sixtek said:
sixtek,
This amp is not designed for PA use. That has been clearly stated in past on a number of occasions. Those wishing to use an/this amp for PA use have been directed to other designs and advised this amp was not designed for PA use.
Regards,
John L. Males
Willowdale, Ontario
Canada
14 January 2007 09:47
Re[01]: Post #1357 14 January 2007 08:58
Opps I made a typo error. Correction to 2nd last sentance in above quote is:
Paralleling the secondaries would double the secondary current of the 20VAC types to 0.09A (90ma) and 0.12A (120ma) for a 14V secondary.
Regards,
John L. Males
Willowdale, Ontario
Canada
14 January 2007 10:22
keypunch said:
Opps I made a typo error. Correction to 2nd last sentance in above quote is:
Paralleling the secondaries would double the secondary current of the 20VAC types to 0.09A (90ma) and 0.12A (120ma) for a 14V secondary.
Regards,
John L. Males
Willowdale, Ontario
Canada
14 January 2007 10:22
Sorry Quasi and Keypunch, then I cancel my previous comment.
On the margin from the pictures of the already built amps it seems to be that these are for PA-purpose... Not for HIFI. Let's read back in the thread, you will find some pictures posted by members.
And I wondering why people tend to build such high power (200-400W) amplifier for HIFI?!
On the margin from the pictures of the already built amps it seems to be that these are for PA-purpose... Not for HIFI. Let's read back in the thread, you will find some pictures posted by members.
And I wondering why people tend to build such high power (200-400W) amplifier for HIFI?!
sixtek said:
sixtek,
Did you make a typo error such that should of read:
"... it seems to be that these are for HIFI-purpose... Not for PA."
With respect to people tend to build such high power for HI there are a few reasons. The biggest reason is handle the dynamic peeks of music material, usually for subwoofer based amplifiers. The secondary reason is speaker low speaker efficiency, speaker impedance dips and passive crossover losses whilst having the dynamic headroom for the peaks. Basically this is about power reserve for peaks, not letting the output devices run too warm and ability to handle the reactive and low speaker impedance dips. That said bear in mind a the 350W version is the 4 ohm speaker impedance rating. Into 8 ohms the power rating is about 180W. When using a speaker rated at 8 ohms impedance one generally designs the amp to handle 4 ohm impediances to account for phase angle and impedance changes/dips in the speaker load the amplifier sees.
I am not an expert by any means, so if you need to understand the impact of speaker phase angle and related matters of speakers/passive crossovers and amplifier load handling ability you will need to seek understanding from those that understand that material. I know enough at this point that it exists and how to factor that into PSU and device selection.
Now as much as you are may be aware of the 350W and 500W version into 4 ohms, there has been a healthy demand for lower power versions of the amp as well. I know this as I created variations of the PCB with only one pair and two pair of output devices over a year ago. I know there were a number of people that downloaded those versions of the PCB. This means there are plenty of people that like the quasi design and do not feel the need for as much reserve power handling for their needs or have more efficient speakers.
Please be aware higher power version amplifiers tend to generate more discussion as higher power versions of any amplifier tend to have greater concerns, PSU, heatsink and part selction demands to ensure the finished project is stable. Lower power versions do not generally pose all of these, if any, considerations to ensure a stable final outcome. For sure if you undersize a smaller design it can have many of the same issues high power designs do, but generally do not as there is better and more suitable parts and heatsinks to enable a successful end result.
Quasi recently posted a few days ago a compact PCB board that uses TO-220 based output devices. The TO-220 devices are smaller than the TO-247 devices and generally have less power handling than the TO-247 devices. As you will see looking back in last few days the reason for the TO-220 version was for HT and/lower power needs many have made requests to quasi directly for. The compact board only has two output device pairs.
If you look back in last week or so there was a request for a two pair version of the TO-247 based board for lower power needs. Shawn and I posted versions of the two TO-247 output pairs and many others downloaded those versions. There was a few months ago a request again for the two output pair version. It seems people were not aware of the posting put up where this and the single pair version are located.
I hope I have at least opened your eyes to the reference impedance used for the quasi designs and the application reasons for higher and lower power version of the quasi amp and why higher power designs by their very demanding nature tend to see more discussions on any amplifier thread than low power versions of same design.
Regards,
John L. Males
Willowdale, Ontario
Canada
14 January 2007 11:20
sixtek said:
That's a really good question and one that maybe needs to be looked at closely.
With modern digital recordings it is not unusual to experience a dynamic range approaching 90 dB or more (not rock & roll etc.) A piano, organ or orchestral piece could have a very great dymanic range when you consider how quite some passages can be compared to the loud bits.
