Can someone check out the AV800
document below (see link) and tell me
if it's possible to bridge two channels
to form a single channel amplifier of
higher power.
I'm not 100% sure, but I thought
I read somewhere on this forum
last year that you will exceed safe
operating parameters of the output
stage transistors. What modification
can be done to make it work ?
Can I reduce rail voltages to make
bridging possible ?
Whether bridging is or isn't recommended, what is an elegant
method to create a bridging module
for this amplfiier with differential input
stage ?
1K watt pic
http://www.aussieamplifiers.com/1kwamp.htm
AV800 watt document
http://www.aussieamplifiers.com/downloads/AV800.pdf
** Last question **
If I want to connect 20 output stage
transistors per rail (40 transistors per channel), do you think the pre-drive
stage is able to drive so many transistors ? if not, what mod needs
to be done ?
document below (see link) and tell me
if it's possible to bridge two channels
to form a single channel amplifier of
higher power.
I'm not 100% sure, but I thought
I read somewhere on this forum
last year that you will exceed safe
operating parameters of the output
stage transistors. What modification
can be done to make it work ?
Can I reduce rail voltages to make
bridging possible ?
Whether bridging is or isn't recommended, what is an elegant
method to create a bridging module
for this amplfiier with differential input
stage ?
1K watt pic
http://www.aussieamplifiers.com/1kwamp.htm
AV800 watt document
http://www.aussieamplifiers.com/downloads/AV800.pdf
** Last question **
If I want to connect 20 output stage
transistors per rail (40 transistors per channel), do you think the pre-drive
stage is able to drive so many transistors ? if not, what mod needs
to be done ?
It's always safe to connect an amp in bridge mode, but with double the normal load impedance.
If the amp is rated to be safe down to 4 ohms with one channel, or in stereo operation, it's always safe to bridge it into 8 ohms, as long as the amp isn't designed in such a way to make that impossible (IC amp with outputs bridged internally, etc.)
However, bridging the AV800 into 8 ohms probably wouldn't be recommended, unless you reduce the rail voltages.
Dividing the recommended rail voltages by 1.414 (square root of 2) would be a good start, which would keep the power dissipation in the output stage when bridged into 4 ohms pretty close to what it would be if you had the higher rail voltages and a single channel running 4 ohms.
However, power output will be reduced accordingly, of course.
However, if you want to add more output transistors, then you should definitely be able to bridge the amp with less (or even none, depending on how many you add) reduction of the rail voltages.
And yes, his predriver stage on the AV800 should be perfectly happy with 10 complementary pairs (20 transistors total) per channel - just make sure to parallel the extras like the others on the schematic with the requisite gate and source resistors, of course. If you're wiring them point to point on the sink to graft them onto his PCB design, make sure to put the gate resistors right at the gate of the mosfet, and not on the pcb.
If the amp is rated to be safe down to 4 ohms with one channel, or in stereo operation, it's always safe to bridge it into 8 ohms, as long as the amp isn't designed in such a way to make that impossible (IC amp with outputs bridged internally, etc.)
However, bridging the AV800 into 8 ohms probably wouldn't be recommended, unless you reduce the rail voltages.
Dividing the recommended rail voltages by 1.414 (square root of 2) would be a good start, which would keep the power dissipation in the output stage when bridged into 4 ohms pretty close to what it would be if you had the higher rail voltages and a single channel running 4 ohms.
However, power output will be reduced accordingly, of course.
However, if you want to add more output transistors, then you should definitely be able to bridge the amp with less (or even none, depending on how many you add) reduction of the rail voltages.
And yes, his predriver stage on the AV800 should be perfectly happy with 10 complementary pairs (20 transistors total) per channel - just make sure to parallel the extras like the others on the schematic with the requisite gate and source resistors, of course. If you're wiring them point to point on the sink to graft them onto his PCB design, make sure to put the gate resistors right at the gate of the mosfet, and not on the pcb.
Why don't you ask Anthony yourself?
