Spec Difference between 2N3055 and 2N3055H ?
Pierre Gauvin
transitors ?
I substituted 2 2N3055 for 2N3055H in a Thermocouple Attaching Unit. It
is a spot welding unit that is battery operated. The 2 transistors are
part of an oscillator circuit that provides AC voltage to a step up
transformer. The rectified DC charge a 2000 mfd 100V Capacitor.
The unit work well up to 80 volts, but the remaining 20 volts is very
slow. Before, the system could generate easily 120V. Now it barely hit
95V. The capacitor is OK.
Pierre.
Trevor Wilson
"Pierre Gauvin" <gau...@citenet.net> wrote in message
news:3afd8e07...@206.172.150.50...
**I don't know of any 2N3055 variants, which can be used at such Voltages
(120 Volts), safely. Which is not to say that they don't exist, of course.
2N3055 devices being produced today, are a long way from the 60 Volt, single
diffused, 800kHz devices originally specc'd, way back in the dark ages.
Personally, I'd suggest a couple of MJ15003s, or MJ15024s, for more reliable
long term performance. Maybe some 2N3773s would be nice. They're pretty much
unbreakable, but their fT is crap (200kHz). BTW: The 2N3055H, according to
my Towers, is rated for 100 Volts.
--
Trevor Wilson
http://www.rageaudio.com.au
Ross Herbert
The best I can do is to refer you to the pages where you can read the data
sheets for yourself;
This page seems to indicate that there are 3 different grades of 2N3055 all
at different prices. Without downloading and analysing the data sheets I
can't tell what the differences are but it is a safe bet that the higher the
price the better the device is able to handle inductive loads as it appears
your spot welder is doing.
Pierre Gauvin wrote:
--
Regards,
Ross Herbert
ros...@vianet.net.au
Pierre Gauvin
<tre...@rageaudio.com.au> wrote:
>
>"Pierre Gauvin" <gau...@citenet.net> wrote in message
>news:3afd8e07...@206.172.150.50...
>> Could someone give me a URL or give me the main difference between the 2
>> transitors ?
>>
>> I substituted 2 2N3055 for 2N3055H in a Thermocouple Attaching Unit. It
>> is a spot welding unit that is battery operated. The 2 transistors are
>> part of an oscillator circuit that provides AC voltage to a step up
>> transformer. The rectified DC charge a 2000 mfd 100V Capacitor.
>>
>> The unit work well up to 80 volts, but the remaining 20 volts is very
>> slow. Before, the system could generate easily 120V. Now it barely hit
>> 95V. The capacitor is OK.
>>
>> Pierre.
>
>**I don't know of any 2N3055 variants, which can be used at such Voltages
>(120 Volts), safely. Which is not to say that they don't exist, of course.
>2N3055 devices being produced today, are a long way from the 60 Volt, single
>diffused, 800kHz devices originally specc'd, way back in the dark ages.
The system is supposed to work at a maximum of 100V (capacitor 100V).
It could work at a higher voltage. The unit is taking a lot more time
to go to close to 100V. It was taking originaly about 5 sec. to reach
the 100V mark. Now it is taking at least 45 sec. and more, depending on
battery voltage. The 2 2N3055 that the clerk gave me seem old (new but
old). I was thinking that since the parts are probably of old design,
they can operate at only 800 kHz and the 2N3055H could operate at higher
than 800kHz. That woud explain the fact that the unit can reach 80V in
4 or 5 sec. and the rest of the time to reach 95V.
>Personally, I'd suggest a couple of MJ15003s, or MJ15024s, for more reliable
>long term performance. Maybe some 2N3773s would be nice. They're pretty much
>unbreakable, but their fT is crap (200kHz). BTW: The 2N3055H, according to
>my Towers, is rated for 100 Volts.
The sujested better parts from online.com is MJ15015. Any comment on
this part. The MJ parts are probably easier to obtain (on shelf) than
the 2N3055H.
Pierre.
Pierre Gauvin
wrote:
>Pierre,
>
>The best I can do is to refer you to the pages where you can read the data
>sheets for yourself;
>
>http://www.onsemi.com/pub/produ/1,2075,products1-1-2,00.html?searchType=product&searchField=2n3055&ProductQuicklinks=
>
Thank you for the URL. I was looking for someting like that.
Too bad they did not had the original 2N3055 data spec.
