Peavey IPR class D amps - Page 17

Thank you. I should be able to switch some stuff around and not have to run another circuit. That is one of the reasons that I have been looking at the more efficient amps.

'Dude,

Catch me up. Do we know the FR yet, or does that info fall under the "you won't be able to tell the difference anyway, stupid" non-answer category?

Bosso

Thank you. Should not be a problem fitting that in with what I have now. I don't plan on using it for a 2 ohm load, but I did want to hear what you had to say.

Sorry but I am late to this thread.

I just read Mortyboy article overkill.talktalk.net/peavey/IPR1600.pdf
It was very interesting. However I do have a couple of questions for Mortyboy that I hope he can clarify for my better understanding:
1. When choosing a similar capacitor why was the 3.9mF, 80 volt capacitor picked? It is both physically smaller than the actual capacitor plus it has much lower Joules. The 4.7mF, 80 volt capacitor seems like the better pick. It is the same size and has similar Joules. The ripple current rating for this capacitor is 4.05A.
2. What is the impact of air flow on ripple current rating of the capacitor?
3. When calculating ripple current multiplier why is it rounded down to two?
4. Using 4.05A and an accurate ripple multiplier I get 8.4A. What is the impact on the life of the capacitor when only 60% of the rated watts are applied (this assumes the 6.5A figure is correct)?
5. The 8.4A ripple current rating is at 45C. Commonly used room ambient temperature is 25C. What is the impact on the life of the capacitor when it is operated at a lower temperature?

This is precisely why the kind of slander from Mortyidiot is so damaging.

Perhaps you missed this post:
http://www.avsforum.com/avs-vb/showp...&postcount=473

According to JD, the F3 is ~8Hz on the IPR1600, and ~16Hz on the other models.

I suppose so. I'm just trying to implement a resonant mode converter 'model' to dangle on my amplifier 'model'.


Muh-Huh


Yup, voltage below surge or at rated might not be a major bigee. What you are after is the heating effect of RMS ripple currents and operating temperature.... plus as you mention, environment, plus 'cooling'.


Indeed! I think I hinted at that when I asked you for more 'specific' data on the capacitors, found some but probably not enough, being used but then you came up with something like nothing and are hand waving once again. Tell me, if a spec like ESR goes out by ~20%, and I will assume that is a rise, does not the dissipation for the same ripple current go up by ~20% and as a result accelerate the failure?


That one is a little bit simplistic. Now you are talking about 'core' temperatures whilst ignoring core to case, case to sleeve and sleeve to ambient thermal resistances. You should agree, but I'm sure you would wish to 'bluster' that a 10C reduction in surface temperature really does equate to a 10C reduction in core temperature, although it does not. If you have the data then you would see that the curves are a bit bendy and don't really work in favour of your 10C doubles things scenario.

If Nippon Chemi-Con are not giving you the full banana then try visiting,

http://www.epcos.com

or

http://www.vishay.com/company/brands/bccomponents/

For some seriously meaningful data. Eeeeeee, when I wer't Lad..

Otherwise, assuming you have them, you might be able to publish your own.......{?}

Sorry folks, I realise this might bore people in AVS but I get the impression that Mr Bennett would rather I kept my tirades in one place. Hopefully you have an ignore button.


Indeed. I assume you knew that already and are not just flossing from the, lack of, information I pointed you at? However you are dealing with both the low frequency audio component and the high frequency switching component. If you wish to operate at limits then you need to check that sort of thing..


Indeed. I assume you knew that already and are not just flossing from the, lack of, information I pointed you at, although the multiplier was stated at 45C?

Oh. Kudos to AVSBama on the question about my choice of 'representative' capacitor. It would seem, waffle waffle waffle, that the devices in question are actually Q50 case size, 25.4mm X 50mm. My Bad. I would not wish to bullpooh about that but I might get the impression that the Volume/SurfaceArea ratio works in their favour in terms of dissipating internally generated heat.


Indeed! Ref the Epcos reference. It's a gentle curve to breakdown with a bit extra left to go. Mind you, it does look like that was the one you were bothered about whilst ignoring the other stuff. I'm sure everyone else does it as well.


Sounds GOOD to me!!


Oooooh that will be sarcasm and I'll say I only mentioned it once so you can find the other one and use it to totally destroy my reasoning.


Works for me...


Muh-Huh...



