Power amp under development

Hi,
is that single polarity 32Vdc or dual polarity +-32Vdc?

A FET amp needs a large voltage overhead to drive the output gate. Expect a peak output voltage about 6 to 8 volts below the regulated supply voltage.

This would indicate 144W into 2r from dual polarity supplies, at best.

I doubt the amp could properly drive a 1ohm (reactive load) to allow bridging.

From single supply rails the output powers will plummet.
 
AndrewT said:
Hi,
is that single polarity 32Vdc or dual polarity +-32Vdc?

A FET amp needs a large voltage overhead to drive the output gate. Expect a peak output voltage about 6 to 8 volts below the regulated supply voltage.

This would indicate 144W into 2r from dual polarity supplies, at best.

I doubt the amp could properly drive a 1ohm (reactive load) to allow bridging.

From single supply rails the output powers will plummet.

Sorry, it's a +-32 of course. The 2 ohm I was thinking about is a normal 4 ohm speaker driven by 2 of nmos200 amps bridged. Taken from the calculation earlier in this thread (32*0.707)*(32.0707)/2 gives about 250W. I'm absolutely not sure that I can build an smps thats capable delivering 500W, so the amp would rather be supply power restricted. Anyways, I'll give it a try. :)
 
not sure that I can build an smps thats capable delivering 500W, so the amp would rather be supply power restricted
then forget it.

(32*0.707)*(32.0707)/2 gives about 250W
you have omitted the voltage overhead that FETs need to drive the gate. In addition there are the usual voltage drops that apply to both BJT and FET amplifiers.
You cannot get anywhere near 30Vpk at the output from a regulated 32Vdc supply.
As suggested earlier, 24Vpk is probably a reasonable estimate to allow decisions on whether to go ahead.
24Vpk gives 144W into 2r. that when bridged will be about 288W into 4r.
Now check whether the Nmos can drive reactive loads down at that level from 32V supplies.
A 1r0 resistor is an easy load compared to 2ohm 45degree phase angle. 2ohm 60degree phase angle is a much more severe load.
What I am saying here, is that a 4ohm load being driven by a bridged amplifier is worse than asking one amplifier to drive a 1r0 load.
You are asking a lot from both the amp and the PSU.
This project is liable to take a lot of development time and a lot of discarded/blown components. Just think time and cost and balance this against your increased knowledge by the end.
 
I reckon the IRF540n's might just do it, although as AndrewT suggested it would be better if you ran the amp single ended into a nominal 2 ohm load. This could conceivably go down to about 1.3 ohms but even that is actually below the 10mS SOAR for 2 IRF540n's per rail. So provided it's not about competition sound-offs things should be ok.

You will only get about 48 volts peak to peak (16.9 volts RMS) so in terms of "RMS" power we're talking about 142 watts. Remember though; provided your power supply (and car battery) can handle it, these are the real McCoy watts (I.e; more than marketing watts).

I do not know however how the amp module will react to a collapsing switchmode supply, and I fear the worst.

Good Luck
Q

PS. Use a fan.
 
Re[02]: NMOS

zlast said:
IN the schematics for the NMOS-200 and NMOS-350, capacitor C7 or C8 (100 uf 63 volt) is shown as NON polar. But in the board layout it is shown as polarized...which is it? Is there any recommendation on the 10pf, 39pf, 330 pf caps...are ceramic ok?

For C7 and C8 of the respective schematics this is ground leg capacitor of the NFP.

C7 (NMOS350/500) and C8 (NMOS200) are the ground return NFB caps. Most builders, including quasi have used Electrolytic caps. Hence why the PCB has a polarity mark on PCB.

Ideally non-polar caps are the desired type, but at the 100uF value such caps are very hard to find, really large, too many to parallel to value desired and really expensive. There are mixed opinions if BiPolar or Electrolytics are the next best choice. The voltage usually seen by this cap is basically the offset voltage at best in normal circumstances. As a consequence of the low voltage seen by the cap it is generally known that Electrolytics behave like BiPolars.

In both designs the cap is in series with a 1k0 resistor that will have a voltage drop lower than the DC offset measured. Overall experience with DC offset for the builders so far these Quasi amps has been usually 3-5mVDC, generally no more than 10mVDC. This means ignoring the voltage drop across the 1k0 resistor in series with the cap the Electrolytic will behave like a BiPolar.

It has been commented a few times that if the DC offset voltage is high or rises for some reason there is some other problem with the amp (bad part, misplaced orientation of part like transistor, power supply polarity, el al) that will likely blow the cap as well as other parts resulting in module failure.

