hi everybody...
is possible to use this amplifier with a 40+40 Vac transformer(56+56Vdc)???
i think that a transistor change is needed...
i want to change all the bc546 with MPSA42, and the output transistors with MJ21193/4
in place of the MJE15034/35 i allready have some pairs of MJE15030/31
do you think that these mods are possible, and with this power supply, the amp can work correctly???
thanks....
http://sound.westhost.com/project3a.htm
is possible to use this amplifier with a 40+40 Vac transformer(56+56Vdc)???
i think that a transistor change is needed...
i want to change all the bc546 with MPSA42, and the output transistors with MJ21193/4
in place of the MJE15034/35 i allready have some pairs of MJE15030/31
do you think that these mods are possible, and with this power supply, the amp can work correctly???
thanks....
http://sound.westhost.com/project3a.htm
Flavio88 said:
As Elliot write in his description MJL21193 / MJL21194 is a good replacement.
And MJE15030 / MJE15031 is the same as MJE15034 / 35.
So no problem with this.
Also replace BC546 with MPSA42 is good!
MJL21194 can take: +/- 125 V
MJE15030 can take: +/- 75 V
MPSA42 can take : +/- 150 V
====================================
But Elliot says:
In short, if you want to use higher voltage than 42V
do not use any speakers with lower imp than 8 Ohm!
You should be aware, that using ONLY 1 PAIR of output transistors
and try to have too high Supply voltage (=much more power)
will make it too heavy for those 2 devices.
They will destroy.
MJL21193 MJL21194 are good.
But they can not do too much.
They will simply get too hot.
Does not help if you use 100 kg of heatsinks.
They will will smoke
============================================
Advice.
If you want to make more power,there is other projects
with many output transistor pairs.
Have a look at Nelson Pass high power amplifier, for example.
He uses here 24 big output transistors!
He does not do this because he likes to - he does it because he must ....
http://passlabs.com/np/400a-4000.tif
You can come a long way with +/- 56 V
if using two MJL21193 and two MJL21194.
There are several good amplifiers at forum, for higher Watt
that parallel several output transistors.
2 + 2 would be good for this.
But I don't think you should try this with P3A.
You need to change many other things in the amplifier to get this to work.
=========================================
Maybe someone at our forum already have built a higher power version of P3A?
I do not know?
Regards
lineup
--------------------------------
http://www.onsemi.com/pub/Collateral/MJL21193-D.PDF
MJL21193 / MJL21194
Max 15 Ampere
Temperature Range, Max: 150 degrees C
Thermal Resistance, Junction−to−Case: 0.7
To get the p3a to operate on higher supply rails you will most definately have to go with an emitter follower output stage which can easily be implemented into the original design with few parts changes. To try and build a high power psa utilizing the sziklai pair will take much more effort to stabalize.
Colin
Colin
Why not use Rod's other amp, specifically for higher voltages?
http://sound.westhost.com/project68.htm
http://sound.westhost.com/project68.htm
paulb said:
I think the first question friend Flavio should ask himself
is:
Do I really need more power??
P3A is giving a lot of power.
All needed for 90% of all audio sound systems.
----------------------
It is my firm believe, that many of these 'power wattage hungry' guys, often the young ones,
fool themselves.
----------------------
It is a bit like young men:
They drive their cars like they were chased by something to escape.
Most persons killed in automobile crashes are 16-25 years old.
Maybe they are trying to impress other guys
- show manhood or something or impress on girls ...
When you are an older lady or and older man
you have learnt, that you will get where you are going
even if drive in nice and proper speed.
No need to risk to kill yourself, just to save 10 minutes or so!
When the use of these 10 minutes will be wasted anyway,
just watching TV, reading a paper or plain doing nothing
----------------------
These Power Hungry Maniacs:
I think,
They have no idea of how much power they would have use for.
Because they can not do a proper math, calculation and estimation
of the power requirements for their speakers.
----------------------
If I build a 150-300 Watt amplifier I know for sure
I would never have any use for, in extreme loudness playing,
more than 20-50 Watt.
Because I have speakers with good efficiency: decibel / Watt
I need for example only like 10 Watt RMS to reach 100 dB SPL, Sound Pressure Level
here in my small room.
And I live in a house with many apartments = neighbours behind my walls.
They would call The Police Force upon me, if misbehaving.
