"I am planning to use 2N 3055/ MJ 2955 pair in the output stage with rail voltage of + - 40 volts."
Ouch!
They are only rated at 60Vceo
"Is this voltage ok for this pair?"
What do you think?
"For safety shall I add another pair parrallel to em."
All fine and good, it just won't help the voltage problem.
How much is safe for the 3055/2955 pair?
Subtract 10% for line voltage regulation and then divide by 2,
±27V
Ouch!
They are only rated at 60Vceo
"Is this voltage ok for this pair?"
What do you think?
"For safety shall I add another pair parrallel to em."
All fine and good, it just won't help the voltage problem.
How much is safe for the 3055/2955 pair?
Subtract 10% for line voltage regulation and then divide by 2,
±27V
"Truthfully, they would likely hold up to +/- 40V rails, those devices are rather diverse in construction, and I believe it has been done before. "
You offering to fix his amplifier and speaker after the smoke escapes?
I thought not.
"and I believe it has been done before. "
NAD tried it.
It didn't work.
I fixed all my blown NAD by using MJ15015/16, a 120V part, the MJ15003/04 are better, and cost more.
You offering to fix his amplifier and speaker after the smoke escapes?
I thought not.
"and I believe it has been done before. "
NAD tried it.
It didn't work.
I fixed all my blown NAD by using MJ15015/16, a 120V part, the MJ15003/04 are better, and cost more.
djk said:
kilowatt did say that it wouldn't be reliable and suggested beefier devices as you did.
I suppose that kilowatt was merely suggesting that it is doable but not advisable.
I have use those devices in over 70v rails reliably and others over their specs with mixed success.
so if the original poster has no other choice, I would suggest him/her proceed with caution.
Why not use reasonable outputs like MJ15022 / MJ15023? They cost more than MJ15015/ MJ15016, but are faarrrrr better. I don't think the difference in cost is a big hit here.
No offence to djk, but the gain differences between the MJE15015 and MJ15016 are huge! I wouldn't use them for that reason alone. The other glaring issue is the hFE vs collector current curves. MJ15022/MJ15023 are much flatter as they are a multi-emitter type, Ft is much higher as well. On top of that, these were designed for audio output applications.
-Chris
No offence to djk, but the gain differences between the MJE15015 and MJ15016 are huge! I wouldn't use them for that reason alone. The other glaring issue is the hFE vs collector current curves. MJ15022/MJ15023 are much flatter as they are a multi-emitter type, Ft is much higher as well. On top of that, these were designed for audio output applications.
-Chris
"Why not use reasonable outputs like MJ15022 / MJ15023? "
At the time they cost 3X as much as the MJ15015/16
" I don't think the difference in cost is a big hit here. "
No, not for DIY. But this was for warranty repair for NAD. Big difference.
"No offence to djk, but the gain differences between the MJE15015 and MJ15016 are huge!"
No different than the 3055/2955, ever notice that the Motorola curves for both pairs are on the same sheet?
" No offence to djk, but the gain differences between the MJE15015 and MJ15016 are huge! I wouldn't use them for that reason alone. "
Gee, the published data shows the same gain at 7A, 80 vs 90 at 1A, and 120 vs 140 at 0.1A
"On top of that, these were designed for audio output applications."
Complementary Silicon
High-Power Transistors
. . . PowerBase complementary transistors designed for high
power audio, stepping motor and other linear applications. These
devices can also be used in power switching circuits such as relay or
solenoid drivers, dc–to–dc converters, inverters, or for inductive loads
requiring higher safe operating area than the 2N3055.
Funny, Motorola thinks they designed it for high power audio, and its on their Audio spreadsheet too.
http://www.onsemi.com/site/products/parametrics/0,4446,796,00.html?Rows=MJ15015
Even today with the lower cost basis on the MJ15022/23 a manufacturer would have to think twice about using them over the MJ15016/16, the difference in cost will add over $20 at retail to a $200 receiver.
Carver used the MJ15015/16 in a couple of 250W amplifiers.
Crown uses the MJ15016/16 in several 300W amplifiers.
At the time they cost 3X as much as the MJ15015/16
" I don't think the difference in cost is a big hit here. "
No, not for DIY. But this was for warranty repair for NAD. Big difference.