You can make a BiPolar using two Electrolytics in series back to back. To make a BiPolar 100uF you would used two 200uFs Electrolytics in series. This would of course require more PCB space to implement.

The down side suggested by some about BiPolars are they are basically two Electrolytics in series and have a higher inductance. BiPolars are said to be larger as well.

Some will use an Electrolytic and parallel the Electrolytic with a high quality film capacator, often a 1uF. Quasi used an electrolytic, hence the reason the PCB has a polarity marking as most builders use an Electrolytic cap.

Concering the 10pf, 39pf, 330pf, you can use ceramics. If you can use silver mica, film (MKP, MKT) and the cost of are not that more expensive than a ceramic then do so. I think Quasi used ceramics and feels there will not be that much different.

As it is DIY project you have to decide what suits your budget, audio quality and part availability. For sure the input capacator one is best to use a film cap, MKP or MKT rather than an electrolytic. I doubt you will find a ceramic in the 1uF value.

If you open the printable version of this thread, then display all posts in one page and use your browser search you will be able to find some of prior discussions and opinions on the questions you asked. If you do an internet, or even a diyAudio, search you will find your question is often asked. Some have definite opinions, others will argue there is no difference in cap impact to sound, others will choose to use film caps in certain parts of the circuit. If you take the time to read you can then make a decision for yourself if the basics I noted above are not sufficient to enable you to decide.

In my opinion I am using silver mica for most small dimensioned cap for thermal stability. There is not really much difference in price to ceramics and easily available, hence my decision. One pf capacitor I will be using a MKP, due to availability locally and price is basically inexpensive. I am using a MKP for the input capacitor.


Regards,

John L. Males
Willowdale, Ontario
Canada
24 March 2007 00:01
 
Re: NMOS

zlast said:
IN the schematics for the NMOS-200 and NMOS-350, capacitor C7 or C8 (100 uf 63 volt) is shown as NON polar. But in the board layout it is shown as polarized...which is it? Is there any recommendation on the 10pf, 39pf, 330 pf caps...are ceramic ok?

In theory the voltage the DC voltage across this capacitor is zero, so a polarised one can be used. Others prefer to use a non-polarised one so it's up to you. I have always used polarised capacitors here without any problems.

The 10pf and 39pf capacitors can be ceramic, the 330pf should be a poly type (although a ceramice will be fine). Some DIY'ers claim to hear a difference when using different types of low value capacitors and I do not dispute this at all, so it's up to you. In the end the amp won't care, it is very stable and tolerant of component choices.

Cheers
Q
 
Listening to Quasi

When I first built Quasi I listened to it with a set of my own DIY Bob Brines Peerlees Pipes and I was using my own DIY Sheldon Stokes valve SDS Line Stage.

Later on, for a few weeks, I direct connected my CD player to Quasi through a pair of Alps POTs and the B.Brines TL/PeerlessPipe speakers.

Further on I hooked up Russ Whites Kookaburra preAmp linestage still with the Peerless Pipes for speakers.

My all time favourite Quasi set up was with the SDS Linstage using NOS GE 6CG7's or(second choice: modern electro-harmonix 12AU7/ECC82's) and of course the Peerless Pipes for speakers.

I mixed in a set of my own DIY bookshelf speakers that I designed and they were terrififc but I messed up the port calculation and I ended up rolling off the bass response much too soon but still I could feel and hear the mid and high-end true as a bell. Still I liked the SDS linestage in there most! I will revisit the little bookshelf speakers in the next year or so....I hope. I have been using two different Vifa Tweeters and they seem to be a huge factor for me but I keep messing up the part numbers. I'll come back ASAP with my fav Vifa Tweeter as soon as I make certain which one is my fave. I just know it is the one with the soft black dome and not the chrome dome.

I have three CD players and my first choice is the Rotel RCD-855, second Rotel RCD-865BX and third is the Teac C-1D. The 865-BX has since been serviced but It seems to still have problems.

My Quasi amp is currently playing through some Martin Logan electrostatic's with sub woofer built in and I think they are Sequence II's but it is over at another chaps home and I didn't get all the model numbers as we were having a few glasses of wine when I went over with the amp. The experience was good and I have since been on a fact finding mission about electrostatic speakers. I think I'm going to try to DIY a set!

Work is off the rails so my time has been research and data sheets so the KSA-50 and such are on the back burner just waiting. Shame! I can't wait to compare more gear. I have a lot going on and alot of nothing at the same time. I will be forcing a work break for "home-time" when I finish up the next two or three work contracts and then I will have the cash to indulge a little deeper. DIY can be expensive!:D

I hope this helps potential Quasi builders to try this clean and accurate sounding amp.