***
with 'good efficiency in my case, I mean,
my regular speakers used, have a sound output of:
90-92 dB SPL per watt
.... this means at 10 Watt = 100-102 dB Sound Pressure
Regards
lineup - wants to remain in good relation with his close neighbours
.... and stay free from The Local Police and out of jail
Hi
I agree with Colin, the output impeadance of an emitter follower is inherently less than that of the common emitter or CFP, consequently there is no voltage gain from the emitter follower circuit. CFP or sziklai circuit is much more load dependent than an emitter follower, and this is important when driving a reactive load like a speaker.
I agree with Colin, the output impeadance of an emitter follower is inherently less than that of the common emitter or CFP, consequently there is no voltage gain from the emitter follower circuit. CFP or sziklai circuit is much more load dependent than an emitter follower, and this is important when driving a reactive load like a speaker.
Hi,
Sorry for my bad english, I will do my best.
It is possible use the P3A with 42 volt rails and a 4 ohm load.
I asked it to Rod and it was no problem, stability was not affected.
You have to double the outputtransistors and add resistors in the emitters to force current sharing. Take a look at Project 03, there you can find the explanation wich is still correct.
With the 42 volt you will end with an output around 150 to 180 watts.
If you want to use 56 volt you will end with something about 350 watts. For that kind of power you will need 4 parallell outputtransistors. If you read the artice of Rod, design guidelines-amplifier desing, section output current you will see that it is not that simple. I think that you better take an other design if you realy need so much power. The AV400 for example?
Sorry for my bad english, I will do my best.
It is possible use the P3A with 42 volt rails and a 4 ohm load.
I asked it to Rod and it was no problem, stability was not affected.
You have to double the outputtransistors and add resistors in the emitters to force current sharing. Take a look at Project 03, there you can find the explanation wich is still correct.
With the 42 volt you will end with an output around 150 to 180 watts.
If you want to use 56 volt you will end with something about 350 watts. For that kind of power you will need 4 parallell outputtransistors. If you read the artice of Rod, design guidelines-amplifier desing, section output current you will see that it is not that simple. I think that you better take an other design if you realy need so much power. The AV400 for example?
lineup said:
lineup,
Would you help me to know how much power I need?
I am using 50Wpc for my woofers and it is good for moderate levels. Whilst listening I set my volume control for normal moderate listening and kept it there then put my preamp onto the bench. It has an active low-pass filter and dipole EQ.
I then went through my CD collection to find bass peaks. Then I factored in the gain of my current power amp and came to these (estimated) conclusions.
Note that with the EQ, 20Hz is more that 20dB higher than 200Hz.
50W will work for moderate levels with most music but it is cutting off the peaks.
According to my scope I would need 500W for moderate levels that will not cut off the peaks (especially where sustained 20Hz is available, etc.) -or- will play loud and cut off the peaks.
I estimate that 5kW will play loud but will not cut off the peaks.
So am I being unreasonable? or does the dipole EQ make all this power necessary? and, Is it OK to cut off the peaks?
Hi,
EQ uses up Xmax of your bass speaker and needs extra power.
+20db of EQ is ten times the voltage and one hundred times the power compared to the "flat" signal sent to the rest of the system.
If the maximum signal sent to the system is 50W then the bass will potentially see 5kW with +20db of EQ.
You must redesign you bass to produce what you want.
Either use bigger and lower frequency drivers that need less or zero EQ or find drivers that can handle +20db of EQ (5kW) to match the loudness of the rest of the system, or just accept that the bass is going to overload on any level much above a whisper.
EQ uses up Xmax of your bass speaker and needs extra power.
+20db of EQ is ten times the voltage and one hundred times the power compared to the "flat" signal sent to the rest of the system.
If the maximum signal sent to the system is 50W then the bass will potentially see 5kW with +20db of EQ.
You must redesign you bass to produce what you want.
Either use bigger and lower frequency drivers that need less or zero EQ or find drivers that can handle +20db of EQ (5kW) to match the loudness of the rest of the system, or just accept that the bass is going to overload on any level much above a whisper.
AndrewT said:
OK, I can see a point there: that there's no need to over drive the speaker.
I use 98dB/W/m 15" pro drivers. Only downside is the (slightly) high fs.
So I need to find how much power drives the speaker to the limit. The 50W doesn't (though it may exceed Xmax). How can I find this out? it would be frequency dependent, wouldn't it?
Hi Jnb,
yes, it is VERY frequency dependant.
If you exceed Xmax with 50W then you have overloaded the driver with just 50W.
You have 98db/2.83V/m and high Fs, this would indicate a PA type driver designed for music. They do not make good sub-bass speakers for exactly the problem you have identified.
They overload very easily on low frequency signals.