>This page seems to indicate that there are 3 different grades of 2N3055 all
>at different prices. Without downloading and analysing the data sheets I
>can't tell what the differences are but it is a safe bet that the higher the
>price the better the device is able to handle inductive loads as it appears
>your spot welder is doing.
Well the data sheets from the site are the same for the 2N3055 types.
It is all the 2N3055A.
Thanks again for the URL.
Pierre.
Trevor Wilson
>
> The system is supposed to work at a maximum of 100V (capacitor 100V).
> It could work at a higher voltage. The unit is taking a lot more time
> to go to close to 100V. It was taking originaly about 5 sec. to reach
> the 100V mark. Now it is taking at least 45 sec. and more, depending on
> battery voltage. The 2 2N3055 that the clerk gave me seem old (new but
> old). I was thinking that since the parts are probably of old design,
> they can operate at only 800 kHz and the 2N3055H could operate at higher
> than 800kHz. That woud explain the fact that the unit can reach 80V in
> 4 or 5 sec. and the rest of the time to reach 95V.
**That does seem a plausible explanation. Whilst most 2N3055 variants
possess an fT of significantly less than 1MHz, some are capable of much
higher performance. Some Japanese variants are/were capable of over 5MHz.
>
> >Personally, I'd suggest a couple of MJ15003s, or MJ15024s, for more
reliable
> >long term performance. Maybe some 2N3773s would be nice. They're pretty
much
> >unbreakable, but their fT is crap (200kHz). BTW: The 2N3055H, according
to
> >my Towers, is rated for 100 Volts.
>
> The sujested better parts from online.com is MJ15015. Any comment on
> this part. The MJ parts are probably easier to obtain (on shelf) than
> the 2N3055H.
>
**I can't find an fT figure for the MJ15015, but I guess it should be OK,
since Motorola doesn't make single diffused transistors.
Robert Strand
> "Pierre Gauvin" <gau...@citenet.net> wrote in message
> news:3afe9840...@206.172.150.50...
> >
> > The system is supposed to work at a maximum of 100V (capacitor 100V).
> > It could work at a higher voltage. The unit is taking a lot more time
> > to go to close to 100V. It was taking originaly about 5 sec. to reach
> > the 100V mark. Now it is taking at least 45 sec. and more, depending on
> > battery voltage. The 2 2N3055 that the clerk gave me seem old (new but
> > old). I was thinking that since the parts are probably of old design,
> > they can operate at only 800 kHz and the 2N3055H could operate at higher
> > than 800kHz. That woud explain the fact that the unit can reach 80V in
> > 4 or 5 sec. and the rest of the time to reach 95V.
>
> **That does seem a plausible explanation. Whilst most 2N3055 variants
> possess an fT of significantly less than 1MHz, some are capable of much
> higher performance. Some Japanese variants are/were capable of over 5MHz.
>
What do you mean by "2N3055 Variants"? Other than A and H suffixes what else is
there, BDY20's ?
For the OP's original problem, there may be a problem saturating the switch
transistors, possibly related to the transistor gain. I'm not 100% clear on
how you are winding-up the voltage but if there is a saturation problem then for
a given input voltage output voltage will be lower. I suspect the circuit is a
simple Royer converter, you may be able to get it to go by decreasing the
base-drive resistor(s) - assuming they exist at all. There transistors may not
even be involved with the 100V part of the circuit. Have tried to trace any
of it?
Regards
Rob
Trevor Wilson
"Robert Strand" <rst...@ihug.com.au> wrote in message
news:3B007308...@ihug.com.au...
> Trevor Wilson wrote:
>
> > "Pierre Gauvin" <gau...@citenet.net> wrote in message
> > news:3afe9840...@206.172.150.50...
> > >
> > > The system is supposed to work at a maximum of 100V (capacitor 100V).
> > > It could work at a higher voltage. The unit is taking a lot more time
> > > to go to close to 100V. It was taking originaly about 5 sec. to reach
> > > the 100V mark. Now it is taking at least 45 sec. and more, depending
on
> > > battery voltage. The 2 2N3055 that the clerk gave me seem old (new
but
> > > old). I was thinking that since the parts are probably of old design,
> > > they can operate at only 800 kHz and the 2N3055H could operate at
higher
> > > than 800kHz. That woud explain the fact that the unit can reach 80V
in
> > > 4 or 5 sec. and the rest of the time to reach 95V.
