This will probably be around 20mA from what I can see on the data sheets.
Looks like I may be using two or three devices in paralell.
dmfraser said:
This device is a high quality matched dual transistor especially designed for audio i/p's.
I am hoping that in my design, the tight thermal matching will, to a large extent, erradicate memory distortion. So using a two single transistors is not a valid option for me.
If anyone knows of a dual matched device that allows higher currents please let me know.
I am now building this cct so I will find out ( possibly the hard way ) if these devices can take more that 20mA. In reality though I will probably run these devices at a current that gives a highest current gain. This will probably be around 20mA from what I can see on the data sheets.
Looks like I may be using two or three devices in paralell.
Re: Reason
I hear what you are saying but I think that the point Christer and I are asking is
'what exactly within the device will breakdown and why ?'
I though that it was heat that damaged or destroyed semiconductor devices. Having studied every perameter that I can find I cannot see why 0.25W will damage this device.
I'm not saying that you are wrong but I do not understand why you might be right.
cheers
mike
dmfraser said:
I hear what you are saying but I think that the point Christer and I are asking is
'what exactly within the device will breakdown and why ?'
I though that it was heat that damaged or destroyed semiconductor devices. Having studied every perameter that I can find I cannot see why 0.25W will damage this device.
I'm not saying that you are wrong but I do not understand why you might be right.
cheers
mike
Current Rating
I am quite surprised you need 35mA on a dual matched device to get best gain and lowest noise. This normally occurs in low noise, high gain transistors in the sub 1mA range.
If the designer rated the transistor at 20mA there may be many reasons to so so. However, if you want a reliable unit, there are 2 transistor specs you never, never exceed. If possible you never even exceed 50% of these specs. These are:
Ic - Maximum curremt
Vce - Maximum voltage drop
Pd - Maximum power dissapation.
Going lower on one does NOT allow you more on the others, Transistors are NOT resistors. If you exceed any of the others, you will likely have an unreliable product. Worse, it may not fail outright, just get flaky.
As well, if you are using this device, trying to get maximum gain from it but putting it inside a feedback loop, you are not necessarily going to make the best sounding product. The highest gain point is often the most non linear point as well and then you are going throw away the gain and depend on feedback to get the thing to have the gain you want. However, this means you are dependent of feedback, which always lags the signal a bit. The most musical designs accept lower gain per stage and operate devices in their most linear ranges so the feedback does not have to work as hard.
As well, the maximm gain point usually has the lowest bandwidth and again, you end up using feedback to clean things up. You need feedback in most designs anyway but the idea is to make the circuit as linear as possible before feedback then put the loop around it.
Also you might find that at that high a current you will not be operating your device at its lowest noise point, which is far more important that operating it at its highest gain point.
Don't sacrifice reliability for squeezing the last bit of gain out of your circuit's open loop gain. If you need a really want a low noise super matched pair, look up the National LM394.
Besides, because you're over driving the device, every time there is anything not quite right with it, you get some noise or something, you're going to worry about this over worked device and eventually, even if it does not fail, it will nag you to the point you're going to do it over anyway. Trust me. I've been there.
Remember, a good designer never does anything to sacrifice reliability.
I am quite surprised you need 35mA on a dual matched device to get best gain and lowest noise. This normally occurs in low noise, high gain transistors in the sub 1mA range.
If the designer rated the transistor at 20mA there may be many reasons to so so. However, if you want a reliable unit, there are 2 transistor specs you never, never exceed. If possible you never even exceed 50% of these specs. These are:
Ic - Maximum curremt
Vce - Maximum voltage drop
Pd - Maximum power dissapation.
Going lower on one does NOT allow you more on the others, Transistors are NOT resistors. If you exceed any of the others, you will likely have an unreliable product. Worse, it may not fail outright, just get flaky.
As well, if you are using this device, trying to get maximum gain from it but putting it inside a feedback loop, you are not necessarily going to make the best sounding product. The highest gain point is often the most non linear point as well and then you are going throw away the gain and depend on feedback to get the thing to have the gain you want. However, this means you are dependent of feedback, which always lags the signal a bit. The most musical designs accept lower gain per stage and operate devices in their most linear ranges so the feedback does not have to work as hard.
As well, the maximm gain point usually has the lowest bandwidth and again, you end up using feedback to clean things up. You need feedback in most designs anyway but the idea is to make the circuit as linear as possible before feedback then put the loop around it.
Also you might find that at that high a current you will not be operating your device at its lowest noise point, which is far more important that operating it at its highest gain point.
Don't sacrifice reliability for squeezing the last bit of gain out of your circuit's open loop gain. If you need a really want a low noise super matched pair, look up the National LM394.
Besides, because you're over driving the device, every time there is anything not quite right with it, you get some noise or something, you're going to worry about this over worked device and eventually, even if it does not fail, it will nag you to the point you're going to do it over anyway. Trust me. I've been there.
Remember, a good designer never does anything to sacrifice reliability.
Re: Re: Reason
not sure exactly what will happen with your particular device, mike. But for power devices (BJT), hot spotting is what ruins a bjt: because of their minority carrier structure, certain parts of the BJT will carry more than their fair share of the current, thus heating up more than the rest of the die. As that happens, its "resistence" goes down, attracting more current -> more heat. I suspect the same is true in a small signal BJT.
Supposedly, hot spotting doesn't happen with MOSFETs (they are majority carrier devices) but I have seen documents that claim otherwise. It is, however, fair to say that most people don't think mosfets suffer from hot spotting.
mikelm said:
not sure exactly what will happen with your particular device, mike. But for power devices (BJT), hot spotting is what ruins a bjt: because of their minority carrier structure, certain parts of the BJT will carry more than their fair share of the current, thus heating up more than the rest of the die. As that happens, its "resistence" goes down, attracting more current -> more heat. I suspect the same is true in a small signal BJT.
Supposedly, hot spotting doesn't happen with MOSFETs (they are majority carrier devices) but I have seen documents that claim otherwise. It is, however, fair to say that most people don't think mosfets suffer from hot spotting.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.