With a very rough calculation I have come to this:
8 ohms, 92 V => 529 W out, 8.1 A, and 140 W losses (at least)
4 ohms, 92 V => 1058 W, 16 A out and 280 W losses (at least)
One pair can take 300 W losses, 14 A
With resitive load, one pair can almost take the whole load!
2 to 4 pairs seems reasonable if we talk technically and for home use.
Don't forget that a speaker is NOT a resistive load which means bigger losses but how much really?
With a very rough calculation I have come to this:
8 ohms, 92 V => 529 W out, 8.1 A, and 140 W losses (at least)
4 ohms, 92 V => 1058 W, 16 A out and 280 W losses (at least)
One pair can take 300 W losses, 14 A
With resitive load, one pair can almost take the whole load!
2 to 4 pairs seems reasonable if we talk technically and for home use.
Don't forget that a speaker is NOT a resistive load which means bigger losses but how much really?
Hi Thy,
Using 40 devices per channel, 80 in total, is a lot of soldering and a lot of wiring! Why not check out some higher power FETs and save all this work? When you double the number of FETs you double the transconductance, double the power handling and double the input capacitance. It's the input capacitance that you must be careful of as this is what loads the drivers and will cause additional phase lag at high frequencies. If you check out the databooks you may find bigger FETs where the input capacitance grows less quickly than the power handling. Perhaps Nelson can give some advice here.
It is not easy to tell what the operating point is of the output and driver stage from the schematic alone. It may be if you double the output devices, thus doubling the load capacitance on the drivers, that R38 may need reducing, perhaps halving to 160-ohms. But you'll need to experiment while monitoring the distortion and you'll need to check the drives power dissipation is not exceeded.
Another consideration that hasn't been addressed is power supply. When you bridge two amps you must look after the routing of the power. Use a single psu for each bridged channel and make sure the ground, V+ and V- use separate wires from the two channels back to the psu caps and star ground. Make sure the input coax ground goes back to the psu star point separately. Things can become unstable easily with bridged configurations: check out the current loop paths (see a thread in here to do with star grounding where I prattle on in more detail about this).
An easy and good method of dealing with the input signals is to use two good quality op-amps, one as a x1 buffer and the other as a x-1 buffer (simple non-inverter and simple inverter). Why use a non-inverter? To ensure the signal path is as identical as possible between the two channels. You'll need to provide a separate psu for the op-amps. This is the simplest approach IMO.
Another approach is to make a discrete LTP input amp and take the signals off the collector/drains via dc blocking caps. Again, a separate psu is needed with good regulation because the signals will be referenced to the psu rail and you don't want the psu signal polluting your channels (even though it will cancel somewhat, in theory).
Let us know what you decide to do.
Using 40 devices per channel, 80 in total, is a lot of soldering and a lot of wiring! Why not check out some higher power FETs and save all this work? When you double the number of FETs you double the transconductance, double the power handling and double the input capacitance. It's the input capacitance that you must be careful of as this is what loads the drivers and will cause additional phase lag at high frequencies. If you check out the databooks you may find bigger FETs where the input capacitance grows less quickly than the power handling. Perhaps Nelson can give some advice here.
It is not easy to tell what the operating point is of the output and driver stage from the schematic alone. It may be if you double the output devices, thus doubling the load capacitance on the drivers, that R38 may need reducing, perhaps halving to 160-ohms. But you'll need to experiment while monitoring the distortion and you'll need to check the drives power dissipation is not exceeded.
Another consideration that hasn't been addressed is power supply. When you bridge two amps you must look after the routing of the power. Use a single psu for each bridged channel and make sure the ground, V+ and V- use separate wires from the two channels back to the psu caps and star ground. Make sure the input coax ground goes back to the psu star point separately. Things can become unstable easily with bridged configurations: check out the current loop paths (see a thread in here to do with star grounding where I prattle on in more detail about this).
An easy and good method of dealing with the input signals is to use two good quality op-amps, one as a x1 buffer and the other as a x-1 buffer (simple non-inverter and simple inverter). Why use a non-inverter? To ensure the signal path is as identical as possible between the two channels. You'll need to provide a separate psu for the op-amps. This is the simplest approach IMO.