> >
> > **That does seem a plausible explanation. Whilst most 2N3055 variants
> > possess an fT of significantly less than 1MHz, some are capable of much
> > higher performance. Some Japanese variants are/were capable of over
5MHz.
> >
>
> What do you mean by "2N3055 Variants"? Other than A and H suffixes what
else is
> there, BDY20's ?
**2N3055A
2N3055C
2N3055E
2N3055H
2N3055S
2N3055U
2N3055V
2N3055-1, thru 2N3055-9
2N3055SD
There are probably lots of others.
The 'E' variant has a quoted fT of 2.5MHz. All the rest are quoted with an
fT of 800kHz. Some Japanese ones, I'vee seen are good to at least 5 MHz
(Hitachi).
Phil Allison
The famous 2N3055 power transistor has probably been made by more
manufacturers than any other single type. In a sense every different brand
is a "variant" of the type as no two are ever quite the same.
The number is JEDEC resistered which means that any device that meets the
relevant specs can be labelled a 2N3055, if it is inferior or superior it
gets a tail on the number. Each maker publishes specs which give the
details. My old Motorola book gives their specs for what has long been
regarded as one of the best '3055s available.
This includes a spec for "Ft" = 2.5 MHz (min) and "fhfe" = 10kHz (min). The
fhfe is the small signal current gain bandwidth while the Ft is the product
of current gain and bandwidth.
A power transistor cannot actually operate at its Ft as it has a current
gain less than unity - this is a limit frequency. A higher published Ft
figure may only indicate a higher small signal current gain and not wider
bandwidth.
For this and other reasons power inverters using '3055s normally operate at
less than 20 kHz if losses are to be kept low.
Regards, Phil
In switching ccts the gain and saturation voltage matter most. The '3055
takes about 1 amp of base drive to give 10 amps collector current at about
1.2 volts.
Trevor Wilson
"Phil Allison" <bi...@bigpond.com> wrote in message
news:yC2M6.27246$482.1...@newsfeeds.bigpond.com...
> Trevor,
>
> The famous 2N3055 power transistor has probably been made by more
> manufacturers than any other single type. In a sense every different brand
> is a "variant" of the type as no two are ever quite the same.
**Correct.
>
> The number is JEDEC resistered which means that any device that meets the
> relevant specs can be labelled a 2N3055, if it is inferior or superior it
> gets a tail on the number. Each maker publishes specs which give the
> details. My old Motorola book gives their specs for what has long been
> regarded as one of the best '3055s available.
**Yes and, no. Motorola used to manufacture their devices with aluminium
cases. Aluminium is not as good as steel, or copper. RCA always used steel
(which Motorola then switched to, after a decade or so of aluminium cases)
and arguably hads the best power devices. Motorola 2N3055s were Epitaxial
Base types, which have a superior fT, but an inferior robustness to that of
Single Diffused (which all the 800kHz types are). RCA, for a long time,
employed superior wire bonding techniques to that of Motorola and their
devices were always the preferred types, for those who cared. Even when they
second sourced Motorola chips (for things like MJ15003/4), RCA devices were
much more reliable.
>
> This includes a spec for "Ft" = 2.5 MHz (min) and "fhfe" = 10kHz (min).
The
> fhfe is the small signal current gain bandwidth while the Ft is the
product
> of current gain and bandwidth.
**Yep.
>
> A power transistor cannot actually operate at its Ft as it has a current
> gain less than unity - this is a limit frequency. A higher published Ft
> figure may only indicate a higher small signal current gain and not wider
> bandwidth.
**Yep.
>
> For this and other reasons power inverters using '3055s normally operate
at
> less than 20 kHz if losses are to be kept low.
**Absolutely. Better still, is to use a more modern, faster device. There
are plenty available.
Robert Strand
> **Yes and, no. Motorola used to manufacture their devices with aluminium
> cases. Aluminium is not as good as steel, or copper. RCA always used steel
> (which Motorola then switched to, after a decade or so of aluminium cases)
> and arguably hads the best power devices. Motorola 2N3055s were Epitaxial
> Base types, which have a superior fT, but an inferior robustness to that of
> Single Diffused (which all the 800kHz types are). RCA, for a long time,
> employed superior wire bonding techniques to that of Motorola and their
> devices were always the preferred types, for those who cared. Even when they
> second sourced Motorola chips (for things like MJ15003/4), RCA devices were
> much more reliable.