Then consider the efficiency within a typical listening room. Whilst the speaker may be rated at 88dB or even 90dB @ 1 watt, this is generally at 1 metre. I sit about 3 - 4 metres away so at 1 watt I might be hearing only 78 or so dB (discounting reflections etc.) Also at 35Hz the efficiency might be a further 6dB down. Now 78 dB is quite loud, no question there, but to me it's the ability to reproduce accurately the faintest sound ("ting") to the loudest (grand piano, organ, drums). It is also reasonable to expect a concert like performance from time to time. I'm not talking about overly loud music, just the ability to faithfully reproduce a competent orchestra.
Given the above, 1kW per channel may not be enough, if absolute accuracy is your thing. But life is a compromise and this is different for different people. For me 200 watts per channel into 8 ohms is enough, for others on this thread 600 watts into 4 ohms isn't. I know though that for some of the classical pieces I listen to at moderate levels, some of the passages are compressed as the power supply starts to reach its delivery capability (thankfully for the output stage).
So large power reserves is not necessarily about loud. To me it's more about dynamic range.
PA applications generally (not always) deal with content with a smaller dynamic range. What is important here is high average power capability and robustness. Remember also that the efficiency of good PA speakers may be 6 or 10dB better than HiFi speakers. But they generally have a narrower frequency range and poorer accuracy. So for small venues where pop music is the fare (parties) 200 watts might be plenty.
Cheers
Q
sixtek said:
I like amps that are excessive and that doesn't mean power and/or size, not all the time. Tiny little amplifiers can be excessive too. I do not know why this design gets branded as a PA amp but I guess I know; The quasi comp topology seems to have been used in the past to build large PA type amplifiers. I guess the Crown DC300A is a classic example(bad rap).
I think more folks should try this design that are apprehensive of its topology. It is easy and fast to assemble and Q has employed the use of universal components easing the process. Build it well and you will have a fine sounding amplifier. Quasi complimentary is not a brand or flavor specific to PA amplifier. It's past was exploited by PA use due to the fact that high power complimentary sets were at one time not available. I think this topology has a bad rap but every topology makes a compromise somewhere to get something.
Contrary to its BIGness, Q is pumping out mini-scale versions so the timing to jump in and talk big is a little weird?
My Quasi...I love it. It is fine. Go build one and listen.
Cheers,
Shawn.
Hi,
BTW, Lars Clausen once used to insist on the fact that his MOSFET amplifiers sounded best with only one output device per rail - in a smallest version.
Can that be generalized and expected from this amplifier? I've wondered if that only was a subjectivist's impression or rather rooted in the imbalance between the FETs...
I like the 5-FET design for it's simplicity and performance, but I still tend to favor the 4-FET TO-220 board over most older (bipolar or mosfet) boards of same size and intended power range (also note the smaller parasitic capacitances inside TO-220 devices and thus potentially lower drive distortions in the VAS/driver stage). This one should be ideal to use in low power (or dynamics) applications with only one pair of FETs - where such a proposed sound quality advantage could really come in handy.
Did you make a comparison regarding sound detail quality, John?
Cheers,
Sebastian.
BTW, Lars Clausen once used to insist on the fact that his MOSFET amplifiers sounded best with only one output device per rail - in a smallest version.
Can that be generalized and expected from this amplifier? I've wondered if that only was a subjectivist's impression or rather rooted in the imbalance between the FETs...
I like the 5-FET design for it's simplicity and performance, but I still tend to favor the 4-FET TO-220 board over most older (bipolar or mosfet) boards of same size and intended power range (also note the smaller parasitic capacitances inside TO-220 devices and thus potentially lower drive distortions in the VAS/driver stage). This one should be ideal to use in low power (or dynamics) applications with only one pair of FETs - where such a proposed sound quality advantage could really come in handy.
Did you make a comparison regarding sound detail quality, John?
Cheers,
Sebastian.
Hi,
The main disadvantage is that output power for a push pull pair is limited to about 30% to 40% of device power. i.e. 80W from a pair of 250W semis or 50W from a 150W pair.
If power at these levels is adequate then the simplicity of single pair is attractive and if that also gives better sound then one gains twice over.
is repeated by a number of designers.
The main disadvantage is that output power for a push pull pair is limited to about 30% to 40% of device power. i.e. 80W from a pair of 250W semis or 50W from a 150W pair.
If power at these levels is adequate then the simplicity of single pair is attractive and if that also gives better sound then one gains twice over.