Another approach is to make a discrete LTP input amp and take the signals off the collector/drains via dc blocking caps. Again, a separate psu is needed with good regulation because the signals will be referenced to the psu rail and you don't want the psu signal polluting your channels (even though it will cancel somewhat, in theory).
Let us know what you decide to do.
I'm in the thinking stage of the project. My goal is
to layout a circuit board for one channel of the amplfiier
to build a single module. Then I can insert the modules
in the chassis to form whatever I need. 2 ch amp, 4 ch amp,
mono bridged amp, etc.
All components will be soldered/mounted on the circuit board.
Since this is my goal, I'm entertaining other ideas to
see if they would integrate into the pcb layout. I need
some feedback from the folks here....
Crazy ideas; Feel free to comment....
********************************************
1. Bridging module circuit.
2. The input stage eludes me. Do I add an op-amp
to the input stage to allow the amplifier to have
input gain adjustment capability so I can level
match *any* source to the amplifier ? Perhaps even
making a op-amp bypass switch if you don't need this option?
Or, should I just abandon this idea and build a seperate
matching pre-amplifier?
3. Do I need speaker protection circuits w/relay on the same
amplifier pcb ? This can occupy a lot of space, maybe it's best
to keep it as a seperate module ?
4. Input signal and/or output clipping indicators using
some op-amp comparators? Should I integrate this
on the pcb and have a connector to allow the LED's
to be mounted on the chassis of the amplifier?
5. Thermal sensor for each channel? If one channel is
getting too hot, do I want the amplifier to disconnect
the input signal, perhaps using relays?
********************************************
The AV800 amplifier uses IRFP240 and IRFP9240
output transistors. Is there an upgrade to these ?
I bought two of Holtons pcb's, I just need to build
the prototype after I sort all the ideas out. I have
a seperate power supply already built for benching,
it has dual toroidals with a switch to output
+75v/-75v or +150v/-150v peak. I think each transfomer
is 6 ampere. Is there a simple way to scale down
the +150v/-150v down to +110v/-110v peak?
I noticed that Holton's 1kw version of the AV800
uses the 110 volt rails. hehe
to layout a circuit board for one channel of the amplfiier
to build a single module. Then I can insert the modules
in the chassis to form whatever I need. 2 ch amp, 4 ch amp,
mono bridged amp, etc.
All components will be soldered/mounted on the circuit board.
Since this is my goal, I'm entertaining other ideas to
see if they would integrate into the pcb layout. I need
some feedback from the folks here....
Crazy ideas; Feel free to comment....
********************************************
1. Bridging module circuit.
2. The input stage eludes me. Do I add an op-amp
to the input stage to allow the amplifier to have
input gain adjustment capability so I can level
match *any* source to the amplifier ? Perhaps even
making a op-amp bypass switch if you don't need this option?
Or, should I just abandon this idea and build a seperate
matching pre-amplifier?
3. Do I need speaker protection circuits w/relay on the same
amplifier pcb ? This can occupy a lot of space, maybe it's best
to keep it as a seperate module ?
4. Input signal and/or output clipping indicators using
some op-amp comparators? Should I integrate this
on the pcb and have a connector to allow the LED's
to be mounted on the chassis of the amplifier?
5. Thermal sensor for each channel? If one channel is
getting too hot, do I want the amplifier to disconnect
the input signal, perhaps using relays?
********************************************
The AV800 amplifier uses IRFP240 and IRFP9240
output transistors. Is there an upgrade to these ?
I bought two of Holtons pcb's, I just need to build
the prototype after I sort all the ideas out. I have
a seperate power supply already built for benching,
it has dual toroidals with a switch to output
+75v/-75v or +150v/-150v peak. I think each transfomer
is 6 ampere. Is there a simple way to scale down
the +150v/-150v down to +110v/-110v peak?
I noticed that Holton's 1kw version of the AV800
uses the 110 volt rails. hehe
traderbam said:
Is this what you mean?