So this will be why the air deflector is positioned to cool something like the lower 80% of 20%, for a whole 16% effectiveness, of one side of one capacitor whilst the other one sees not a lot in terms of 'cooling' air. It will also explain why the, one, capacitor, being cooled receives its cooling air directly through one of the, exceptionally, close proximity output inductors.

Still, I assume that at your 'rated' output power levels those inductors don't get very warm at all. Not a problem then. Probably applies to the other capacitor receiving not a lot of airflow and its, not hot, inductor sitting next to it either.


AHAAAAAAH, I See you have entered Pro-Audio Waffle Mode.

'Our power supply recharges stuff absolutely HUGE amounts per second for most excellent things and stuff like pumpy base and dynamics with throb when you need it.'

Yay!! We are believers. Gosh 210,000 times a second.


Excuse me, sorry to get Techy, but resonant mode conversion slaps the capacitors at about 1.53 times the RMS of the load if you get it right and you are running up to 30 Amp 5uS half sine pulses based on rated power, I'm still looking to get proper figures for that one, into your secondary side main ponies and are expecting some 'poly' capacitors to save them from the onslaught?

Then you say


Works for me... not.

Do you wish to get into a conversation about 'de-coupling' capacitors and their impedance characteristics along with circuit trace and device inductance and resonances and...

That reminds me.

ArthurG. Next time try not to 'bad off' Eva


Nicely delayed. Obviously I can't 'prove' it yet but your rated figures take those capacitors down close to their 2000Hr Limit.


Well done. It seems you were truley focussed on that one. Design to the limits you understand or someone else has told you about, protect against them and miss the other important things.


The former will have little to no effect on the bus pump voltage and the signal frequency imposed is most exceptionally important. The latter is my guess at a value for your output inductor and its possible DCR. It would 'seem' that I was a whole factor of two out which isn't bad going for looking at published pictures.

[Regarding that one. Being someone who does not listen to monsters making lots of noise at low frequencies whilst they rip loads of guts out of their human, or other, victims and you are watching them doing it 'on screen' or even know what F3/FR might be I guess I might qualify as stupid and I am, almost, not really prepared to learn but if your sub is ported to suck power down low and your IPR cacks out then.....]


So it's a door stop... but not heavy enough to do that either?


That's all right. They sell millions of them and probably run HDBK-217 or the commercial equivalent on the designs whilst making sure that the F in MTBF is not a four letter word.

Boy am I glad this forum has an ignore feature.

I received the IPR 1600 today. I ordered it yesterday from Parts Express. This thing is so light that it feels like the box is empty. I have to make up a 1/4" to male banana plug before I can use it. I did not expect this amp until Monday or Tuesday.

Testing my new Peavy IPR 1600 and it gets a grade of zero. It is dead as a door stop. I plugged it in with nothing hooked up and no lights come on. I remember Ivan saying the lights are on even when off, but mine is dead with the switch on or off. It is not a problem with the power cord because I tried a power cord from one of my other amps. I have also verified that the dedicated outlet is hot. This sucks.

Added
I pressed on the circuit breaker and it does nothing so I assume that it is not tripped.

Thanks for the infomation. I looked on the Epcos page and there is some good info. However since these capacitors are from Nippon Chemi-con I thought it best to use their information. It is very similiar and bit more conservative:
They show the following formula to calculate life:
Lx=L x Ktemp x Kripple
It appears it is necessary to multiply the Base life by factors to account for temperature and Ripple.

KTemp = 2^((T0 - Tx)/10) = 2^((85-25)/10) = 2^6 = 64
T0 = Maximum rated temperature of capacitor = 85C
Tx = Ambient Temperature = 25C

KRipple = 2^((ΔT0 - ΔT)/5)
ΔT0 = 5C
ΔT = ΔT0 x (Ix/I0)^2
I0= 4.05(Rated ripple of similar item) x 1.2 (Freq Multiplier)
I0 = 4.86A (note this does not account for the benefit of air flow)
Ix = 6.4A
(Ix/ I0)^2= 1.73
Thus: ΔT = ΔT0 x 1.73 = 5 x 1.73 = 8.67
Thus: KRipple = 2^((ΔT0 - ΔT)/5) =2^((5-8.67)/5) =0.6

So with the above multipliers calculated we can now determine the life time calculation:
Lx = 2000 x 64 x 0.6 = 76,800 hours
Based on post by Sickneedhelp clubs that operate at maximum level for perhaps eight hours per day. So it looks like this amp should last 26.3 years.
Therefore after 76,800 hours or 26.3 years the capacitors have not failed but will have less than the specified parameter drift.
I am a bit confused where does the 4 months come from??