Using TO3P (plastic) instead of TO3 (metal)
As I must replace two output bipolar transistors on my power amp, with the basic design very much alike the original circuit, I wonder if I should try more modern (fast) plastic types instead of the (slower) metallic types.
This seems to have been the final considerations of this thread: whether to go for faster type devices or not.
As I must replace two output bipolar transistors on my power amp, with the basic design very much alike the original circuit, I wonder if I should try more modern (fast) plastic types instead of the (slower) metallic types.
This seems to have been the final considerations of this thread: whether to go for faster type devices or not.
Hi,
if you have the knowledge and resources to test the outcome after the redesign then go ahead.
Otherwise.... leave well alone.
if you have the knowledge and resources to test the outcome after the redesign then go ahead.
Otherwise.... leave well alone.
Questions Conserning the to-220 Version
Hello Quasi,
I got some Question concerning your small-layout.
The flat pins you use for the PSU/GND/Speaker Connection have two pins in 5mm distance don't they?
Is it a good idea to add the ouput-inductor apart from the PCB-Board?
Do you know the distance of the two pads at the Input. (where the signal input is connected)?
What is the distance between the pins of the PSU-Bypass-caps?
Do all foil-Caps have 5mm distance between their pins?
The 1R0 Resistors are 2W each-aren't they? What space do they have?
You stated, you are going to use the IRF840 Mosfets on your prototype. I suspect, you are going to use them, as you dont have to buy them. I will have to buy my Mosfets, wich type schould bring the Best performance? Wich criteria are important for the to-220 Mosfets?
BTW the Jumper upon the positive supply-rail is a great idea to increase the amount of copper without loosing too much space!
Thanks alot.
Black
Hello Quasi,
I got some Question concerning your small-layout.
The flat pins you use for the PSU/GND/Speaker Connection have two pins in 5mm distance don't they?
Is it a good idea to add the ouput-inductor apart from the PCB-Board?
Do you know the distance of the two pads at the Input. (where the signal input is connected)?
What is the distance between the pins of the PSU-Bypass-caps?
Do all foil-Caps have 5mm distance between their pins?
The 1R0 Resistors are 2W each-aren't they? What space do they have?
You stated, you are going to use the IRF840 Mosfets on your prototype. I suspect, you are going to use them, as you dont have to buy them. I will have to buy my Mosfets, wich type schould bring the Best performance? Wich criteria are important for the to-220 Mosfets?
BTW the Jumper upon the positive supply-rail is a great idea to increase the amount of copper without loosing too much space!
Thanks alot.
Black
AndrewT said:
Well, alone I can't leave them, because I have to replace them.
My main concern is with heat dissipation, as plastic types are not too effective there. Most power amps are designed to use almost any parts, as long as the pcb is well done. The transistors are soldered directly to the pcb, no wires.
In any case I think I will risk less and just go for another metal TO3.
Hi
"Audible difference
Difference between 100 - 200 Watt sound system, is minimal (3dB)
Difference between 100 - 1,000 Watt sound system, is twice as loud (10dB)
"
I want to try this amp but i see that are many versions around here 😀
I want to buil the lowest power version , i think it is "nmos200-c layout & tracks.pdf"
Quasi , or enyone else, can you tell me some specifications about this ?
I em right ? it's the lowest power version ? which is the schematics ?
thank
"Audible difference
Difference between 100 - 200 Watt sound system, is minimal (3dB)
Difference between 100 - 1,000 Watt sound system, is twice as loud (10dB)
"
I want to try this amp but i see that are many versions around here 😀
I want to buil the lowest power version , i think it is "nmos200-c layout & tracks.pdf"
Quasi , or enyone else, can you tell me some specifications about this ?
I em right ? it's the lowest power version ? which is the schematics ?
thank
sek said:
Sek,
Aside from the Holton N-Channel amp that caused me stability concerns form reading others experiences, I had looked at the initial Zeta design by Lars. The initial Zeta had 4 or 5 output devices which was beyond my needs. That said, it was easy for one to reduce the number of output devices to what one needed.
I did not persue the Zeta design for a few reasons - local part availability of a couple of the transistor types, was considered more PA quality which the notion of HiFi quality suggested with only one pair, lack of PCB or one geared for surface mounting, and a certain aspect of the design that even with my (at time) limited knowledge (and even now more aware knowledge) did not agree with me, and certain parts it was very dependant on must have.
Added to this it seemed the Zeta design was more of an experiment. In end there was limited to no sustained interested in the design compared to the quasi amp that followed not to long after the Zeta when into a declined of interest compared to the limited interest when started out. On the other hand the quasi amp has continued to sustain interest and grow interest.