The concept is to feed channel #2 amplifier a 180 degree signal
that channel #1 receives...
source signal #1 --> input AV800 channel #1
source signal #1 --> input buffer #1 (non-invert) ---> buffer #2
(invert) --> input AV800 channel #2
If this is true, then the part that confuses me is the balanced
input stage of the AV800.
Does input buffer #1 also need to be balanced?
How do I connected buffer #2 to the balanced input
of AV800 channel #2 ? Do I convert it from balanced input
to un-balanced input so the op-amp interfaces?
traderbam said:
My electronics is rusty, I haven't focused on it in almost 15 years.
What is LTP ?
IRF does offer a IRFP250N which is a good upgrade from the 240. You will have to use the PChannel 9240's though as thier is no upgrade from IRF for those.
I too am building a pair of AV800's and so is Jason L. I have machined the center of the boards out as A.H. recommends and am now in assembly mode. I have posted pics elsewhere on this site of the Chassis I built for a stereo pair driven by two 1KVA Toroids.
At over 400 Watts into 8 Ohms why would you want to bridge them? This would require two dedicated 15 Amp circuits to provide enough power to run them!
Anthony
Do not use the IRFP250N. IR Mosfets with `N' markings mean that they are meant for switching applications. The current rating is higher but dissipation is lower than equivalent devices without the `N' markings. In any case, since there is not yet an upgrade to IRF(P)9240, stick to its complementary, the IRF(P)240.
In case you want to always use the amp in bridged mode, you can either use a true balanced singal or, since the amplifier's inputs are differential pairs, feed the signal to the inverted input of the second amp channel. Anthony posted the values of some passive components to be added for this mod in a thread relating to his N-channel amplifier which may be instructive in this context. Email Anthony and he is sure to reply. I have periodically corresponded with him regarding some mods to his N-channel design. He is a very helpful guy.
Yes, why do you want to bridge an already very powerful amplifier? I am also looking for a design (single channel or bridged) that can put out 1200watts RMS into 8 ohms to drive EVX180A woofers in a folded horn design. I am looking for designs with full protection, particularly short circuit and overload protection inbuilt. DC protection, thermal protection etc., can be added to any amplifier through separate modules.
In case you want to always use the amp in bridged mode, you can either use a true balanced singal or, since the amplifier's inputs are differential pairs, feed the signal to the inverted input of the second amp channel. Anthony posted the values of some passive components to be added for this mod in a thread relating to his N-channel amplifier which may be instructive in this context. Email Anthony and he is sure to reply. I have periodically corresponded with him regarding some mods to his N-channel design. He is a very helpful guy.
Yes, why do you want to bridge an already very powerful amplifier? I am also looking for a design (single channel or bridged) that can put out 1200watts RMS into 8 ohms to drive EVX180A woofers in a folded horn design. I am looking for designs with full protection, particularly short circuit and overload protection inbuilt. DC protection, thermal protection etc., can be added to any amplifier through separate modules.
Samuel Jayaraj said:
I'm working on the speaker system .. slowly .. it doesn't
require monster power to drive ........
but, I have 19
18" woofers in storage...
1kw - 2kw of amplification is pretty common in
car audio today for a reasonable price. JBL has a
new 6kw car amplifier.
therefore, you'll need about 1kw - 2kw (home amplifier)
to drive the new high x-max monster subwoofers.
thylantyr said:
I'm working on the speaker system .. slowly .. it doesn't
require monster power to drive ........
but, I have 19
18" woofers in storage...
1kw - 2kw of amplification is pretty common in
car audio today for a reasonable price. JBL has a
new 6kw car amplifier.
therefore, you'll need about 1kw - 2kw (home amplifier)
to drive the new high x-max monster subwoofers.
I am into Hi-Fi and Pro-Audio in quite a big way. However, I am yet to come across 18" woofers with bigger than 5" voice coils; Beyma subs with 5" voice coils are rated at 800watt RMS. The more common 18" Pro speakers all have 4" voice coils and are rated between 500 and 600watt RMS. EV claims that their EVX180A, though rated at 600watts can actually be brought to its limits with a 1200watt RMS amplifier.