Hello Bama,

The potential problem with the output capacitors is twofold. Assuming the worst case conditions I mentioned in my last post, if the output capacitors are going to fail at all, most likely is that they could go into thermal runaway due to a temperature dependent wear-out rise in ESR. Ripple current is fixed by the loading profile (here we are assuming worst case) and internal heating goes by current squared times ESR. ESR rises due to operation at elevated core temperature and heating rises due to ESR rising. At some point this becomes self re-enforcing and runs away (until the capacitor boils). Worst case occurs with maximum loading (the 8 hour daily disco profile), maximum ambient temperature at minimum cooling (summertime in a closed equipment closet).

The second scenario is overvoltage runaway due to shunt leakage currents (which also go up with both temperature and operating voltage). If the power supply is unregulated, this would occur at high line and maximum power at the lowest audio frequency within the pass band (where power supply pumping is often the worst). Factors that could mitigate this are:

1) the power supply is regulated so the output caps don't see line variations; and

2) the power supply output "diodes" are really bi-directional active switches so that they can equalize voltage both between each of the two output capacitors and between the output capacitors and the power supply.

In fact, looking at the pictures on the talkbass.com forum, the power supply capacitors are several time larger (likely storing as much as 10x the energy) than the output capacitors. If the power supply is able to electronically put its capacitors effectively in parallel with the output caps then it is likely none of these problem scenarios would exist, even at worst case. Because this amp is cheap it is not likely the power supply has active diodes, but it is quite possible (and even likely) that the power supply provides well regulated dc to the output capacitors. This means the output capacitors would see virtually no 120Hz line ripple current and that the power supply would supply a lot of the audio current up to the likely bandwidth of the power supply, say somewhere in the low kHz range.

It certainly is not optimum that there is an airflow deflector blocking off one of the output capacitors. Also, the output toroidal inductors are very close to the capacitors, so that they may heat the caps somewhat by thermal conduction (such inductors often run quite hot). Another possible bad effect is induction heating. Outside of its winding, a toroid is equivalent to a single loop of wire the same diameter as the core. This single turn loop carries the full class-d switching current and may induce mirror-like high frequency eddy currents in the aluminum case of the capacitor. (I am not claiming this is a significant effect - just that it bears consideration).

Perhaps Mortyboy is being a "sour grapes" alarmist, but he is asking legitimate questions in my opinion. Mostly empty ridicule is not a very reassuring or convincing way to lay to rest the questions raised. As I stated before, I don't think there is any shame in a high bang-for-buck product having some limitations under extreme conditions - most users would be glad to save a few dollars for something that works well enough for them - just don't misrepresent the product (not claiming this is the case - just that the questions raised by Mortyboy were valid to ask).

Here are some things that I have been wondering about this "discussion".

Mortyboy-so exactly what do you do for a living-ie who do you work for? My appologies if I have missed this.

What commerical amplifiers have you designed-ie model numbers? Again my appologies if I have missed that one also. Sales are the only way to "prove" a design to the world. You may have the greatest thing ever designed-but if nobody is buying it-then how can your prove it (except to yourself)? What about a track record of some sort?

If you are competitor of Peavey-then why are you so worried about the life of the capacitors. Let's assume for a second that you are correct. I would think you should be happy knowing that in a little while your competitors product will start to fail and they will get a bad reputation because of that. And then you could use that problem to your advantage. Why help them improve their product? Seems like a bad idea to me.

Your argument sounds like it is between 2 of the engineers of the SAME company-trying to figure out what is best for the design and the company.

Let's assume you are just a DIY kind of guy. Fine. But can you produce your design-with the changes you recommend-and sell it for a good enough profit to stay in business-at the same price as the IPR1600?

Peavey knows very well what their customer base is-and what they are willing to pay. In a DIY type of thing you can do all sorts of things to improve the product-without much regard to price. But in manufacturing-every dollar counts-as it get multiplied by the time it gets to the price the actual customer has to pay for it.

An old saying goes "It is pretty easy to design and build a bridge to carry a particular weight (and the added expense to go along with it)-but it takes REAL engineering to design a bridge (for a particular price) that will just barely carry that weight".