Ah so that's why there are both 2.5MHz and 800kHz fT types. I've never been
able to work out why both exist, I just out it down to the bad old days of
semiconductor manufacture. Most are simply marked 2N3055's without a suffix and
there's no way to tell, without perhaps chasing up an obscure manufacturer data
sheet - which probably doesn't even show an fT! Incidently, where did you find
that info?
PS: Thanks for 2N3055 variants list. I haven't seen any of those around except
for the -1..-9 etc. which I believe are gain sorts.
Regards
Rob
Phil Allison
do you have a reason for believing that Ft and good performance in a
switching inverter are closely related? If so explain.
What is the important difference between the bonding wire method used by
Motorola and RCA?
What does your "yep" mean?
Regards, Phil
Trevor Wilson
**Various places. I investigated fT figures, many years ago, when I
discovered that 2N3055s were lousy replacements, in some audio amplifiers.
Later I found some 2N3055s were OK. Hitachis were excellent. I obtained data
sheets (long since dissappeared) and learned why they were better than
almost any other variant.
>
> PS: Thanks for 2N3055 variants list. I haven't seen any of those around
except
> for the -1..-9 etc. which I believe are gain sorts.
**A combination of factors. Breakdown Volts, mostly and HFE.
Trevor Wilson
> Trevor,
>
> do you have a reason for believing that Ft and good performance in a
> switching inverter are closely related? If so explain.
**I had always assumed so. I know little about inverter PSU's (except that
high speed components are required). If I am incorrect in my assumption, I
will be happy to be educated.
>
> What is the important difference between the bonding wire method used by
> Motorola and RCA?
**Solid silver strap, pierced by the Base and Emitter legs. The other end
was securely bonded to the Silicon, with a relatively large contact area.
The RCA devices proved to be SIGNIFICANTLY more reliable than Motorola (and
just about any other brands) in real use. The specific devices, I am
referring to, are the RCA variants of the MJ15003/4 and MJ15034/5. AFAIK,
the actual Silicon was manufactured by Motorola, but the transistors were
assembled by RCA. I thought it was common knowledge that RCA made better
devices.
>
> What does your "yep" mean?
**Yes.
Phil Allison
as you have so politely asked about switching transistors the specs that
matter most are the saturation voltages and currents and the switching
times. Particularly the "charge storage time" and "fall time" as slow turn
off causes most of the heat dissipation. These figures are quoted in data
sheets and also the Farnell cat in brief form.
Switching transistors are a special class and quite unlike amplifer types
which must have high gain with good linearity and SOA figures.
Transistor Ft is a small signal parameter and not very relevant to high
speed power switching.
Some RCA TO3 transistors have the solid metal lead frames you mention
connecting B and E pads to their pins but others dont. The big disadvantage
of this method is the very high fusing current. Aluminium bonding wire as
used by Motorola (and most others) has a specified fusing limit so that
overstressed and failed devices can disconnect themselves and limit damage.
Figures are given in the published SOA curves.The lack of internal fuses is
a notable problem on plastic pack devices too.
Steel TO3s are preferable to aluminium alloy and Motorola finally changed
over in about 1987. This is despite the much better thermal conductivity of
aluminium. Hitachi's copper TO3s ARE good, they used to put their famous
Mosfets in them!
Regards, Phil
Phil Allison
Trevor Wilson
> Trevor,
>
> as you have so politely asked about switching transistors the specs that
> matter most are the saturation voltages and currents and the switching
> times. Particularly the "charge storage time" and "fall time" as slow turn
> off causes most of the heat dissipation. These figures are quoted in data
> sheets and also the Farnell cat in brief form.
>
**Thanks for the explanation. Nevertheless, it seems clear that a standard
2N3055 is a very poor choice for a SMPS.
Robert Strand
> **Various places. I investigated fT figures, many years ago, when I
> discovered that 2N3055s were lousy replacements, in some audio amplifiers.
> Later I found some 2N3055s were OK. Hitachis were excellent. I obtained data
> sheets (long since dissappeared) and learned why they were better than
> almost any other variant.
Thanks!
Regards
Rob
Russell Shaw
Phil Allison wrote:
>
> Trevor,
>
> Transistor Ft is a small signal parameter and not very relevant to high
> speed power switching.
Ft is as important as any other parameter for switching. It relates
charge required to switch a certain current.