"No offence to djk, but the gain differences between the MJE15015 and MJ15016 are huge!"
No different than the 3055/2955, ever notice that the Motorola curves for both pairs are on the same sheet?
" No offence to djk, but the gain differences between the MJE15015 and MJ15016 are huge! I wouldn't use them for that reason alone. "
Gee, the published data shows the same gain at 7A, 80 vs 90 at 1A, and 120 vs 140 at 0.1A
"On top of that, these were designed for audio output applications."
Complementary Silicon
High-Power Transistors
. . . PowerBase complementary transistors designed for high
power audio, stepping motor and other linear applications. These
devices can also be used in power switching circuits such as relay or
solenoid drivers, dc–to–dc converters, inverters, or for inductive loads
requiring higher safe operating area than the 2N3055.
Funny, Motorola thinks they designed it for high power audio, and its on their Audio spreadsheet too.
http://www.onsemi.com/site/products/parametrics/0,4446,796,00.html?Rows=MJ15015
Even today with the lower cost basis on the MJ15022/23 a manufacturer would have to think twice about using them over the MJ15016/16, the difference in cost will add over $20 at retail to a $200 receiver.
Carver used the MJ15015/16 in a couple of 250W amplifiers.
Crown uses the MJ15016/16 in several 300W amplifiers.
Cost
Well, it depends on what you are building. If you measure the DC current gain for MJ15015 types, you run around 20~60, MJ15016 types run over 140 typically. I do not see amplifiers designed for sound quality using MJ15015/15016 "economy" (Motorola's words) transistors. I've done warranty service for over 20 yrs. The failure rate for devices using MJ15015/MJ15016 is much higher. Such great makers such as Linear Power (car amps) used these. Man, they were a cheap bunch!
Now, I didn't realise you were behind the 8 ball with NAD warranty repairs, but if I'm charging a customer, they get decent parts. For warranty, I agree, you use what is approved. No argument there.
As far as the 3055/2955 pair, they aren't great. If you use the "H" versions, they are rated for 100V C-E (could be 80V, memory is fuzzy here). But, you are talking about a warranty service issue, which I don't think any of us will argue with.
BTW, quantity manufacturing pricing will cut that price difference considerably. (To the MJ152XX series)
Now, I was Canadian Factory warranty for Carver for years. MJ15015/15016 wouldn't survive in a Carver. They also used MJ15024/15025. These in all the TO-3 based amps. The cubes are another matter, Carver learned fast.
Anyhow, we are talking specifically about a single DIY service issue. Why not use the better part? This is what we're all about aren't we??
Just out of curiousity, have you actually measured these different transistors yourself? For over 25 yrs, I needed to install matched sets. The data sheets aren't always a good indicator of actual parameters.
Also, please don't interpret this as a personal attack, just observations on parts over the years. I don't think we are in disagreement here.
-Chris
Well, it depends on what you are building. If you measure the DC current gain for MJ15015 types, you run around 20~60, MJ15016 types run over 140 typically. I do not see amplifiers designed for sound quality using MJ15015/15016 "economy" (Motorola's words) transistors. I've done warranty service for over 20 yrs. The failure rate for devices using MJ15015/MJ15016 is much higher. Such great makers such as Linear Power (car amps) used these. Man, they were a cheap bunch!
Now, I didn't realise you were behind the 8 ball with NAD warranty repairs, but if I'm charging a customer, they get decent parts. For warranty, I agree, you use what is approved. No argument there.
As far as the 3055/2955 pair, they aren't great. If you use the "H" versions, they are rated for 100V C-E (could be 80V, memory is fuzzy here). But, you are talking about a warranty service issue, which I don't think any of us will argue with.
BTW, quantity manufacturing pricing will cut that price difference considerably. (To the MJ152XX series)
Now, I was Canadian Factory warranty for Carver for years. MJ15015/15016 wouldn't survive in a Carver. They also used MJ15024/15025. These in all the TO-3 based amps. The cubes are another matter, Carver learned fast.
Anyhow, we are talking specifically about a single DIY service issue. Why not use the better part? This is what we're all about aren't we??
Just out of curiousity, have you actually measured these different transistors yourself? For over 25 yrs, I needed to install matched sets. The data sheets aren't always a good indicator of actual parameters.