Sales are what drives all commerical designs. Speculating is fine-but unless you can do better (and sell it) for the same price-then it is very hard to find fault with somebody elses commerical design.

So if you have a better design-that will last longer-and sell for the same price-then please let us know what the model number is-so people can order one and play with it and see how well it really performs.

BTW most of the amplifiers in use are not in heavy dance clubs (no matter what you think)-so most of the amps will not be subjected to that kind of abuse.

I will agree that heat is a major killer of amps in installed clubs. But not directly from the music and load-but rather from negligence on the clubs part. The fan intakes get all clogged up with bar dust-causing a reduction in air flow and a heat build up.

You would not believe how clogged up with dust some of the amps would be when they would come into my repair shop. Dust so built up the fans would not even turn.

As I stated earlier-my tests on the IPR1600 were kinda weak-but still gave me any idea of what the amp would do as compared to "normal" amp. It ran VERY VERY cool (under HARD clipping)-I was quite impressed. I did not drive it at 2 ohms-and any good design shouldn't either. Pushing an installed amp (or loudspeakers) to the limit is just asking form trouble. At least that is how I design my systems-and failures are a very rare occurance-and not from overdriving.

As I have mentioned before
When A mate of mine brought over A Demo 240Vac Peavey IPR-1600, we ran it pretty hard for A couple hours on the bench driving PR-12's, it did'nt even break A sweat, we were both real surprised how cool/efficient this thing runs and how clean it sounded for music playback...

IMO this amp offers Huge Bang for your buck, and weighs A mere 2 pounds..

If I was in the market Would I use the IPR-1600 to run huge power hungry Subs ? No
I would choose "wait" for its bigger brothers(6000) or something more suited to the task such as my Peavey CS4080 or better..

Cheers..

I think Sick mentioned earlier in the thread that the 1600 was originally intended for a minimum of 4 ohm loads, but that it turned out to have enough margin to be rated at 2 ohms as well. So at 4 (or at 3+) ohms it runs to the stops all day long without breaking a sweat, thus your experience makes perfect sense.

Many of the worst amp stresses go up with current squared so, at 2 ohms and nearly twice the current, it's not surprising that is where the sweat starts and the fan comes on.

As an SMPS and (former) class d designer, I always thought it would be fun to attempt to do a 10kW+ peak capable, self-oscillating (non clocked), multi-phase, full bandwith class d amp with an input power factor corrected power supply and enough energy storage to provide several seconds of 10kW+ audio without tripping a 15 amp circuit breaker (this would require 10 to 50 times the number of expensive capacitors contained in a normal power supply - maybe a battery would be better).

What a great demo such an amp would make integrated into the base (foot) of a pole stand for a three to four foot or so diameter sphere of a few dozen 8 or 10 inch speakers all driven in parallel. The sound fueled by one skinny little 15A line cord would be as amazing as it would be deafening. [end of engineering fantasy/digression]

3.2KW half bridge resonant converter designed... Sure, in my dreams like that's fairly mental. Write up and results some time later on today or Monday.

Laterz

AVSBama

Thanks for the info. I'll have to hunt down a link to the relevant information and try it myself.

I don't want to sound like I am trying to break your sums but I'd be wary about using 25C as ambient. 'Consumer' electronics are normally specified for 0C-40C operation so the base starting point would be 40C. Without doing 'real' temperature measurements you might 'guess' at 50C in the box and then things depend on the local internal environment.

IvanB

As mentioned before. I'm unemployed/unemployable. In a previous existence I did anything from electronics assembly through test to design. I have experience of SMPS design in consumer electronics up to MIL-Spec equipment. I think my fuse popped as a result of the continual battle against 'others doing silly things'. Designing the stuff is 'hard' enough without the added cr*p. I'm not going to point you at 'product'. Much of what I have done would be classed as sub-systems.

No... I have little to no experience of Pro-Audio just the interest in Class-D specifically and SMPS in general. Unfortunately, at the moment, I'm not going to be able to point and say 'buy one of these instead'. However having spun through some of the design process as a result of this I might be tempted.

No... I probably would not be able to do it for the 'Peavey' price but if I found myself on the receiving end of an idiot like myself I'd be able to post or otherwise refer to solid design and test data to back up any 'claims' I had made.


Hmmm.... ftp/webserver appears to be flakey Probably using a power supply designed by me

If and when it comes back I'll upload the words and the models then anyone who wishes to have a look, play with things or even rip the p*ss out of it can do so.