Let me see if I can obtain these "new high x-max monster subwoofers" and load them in all of 8 folded horn bass bins that comprise my present system, apart from the mids and highs.
However, I don't think I'd want to waste my time and money!!!
Banned
Joined 2002
Hummm.. where did you get this information from? I have read most of the IR technical docs and did not come across this. I have even asked IR for direct replacements in an Audio desing application and was referred to the new N series parts.
If you read the tech docs both the N series and non N series are designated as High speed switch devices.
BTW all transistors are essentially switches
Anthony
Samuel Jayaraj said:
Partial list of high x-max monster
subwoofers;
Resonant Engineering
http://respl.com/xxx.html
18XXXQ1.2
Xmax 32mm one way
1600w rms
http://www.edesignaudio.com/ep/aseries.htm
http://www.adireaudio.com/diy_audio/drivers/adire/tumult.htm
Tumult
Xmax = 34mm one way
Pmax = 1600w
http://www.ddaudio.com/caraudio/woofers.asp?series=9900a
Xmax = ~30mm one way
1500w rms
http://www.ddaudio.com/caraudio/woofers.asp?series=9500e
Xmax = 29mm one way
1200w rms
http://www3.sk.sympatico.ca/hebig1/index2.html
Xmax = 25mm one way
1200w rms
http://www.jlaudio.com/subwoofers/w7preview.html
13w7
Xmax = 32mm one way
1000 watt
Stryke AV15
http://www.stryke.com/store/shop.cgi/page=preorder.html/SID=PUT_SID_HERE
Xmax = 23mm
500w rms
****************************
Even my Lambda TD15H, Apollo version can handle 1kw of power.
But, Xmax = 10mm. It's not really
a monster subwoofer, it's a wideband
driver. The VC barely gets warm
when driving it with my 600w amp.
http://www.lambdacoustics.com/drivers/TD15H.html
.... fun stuff .....
JasonL said:
Power output of an amplifier
is proportional to my finger
on the volume knob
Easier to scale down the big
amplifier design, than to scale
up. The 1kw amp can be
configured to operate at 250w
easily, but it will be much harder
to take a 250w design and scale up
to 1kw.
Since I will eventually build my
own PCB on the CAD system,
might as well think big since the
PCB cost is the same
****************************
I have 600w mono bridged amplifier,
but my Lambda TD15's can go to
1kw. That's why I want to DIY
a high powered amplifier, the commerical offerings are way too
expensive. Seems like the only brands
that are affordable are some
prosound amplifiers, but still alot of $$$$$$
I'm here so you folks can bring
order to my chaos
Any interesting ideas to integrate
to the AV800 design
Banned
Joined 2002
JasonL said:
I bought a used ADCOM 555II
off Ebay for $400 US.
But if you look inside, it's just "OK",
DIY is way better.
DIY amplifier
DIY speaker system
I'm rusty in electronics, the last
amp kit I build and modded
was 12 years ago. Since then,
have not excercised my brain and
I forget alot about electronic theory
- LOL
Was 0.5 ohm stable per channel.
http://dreadlordpk.50megs.com/2kw.html
This amplifier is old news and boring.
I like the Holton designs....... It seems
to fit my personality.........
Samuel Jayaraj said:
Sorry Samuel, you are completely wrong on the "N" classification. Here is a response from the IRF technical department.
Subject
What is the N designation for in the newer series HEXFETS.
Discussion Thread
Response (Diana Pinto (Americas-TAC)) 03/20/2003 04:06 PM
Anthony,
The "N" suffix designates that this is a new generation of MOSFETs - we are using a different die process for these MOSFETs newer MOSFETs which reduces the Rds(on)
Regards,
Customer (Anthony Tagg) 03/20/2003 02:48 PM
What is the N designation for in the newer series HEXFETS. ie IRFP250 moved to IRFP250N?
International Rectifier
Americas-TAC
101 S.Sepulveda Blvd
El Segundo, CA 90245
(310) 252-7105 Fax (310) 252-7903
D.Pinto - Applications Engineer
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.