--
___ ___
/ /\ / /\
/ /__\ / /\/\
/__/ / Russell Shaw, B.Eng, M.Eng(Research) /__/\/\/
\ \ / Victoria, Australia, Down-Under \ \/\/
\__\/ \__\/
Robert Strand
> as you have so politely asked about switching transistors the specs that
> matter most are the saturation voltages and currents and the switching
> times. Particularly the "charge storage time" and "fall time" as slow turn
> off causes most of the heat dissipation. These figures are quoted in data
> sheets and also the Farnell cat in brief form.
I wasn't going to be picky but the switching characteristics of BJTs are a
strong function of the base drive circuits, basically how much charge you force
in and out of the base. The tabulated figures are for specific conditions and
are not consistent between transistor types, they can only be used as a rough
guide.
> Transistor Ft is a small signal parameter and not very relevant to high
> speed power switching.
As Russ hinted at, fT represents a large part of the stored charge which means
it does affect switching. However, it doesn't represent the all of the
charge. Under saturated conditions there is extra charge in the base due to
the forward biased bc junction. This charged has an effect on the time to get
out of staturation and doesn't appear in the (forward) Ft. At high voltages
depletion capacitances can also affect switch times. Although the depletion
capacitance is represented in fT it contributes very little to it's value under
normal conditions. Still not the full story but perhaps a little clearer.
Regards
Rob
Phil Allison
Trevor Wilson <tre...@rageaudio.com.au> wrote in message
news:qLhM6.55342$ff.4...@news-server.bigpond.net.au...
Trevor,
in the question that started this thread we had an old battery operated
inverter that charged a 2000uF cap to 100 volts in 5 seconds. The energy in
the cap is about 10 joules - similar to a small photo flash unit.
Using a pair of '3055s is overkill for such a job - my old camera flash
uses one tiny germanium.
Do you remember the CDI ignition kits? - '3055 inverters were the go in
them. It all a matter of horses for courses.
Regards, Phil
Trevor Wilson
> Trevor,
>
> in the question that started this thread we had an old battery operated
> inverter that charged a 2000uF cap to 100 volts in 5 seconds. The energy
in
> the cap is about 10 joules - similar to a small photo flash unit.
> Using a pair of '3055s is overkill for such a job - my old camera flash
> uses one tiny germanium.
**Sure and the 2N3055s failed. IT would seem sensible to replace them with
more capable units. The cost of service is significantly higher than the
added cost of superior (and presumably more reliable) transistors. Maybe
it's just me. When I service a product, I don't like to see it back. I
install devices which at at least the equal of what I remove. Usually, I'll
replace them with the next higher rated device (which is easy enough, with
Japanese semis).
>
> Do you remember the CDI ignition kits? - '3055 inverters were the go in
> them. It all a matter of horses for courses.
**True enough. I recall building one, in the dim, dark ages. I used high
Voltage, high speed transistors in it, too. It's probably still operating,
in someone's Ford Escort. I pretty much don't use 2N3055s where the
operating frequency is much above DC. They're an excellent choice, down
there.
Phil Allison
Trevor,
we dont know how or even if the original '3055s failed, maybe the battery
was connected in reverse. Maybe the problem Pierre now complains of was the
one he was trying to fix and maybe there is another fault in the unit that
is causing all the trouble.
Like you, I often upgrade when doing repairs. Most makers use the cheapest
components they can just get away with and that keeps me in a job. My
repair stock consists of the best available semis in each category, bought
from Farnell or authorised agents only!
The germanium transistor in my camera flash has been there for 25 years and
still going strong. It looks similar to an AC128, if it failed what do you
suggest for a substitute? An RCA brand MJ15003?
Regards, Phil
Trevor Wilson
>
> The germanium transistor in my camera flash has been there for 25 years
and
> still going strong. It looks similar to an AC128, if it failed what do you
> suggest for a substitute? An RCA brand MJ15003?
>
-:) I've got some AC128s, if you ever need 'em.
Robert Strand
> The germanium transistor in my camera flash has been there for 25 years and
> still going strong. It looks similar to an AC128, if it failed what do you
> suggest for a substitute? An RCA brand MJ15003?
Crikey Phil that comment's a bit over the top :O.
Regards
Rob
Phil Allison
-
> http://www.rageaudio.com.au
Hmmm, what upgrades do you use them for?
Could use the odd one in old Phase Linear's VI limiters. They use switching
types there - 2N1304/5. I'll let you know.
Regards, Phil
>
>