The second generation of the Zeta was only one pair of output MOSFETS. The PCB was only a protype board, never published basically surface mount approach. As a DIY person, my opinion is the surface mount is more problems and not as good as using the "real" components.
I am inclined to build a TO-220 version using the TO-247 board design. It is actually easy to do and I would leave the spacing as is and just move the MOSFET output device holes closer to accomodate the TO-220 devices. One reason it to allow more heatsink between the TO-220 devices which may run a bit warmed then the same power and number of devices for a TO-247 version. The extra spacing would offset the possibility warmer run temperatures of the TO-220(s) from heating up the devices beside them, in my opinion. I wish I could get my hands on the IRF640N as I really feel this is the ideal TO-220 device and certain it meets and exceeds your comments for smaller drive distortions and drive capacitance. My next best choice if the IRF540N.
As far as having a 4 pair version of the quasi PCB, I made and put it online over a year ago. It was the the TO-247 version of the board which is easy to adapt for TO-220 device use for the output stage. I have a more current version of the stock board with the corrections since uncovered by Shawn, but I may not be able to post due to diyAudio file size limits for attachments to forums.
I recall reading that on the plus side using a single pair of output devices one gains some extra power handling ability as one does not need the source resistors often used to force current sharing of a multi-device output staage.
On flip side there have those that have stated one still needs, perhaps at lower resistance, the current sharing source resistors even with a single output pair MOSFET amplifier. I have also read that there is better speaker damping using multiple output MOSFET pairs (not sure if same applies to BiPolars).
I do not know if any of the above points concerning single apir or multiple pair output devices is fact or fiction, nor do I have the knowledge or skill to evaluate any of these are fact or fiction. Perhaps others far more in know can help clarify.
I have had no time to even build a module. I had an extensive set of life demands on me for most of the last 6 months of this year and last year I had a dead (my end, not at exchange) phone line as a consequence to a Bell strike here locally that lasted about 6 months. The fallout of that affair had a number of long lasting side effects for several months over and above a major cat emergency that happened right in the middle of the 3 months my phone line was dead that caused serious problems in being able to effetively address Mahogany's medical emergency that lasted a number of days and then for few months in a less acute condition, but just on the edge of acute.
I am not sure even if I did build a single and multipair if I notice any sound quality difference. My motivation for a single pair module was driven by my desire to build a HiFi 3 way active speaker system and not needing the high power for mid and tweeter amps. This also made it attractive for various reasons to use soem TO-220 devices that in my opinion would be ideal for such low power single output device pair modules.
One other challenge is I do not have a suitable case and heatsink solution for any modules
Hope my comments are helpful Sek.
Regards,
John L. Males
Willowdale, Ontario
Canada
15 January 2007 19:59
15 January 2007 20:03 Typo corrections
15 January 2007 20:09 Typo corrections
Hi Black Chicken
.
I have uploaded the layout in sprint layout format. This will allow you to see all of the dimensions using the software. This layout also has a bit more detail than the PDF. You can download a trial version and a free viewer from here:
http://www.abacom-online.de/uk/html/sprint-layout.html
The 1R (ohm) resistor are 2 watt types (high temperature types) and their spacing is shown on the layout (use the measure tool).
Alternatively you can print direct from the PDF showing the layout and measure all dimensions with a ruler.
Whn choosing FETS its a trade off between a few criteria. I look for FETs with low Ciss, high power, availibility and price. Rds on and current handling are actually less important and will be taken care of automatically via the power handling criteria.
All builders please note; whilst I have posted the layout, I have not built and tested this version yet.
Cheers
Q
.
I have uploaded the layout in sprint layout format. This will allow you to see all of the dimensions using the software. This layout also has a bit more detail than the PDF. You can download a trial version and a free viewer from here:
http://www.abacom-online.de/uk/html/sprint-layout.html
The 1R (ohm) resistor are 2 watt types (high temperature types) and their spacing is shown on the layout (use the measure tool).
Alternatively you can print direct from the PDF showing the layout and measure all dimensions with a ruler.
Whn choosing FETS its a trade off between a few criteria. I look for FETs with low Ciss, high power, availibility and price. Rds on and current handling are actually less important and will be taken care of automatically via the power handling criteria.
All builders please note; whilst I have posted the layout, I have not built and tested this version yet.
Cheers
Q
Regarding the output coil....
or the lack of it.
Well there was no room on the board. The remaining output network will provide an AC load on the amp to aid in stability (although the amp is stable without it) and will also shunt RF to ground. It won't though present a higher impedance to RF. If an output coil is desired, one can be put in series with the output lead from the amp module to the speaker post.