Cheers,

Shawn.

Edit: typos
 
Re[02]: Listening to Quasi

TomWaits said:
My Quasi amp is currently playing through some Martin Logan electrostatic's with sub woofer built in and I think they are Sequence II's but it is over at another chaps home and I didn't get all the model numbers as we were having a few glasses of wine when I went over with the amp. The experience was good and I have since been on a fact finding mission about electrostatic speakers. I think I'm going to try to DIY a set!

I hope this helps potential Quasi builders to try this clean and accurate sounding amp.

Cheers,

Shawn.

Edit: typos

Shawn,

A few quick questions if I may:

1) Is the amp at friend's place with Martin Logan's the first quasi amp modules you built with the 5 pairs of MOSFETs and the 44-0-44 927 VA toroid?

2) Are the Martin Logan's a very reactive load impediance?

3) When you say the "experience was good" with the Martin Logan's is that you are surprized at the quasi amp modules you built to handle the Martin Load and/or your your own opinion about Martin Logan's? Some people do not like Electrostatic. I am not suggesting you are one of them, but there may be other elements not related to the amp or ML's that results on your "experience was good" rating. It is also possible you may be comparing the ML driven by the Quasi vs some High End Tube amps. I do not know, but why I am asking. Model numbers or such is nto needed. Just a sense of what you feel is important.


Regards,

John L. Males
Willowdale, Ontario
Canada
24 March 2007 04:32

P.S. Hope I have no typos to correct ;)

24 March 2007 04:35 Opps bad cut of the bold token and system still on standard time corrections. jlm
24 March 2007 04:37 Opps few more typos corrected. jlm Too late and tired
 
Re: Re[02]: Listening to Quasi

keypunch said:
1) Is the amp at friend's place with Martin Logan's the first quasi amp modules you built with the 5 pairs of MOSFETs and the 44-0-44 927 VA toroid?

Yes.

2) Are the Martin Logan's a very reactive load impediance??
I'm not certain but..yes they are because they are driven by a transformer that is driven by the amp. Kinda like a tube job in reverse I guess.;)

3) When you say the "experience was good" with the Martin Logan's is that you ...

The guy has a Sonic Frontiers tube preamp! His CD player is some thing I can't even pronounce. There was a nice turntable with some boutique pre-amp box attached. He has really nice gear and spent some good cash on it. I don't really view audio the same way as him so it is a little awkward for me. I think he should jump on the forum and post his own impressions but he doesn't seem inclined.

I don't really care what gear he is using but the ML Electrostatic's made a huge impression on me. I remember having ripped a few broken electrostatic's apart when I was a kid and I thought they were very interesting to say the least. I talked to a guy this past week about purchasing a second hand set of ML's but what I read on the net says that they are not as good as the price tag but I could be wrong. I did Like the sound very much but perhaps I am infatuated with the concept of ESL even more so. I was kinda blown away but I need to listen more without the bottle of wine on the table.

For the record, my Quasi has no problem driving the ML electrostatics and there is no reason that it should or most other amps as far as that goes. I have read that electrostatic's are power hungry and demanding but I didn't get that impression when i was listening. Perhaps it is true but not as bad as folks tend to incline or the quasi amp was designed to do the job?

Cheers,

Shawn.
 
Working On It...

Quasi:

Ok so i did the pcb and i'm currently working on the Nmos350, and so far so good. I'm going to get the transistors on these next days but i can't get the 2sc1845. I searched for the replacements you told me about and the only ones i can get are the 2n5551 and 2n5550, i can't get any of the other. The problem is, 2sc1845 is a low noise transistor and the 2n are not, is this a problem? Are there any other replacements? It kind of worries me:(
 
i to all, i'm back after 2 weeks of being sick.

Ok, got a question about the nmos200

I make the pcb and i notice that the 2w resistors (carbon) don't fit, so i check the thread and found that the pcb is the right size (acording quasi page 136 of the thread), about 75mm 95mm. So why my 2w resistor don't fit??.
I search the thread and found that black chicken make the same question but i don't understand the answer (someone who wants to explaine to me?).
My problem is only with the 2w (carbon) resistor im sure the resistors i have are 2w and i will use them anyway if you say me that theres no other problem than the size , for the rest of the resistors i will use 1/4 metal film that acording to quasi who answer to a korean guy, i can use, and quasi's word is holy word.

And for the last question. can't get any 100uF 50v so i guess i can use a 100uF 63 instead right?

Just to know before answer me (if you want to ;) ) the only resistors i can get are metal film 1/4 only, and standard cabron resistors.

Thanks to all this thread is great
sorry for my poor english