I bought a Tannoy B950 as my sub-bass. It was nearly a big mistake.
After a lot of time using Win ISD and testing the T/S parameters I found an alignment that cost almost nothing to implement that brought up the 20Hz response by about +18db without any extra power input and using the same low Xmax of this PA type speaker.
Search to see how I did it. More info in DIYsubwoofers.
yes, it is VERY frequency dependant.
If you exceed Xmax with 50W then you have overloaded the driver with just 50W.
You have 98db/2.83V/m and high Fs, this would indicate a PA type driver designed for music. They do not make good sub-bass speakers for exactly the problem you have identified.
They overload very easily on low frequency signals.
I bought a Tannoy B950 as my sub-bass. It was nearly a big mistake.
After a lot of time using Win ISD and testing the T/S parameters I found an alignment that cost almost nothing to implement that brought up the 20Hz response by about +18db without any extra power input and using the same low Xmax of this PA type speaker.
Search to see how I did it. More info in DIYsubwoofers.
Thanks, Andrew. The search turned up a good read. I'm not sure I'm ready to try cone mass mods with the OB and the OB itself precludes tuning the response but I am wary of the laws of nature and will have to "suck it and see".
My Delta15LFAs have a very similar FR plot to the Alphas which are popular with OBs.
I found this
So I think I'll build a 100W amp and if I can't make it satisfy me then I'll just need to ease up on the filter, or pull out the ol' saw and router. Currently, both of these, I think, are really pushing things to their limit.
My Delta15LFAs have a very similar FR plot to the Alphas which are popular with OBs.
I found this
I don't mind going past Xmax, especially as the pro drivers were built for it (even if they weren't really).James D.
The Eminence 15" drivers all achieve 95dB at 1W, 1m - some will do more. Measured in practice and according to the spec sheet. All of them hit xmax limits of between 10 and 40 W at 40Hz on an OB and all are rated between 150 and 600W per driver... so they are fine for OB use with low power amps - which is what we all tend to use
So I think I'll build a 100W amp and if I can't make it satisfy me then I'll just need to ease up on the filter, or pull out the ol' saw and router. Currently, both of these, I think, are really pushing things to their limit.
AndrewT said:
Hi Andrew,
Yeah, PA drivers always use "W" surrounds because in concert PA use (not disco use) the range below about 40 hz is simply not used and usually filtered out quite steeply. Thus, PA driver manufacturers rarely extend the response down lower, preferring to favour power handling and SPL, thus XMax isn't as big of a concern as say heat dissipation.
If, as you've witnessed, the PA driver is put into duty with recorded music, by a typical DJ with EQ knobs to play with, you can expect those bottom octaves to be extremely boosted trying to get the subs to produce the same sound the DJ gets at home or in their car, and the surrounds end up torn off the cone.
That also commonly happens when PA techs replace speakers in bass cabinets without regard for the T/S parameters, which many PA techs have never heard of. All they know is, "it fits physically, so lets see what it sounds like. Hmm...
not much bottom end, better boost up 20 and 40 hz on my EQ. That's better." 
..Todd
Hi Jnb,jnb said:
adding weight reduces Fs from typical sound reinforcement frequencies to nearer typical Sub-Bass duty requirements.
But it also increases the Qts.
For my drivers I roughly doubled the weight of the cone/moving mass and would expect both Fs & Q to change by sqrt(2). But when I tested the T/S at each stage of increasing the mass in steps of 20gms. I found that Fs did indeed change as predicted but that Q only changed by about 90% of the expected change.
The resulting Fs & slightly lower Q gave a nice set of graphs when input into WINisd. It took lots of re-testing and re-simulating to ensure the low Xmax of the two drivers could be used to best advanatge. I am very pleased with the result.
The OB would suit the raised Q resulting from added mass. Lots of drivers solves the Xmax limitation. Lot's of drivers also increases sensitivity and you may find that putting a 25W amp on EACH driver will give lots of sub-bass output, if you have enough of them.
Is there some benefit to fs being lower than the signal being fed to it?
Increasing cone mass will reduce the power needed at fs, but will reduce overall sensitivity as well. I can't imagine it requiring less cone travel for a given SPL.
I'm trying to correlate all this to the context of the OB. If I change the behaviour of the driver, I'll be developing an EQ to suit it.
Increasing cone mass will reduce the power needed at fs, but will reduce overall sensitivity as well. I can't imagine it requiring less cone travel for a given SPL.
I'm trying to correlate all this to the context of the OB. If I change the behaviour of the driver, I'll be developing an EQ to suit it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.