Cheers
Q
or the lack of it.
Well there was no room on the board. The remaining output network will provide an AC load on the amp to aid in stability (although the amp is stable without it) and will also shunt RF to ground. It won't though present a higher impedance to RF. If an output coil is desired, one can be put in series with the output lead from the amp module to the speaker post.
Cheers
Q
marus said:
Hi Marus,
This module can be connected to different power supplieas and sound quite different. This smaller module connected to a quality powerful supply should sound great, but is limited to 50 volt rails, maybe 55 depending on the application.
I should deliver just over 100 watts into 8 ohms and maybe 180 into 4 depending on your power supply.
In terms of specifications, I have not yet built one so I cannot provide real specs. I do expect it have a very good signal to noise ratio and low distortion (below 0.05%). I have posted some specs somewhere in this thread for the other module, you could use these as a guide. JLM is normally very good at finding past posts....maybe if your lucky he might ......please John?
Regarding the schematics it is very similar to the latest schematic posted in this thread, with just a few component changes. I'll post it tomorrow maybe.
Cheers
Quasi
sek said:
This new small module could easily be changed to use 2 x TO247 (TO3P) opr TO264 devices and there is quite a good selection with a Pd of 250 to 300 watts (Tc=25 degC). These powerful devices have a much higher input gate capacitance but offer an interesting alternative.
If anyone is interested I'm happy to post a 2 FET version.
Cheers
Q
I will post pictures of my Quasi NMOS amp, perhaps, tomorrow. In the meantime, as I had already reported, the fully completed amp with two channels does distort when both channels are driven (beyond a minimal point of volume), whereas, when only one channel is driven, there is absolutely no hitch. I have not yet had time to sort this out.
However, I'd like to report that during the building phase, I ran a single channel (without any chassis) on a PSU of +-42 volts with a single pair of IRFP460s. They sounded very musical and detailed as compared to the fully finished amp which now has 3 pairs of IRFP450s. The latter has a lot more punch and control in the low end of the audio spectrum, but the musicality of the single OP pair is definitely missing. It might have also had to do with the higher capacitance of the 460s compaired to the 450s.
Just for the record, a single pair of IRFP450s were also tried but the weight that the 460s presented were missing; the background was darker with the 460s and timing seemed more accurate.
However, I'd like to report that during the building phase, I ran a single channel (without any chassis) on a PSU of +-42 volts with a single pair of IRFP460s. They sounded very musical and detailed as compared to the fully finished amp which now has 3 pairs of IRFP450s. The latter has a lot more punch and control in the low end of the audio spectrum, but the musicality of the single OP pair is definitely missing. It might have also had to do with the higher capacitance of the 460s compaired to the 450s.
Just for the record, a single pair of IRFP450s were also tried but the weight that the 460s presented were missing; the background was darker with the 460s and timing seemed more accurate.
Hello Quasi,
Yes, I would do that, but at the moment I dont have the ability to use a printer.
Thank you very much for the files!
As you only need 4 Devices, the price for the Mosfets is not very high. I.E. I pay 1.30 Euros for irf640n and 80 eurocents for the irf540n. The irf840 is in the same Range between those two.
So is it worth "inversting" the money for the irf640n?
For me it would be interesting to find a full isolated FET, as they are really easy to mount. Still TO-220 devices are easier to mount (isolated) as the bis ones.
Still the question remains: "is it desirerable".
As its no Problem to fit it on the Speaker-Terminals I tend to include one in my Build.
I surely would be very interested into that layout. But I like the to-220 Version alot. Because of the 2W Source-Resistors (I mounting those big ones) and the easy to mount to-220.
Maybe two to-220 have a better power handling ability than one to-247?
Greetings
Black
Yes, I would do that, but at the moment I dont have the ability to use a printer.
Thank you very much for the files!
As you only need 4 Devices, the price for the Mosfets is not very high. I.E. I pay 1.30 Euros for irf640n and 80 eurocents for the irf540n. The irf840 is in the same Range between those two.
So is it worth "inversting" the money for the irf640n?
For me it would be interesting to find a full isolated FET, as they are really easy to mount. Still TO-220 devices are easier to mount (isolated) as the bis ones.
Still the question remains: "is it desirerable".
As its no Problem to fit it on the Speaker-Terminals I tend to include one in my Build.
I surely would be very interested into that layout. But I like the to-220 Version alot. Because of the 2W Source-Resistors (I mounting those big ones) and the easy to mount to-220.
Maybe two to-220 have a better power handling ability than one to-247?
Greetings
Black