Otherwise, since I have demonstrated my ability to 'make friends and influence people' and you lot are probably more interested in other things, sorry to have disturbed, I'll keep my blather, and any update, in this post.

Yay... files finally went through...

http://www.overkill.talktalk.net/peavey
http://www.overkill.talktalk.net/peavey/IPR1600revA.pdf

Capacitor paranoia ?

This is a DIY forum, therefore modify the amplifier with a more robust
capacitor like this one.

http://www.ehow.com/video_4936272_el...capacitor.html

If it ever goes bad, you get french fries as a byproduct.

Okay, you linked, and against my better judgment I clicked and I watched. And I think something besides that guy's potato may have been fried.

No matter how you slice it (or in this case skewer it with nails), a potato is going to look like more of wobbly resistor than a capacitor. Er, I, myself don't have access to that exact model of potato to measure, but judging from the pictures posted on the net, estimating from relative size and the probable biasing conditions evident from the rest of the circuitry, I say it is most likely a 56k polyamylopectin resistor to be precise (give or take a few carbohydrates, of course).

Engineers humor.

[quote=analogspiceman;18347975]Hello Bama,

The potential problem with the output capacitors is twofold. Assuming the worst case conditions I mentioned in my last post, if the output capacitors are going to fail at all, most likely is that they could go into thermal runaway due to a temperature dependent wear-out rise in ESR. Ripple current is fixed by the loading profile (here we are assuming worst case) and internal heating goes by current squared times ESR. ESR rises due to operation at elevated core temperature and heating rises due to ESR rising.


Actually the delta T is very small. It calculated to less the 9C and this was with no air flow. With air flow it would be most likely less than 5C. Plus the above calculations indicated it would take 26.3 years for the ESR to double. So after 26.3 years the delta T is now 10C (5 x 2); no big deal!! Also since ESR decreases with temperature it really will not go into this run away condition.

Thanks. Think I'll notice any difference?



Bosso

lol bosso, you kill me. posting such a log graph that gives equal horizontal linear space to the 3-4hz range as the 30-40hz range implies that both are equally important. f_cking laughable.

Depending on the program material I doubt you will notice any difference.

And you are also assuming that everybody has loudspeakers that are flat that low in response.

You have to remember that the TARGET buyer for that amp will never know the difference. That amp was designed for the working musician. They don't have the sources or the loudspeakers that would make any difference.

I have not seen any claims that the amp was designed for home theatre usage with response down into the single digits-so why accuse it of something it was not designed to do.


That is like complaining that your sports car won't carry a bunch of sheets of plywood and bags of cement. That is not what it was designed for.

Different products are designed for different needs of the particular customer.

So what other amp do you know of that sells for the same price that has response down to 5Hz? Especially at the same power.

When I was measuring it-I was getting flat response to 11Hz (If I remember correctly)

"That is like complaining that your sports car won't carry a bunch of sheets of plywood and bags of cement. That is not what it was designed for."

i can haul 4x8 plywood in my c5. it's not pretty, but possible. ;-) i know what you mean, and agree with your point, though.

I hope that you are kidding. The C5 is made 30 miles from my house. I was in the museum last week.

I was telling fibs..

http://www.avsforum.com/avs-vb/showt...9#post18349569

Really, you mean it will actually be program related? Thanks for the heads up there, Ivan.


I'm not assuming anything, nor am I interested in what anyone else does.


Accuse it? Man, take a Mydol, will ya? I asked what the FR is and someone (apparently as holy as thou) remarked that it basically doesn't matter. I'm responding to that idiotic statement.

And, Take a quick look around you, Ivan. See any "working musicians" threads in this "DIY Speakers and Subs" Forum?



Sports car, plywood, cement, different products, needs... gotcha.


QSC/Behringer/American Audio/Marathon, all have 2nd order roll off at 5Hz. In fact, it would be easier to list what amps don't. This is the first I'm aware of that rolls off that high which would eliminate it as a possible power plant for lots of DIY subs.


You have the power test results already?

Bosso

Bosso,

What's the point of the question? Do you think people need to be warned not to use the IPR's in those cases?

Obviously it will limit your system and others w/similar capabilities.

Yes the freq response is only important if you have content that goes that low. The target market for the amp doesn't need it-nor does 99.99% of the rest of the world.

Just because people are talking about the amp here-does not mean that it was designed HT usage.