Also, please don't interpret this as a personal attack, just observations on parts over the years. I don't think we are in disagreement here.
-Chris
TIP35C and TIP36C may be an option and they aren't much more expensive than TIP3055/TIP2955
Most car-audio amplifiers today use TIP35C and TIP36C. 80Wrms@4 ohms and 120Wrms@2 ohms are reliably achievable with a single pair. 75Wrms@8 ohms may be also achieved with a single pair and +-40V rails
Most car-audio amplifiers today use TIP35C and TIP36C. 80Wrms@4 ohms and 120Wrms@2 ohms are reliably achievable with a single pair. 75Wrms@8 ohms may be also achieved with a single pair and +-40V rails
I've run pspice simulations using 40 +/- rails running a simulated speaker load with emf clamps on the output and never saw 60v + deltas on the emitter to collector driving 4 ohm loads up to 75 watts which is a good peak (a-litle over 50 watts rms) value to shoot for with the tip3055/2955 pair.
I know the mj/2n combo are 115w but all I have are tip's and so this is all I have ever tested.
I did all this only becuase I have a bunch Onsemi tip3055/2955 pairs and wanted to know when there low vceo would be an issue, and have since built a few amps with no troubles yet.
But, the reliability of an amplifier is the responsibility of the designer/builder and not me. So if you think there may be a problem and it stinks up your work station with smoke don't be surprised.
Hi Easyamp,
The problem occurs when the amp is unloaded and the outputs get driven to clip. This can happen at turn on before things stabilise, or when the speakers are disconnected & the volume gets cranked. A manufacturer will rely on supply sag when the amp is loaded.
Ain't service wonderful?
-Chris
The problem occurs when the amp is unloaded and the outputs get driven to clip. This can happen at turn on before things stabilise, or when the speakers are disconnected & the volume gets cranked. A manufacturer will rely on supply sag when the amp is loaded.
Ain't service wonderful?
-Chris
In case you already have a lot of 2N3055/MJ2955, or TIP3055/TIP2955, or MJE3055/MJE2955, their reliability could be seriously improved by using two sets of two series devices [totalling 4 devices of each polarity per channel]. That would allow for up to +-60V rails and down to 4ohm operation I think, maybe 150W@8 ohms and 250W@4 ohms if power supply sagging is kept small
This schematic appeared some days ago in a post titled 'Crown XLS' and shows a pretty simple way to use output devices in series connection [with the help of bootstrapping capacitors to improve clipping behavior I think] :
http://wardsweb.org/audio/docs/MK 24 2400 schematic2.jpg
The Double Barreled Leach Amp shows a similar way to use series connected output devices but also with series VAS and triple darlington output :
http://users.ece.gatech.edu/~mleach/superamp/circuit.pdf
Both circuits should provide voltage drop sharing and equal dissipation for upper and lower device of each series pair [compare MJE3055/MJE2955 SOA at 60V to SOA at 30V]
http://wardsweb.org/audio/docs/MK 24 2400 schematic2.jpg
The Double Barreled Leach Amp shows a similar way to use series connected output devices but also with series VAS and triple darlington output :
http://users.ece.gatech.edu/~mleach/superamp/circuit.pdf
Both circuits should provide voltage drop sharing and equal dissipation for upper and lower device of each series pair [compare MJE3055/MJE2955 SOA at 60V to SOA at 30V]
The first schematic is huge, maybe hard to understand at first, and scanning is very poor, but the interesting part of the circuit is very simple, just look at the stacked pairs of output devices, the stacked pairs of drivers and the R divider [between supply rails and output] with bootstrap C that drives the base of the outer drivers. Actually I haven't seen the remaining pages of the schematic but I think that the bases of the inner drivers are driven from a traditional VAS and a Vbe multiplier, thus making the circuit simple and keeping a moderate component count
I did some testing last night to see loaded and unloaded voltage swing and only under extreme output and low z did I get a noticeable increase voltage swing on the output when I disconnected the load but it was less then a volt rms. I did these test on my proto board amp bjt tester it's a Sziklai pair out put with standard ccs diff and vas.
I also did this same test on 2 recievers curious about commercial stuff and had similar results. So I think the biggest problem would be only if you had an amplifier that needed time to settle after turn on or off when one of the rails maybe seen on the output.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.