Peavey IPR class D amps - Page 13

Thanks, Arthur!

Neo,

which amps are you talking about ? Peavey and Behry stuff are made at different places. Peavey uses a sub-contractor when Behry owns a factory in Guangdong.

What does it mean when on another forum you see this comment:


found here:


http://forums.peavey.com/viewtopic.php?f=27&t=10992

I am looking at an IPR-3000 and wanted to grab one tomorrow, that comment makes me nervous.

Jeff

Looking at 'the' pictures, third post.... link might work?
Ooooops, add the necessary..... and the URL in the quote...

talkbass.com/forum/showthread.php?t=617982

Then given it is a 2U rack then the fan will be 80mm which will make the secondary 87V/4000uF capacitors 50mm high. The output chokes are kind of half this so, mixing units I'll call them an inch OD, outside diameter. The hole in the middle is about 0.5 inch. That gives the MLT, mean length of a turn, for the wire as an inch, 4 quarters.

Counting turns... I guess at about 60 so the length of the wire is... 60 inches or 1.5 meters, sorry about the units again. The fill on the inside circumference looks fairly tight. 2piR dah-de-dah says the wire diameter is about 0.66mm. Including insulation that is about AWG 22.5 with a DC resistance of 0.068948 ohms/meter at 60C.

That gives the DC resistance of the choke as 0.1 Ohms. That will become 0.2 Ohms when you bridge. DF, damping factor, is RS/ROUT so for a nominal 4 Ohm speaker that would be 40 single channel and 20 bridged.

That assumes the amplifier does use pre-filter feedback since under those circumstances the choke would be external to the loop and therefore its impedance would not be corrected for.

I've got a post in towards the end of that thread that discusses the way these amplifiers need to be connected to be bridged which does suggest that bridging for Bass might be the way to go. It also references someone elses post about one of the channels shutting down which might be down to the way the cooling cardboard works. However if the [poor] damping factor is going to give you wooly bass then maybe it is not.

I'm not sure that helps you much. I think for these it is very much a case of get one and see if it suits your application. If not, give it back.

I meant to follow up on this way earlier, but I just noticed that I hadn't....

If you're using more than one driver, then having multiple amps gives you redundancy. I would much rather lose some of my drivers rather than the entire system...

Did you have a chance to look up the numbers? If you're too busy, then don't worry about it...I'll probably end up buying one anyway just to see what you guys are up to.

I've just revoked your B and downgraded you to a C-. Some of your assumptions may scare people off. That will require detention.

I like that better. We need the $$

Hi Jeff

The 3000 is not available yet. Try a 1600 on your subs. They certainly work well here driving four dual 18" bass reflex cabs. The only question is if it has enough power for you.

Not entirely my intention but C- sounds like it is going in the right direction. I shall now go get some ugly sleep before a bit of analysis.

Here you go...

This is a quick mock up of something to get an answer. Basically on the left an input source with feedback pre-filter drives some switches, mosfets, through a buffer. Middle is LC filter and 2R load. Right two 4000uF caps pre-charged to 80V coupled through rectifiers to two 80V voltage sources. The waveforms are for 75V out at 20Hz.



You can see the power supply pumping on VDD and VSS bringing the rails up to 102V. That is below your 110V surge rating but above the 87V continuous rating of the capacitors.

I wouldn't be able to comment about reliability under those circumstances but then it is a consumer product. I suppose you will have some form of overvoltage protection in place to limit hings to what you would consider to be an appropriate level.

Of course it isn't 'accurate' in the sense that the 80V supplies are stiff and all other devices are 'ideal'. This is obviously single channel so for 'normal' stereo music use you are going to get some cancellation due to the internal signal inversion and achieve the same if both channels are driven with the same signal.

If however someone decides to use one channel for Bass or Low Synth with the other channel for Higher Frequencies or some other combination then they might run into problems. I have no idea what the state of 'club' music is these days, old fogey. I assume the amplifier it is not going to kill itself as such but might just cut out every so often.

Obviously it is engineered to do a job for 99% of the people 99% of the time, percentages might vary. You appear to be dealing with 'enthusiasts' in these and other forums so they might be expected to push the limits.

Oh, as to my assumptions. Someone was concerned about damping factor.

Apparently it was one of your representatives who suggested bridging was not recommended for those particular reasons.

I took a 'guess' as to what the DCR of the chokes would be for the 1600 to get an estimate on the possible DF single ended and bridged. I don't think I was too far out and presumably it is stated in the data sheet somewhere..

Have I got down to 41Hz yet.

The voltages you show are not possible to obtain with the 1600. I'm not sure what the intent is, especially since you are doing the same thing in multiple forums. You've been very polite, which is certainly appreciated, but is the goal to:

1) Analyze the details of the design so you can build your own?
2) Scare off potential buyers?
3) Have us take note of your switch mode chops?
4) Beat Arthur in a contest of I can do a better analysis?

This is supposed to be a helpful Q&A & discussion forum, not a resume posting forum. Waddup?

Sickneedhelp, I am going to hold out for the 3000 that is due at the end of the month.

I am trying to keep the ratio of 4x sub power versus front channel power, so it fits the needs closer.

I appreciate the feedback.

GTG!

Jeff

MortyBoy,
I believe between you and ArthurG, the speculation has approached ad nauseum.

Two suggestions for you both:
1. Buy one or get one on loan and see if your speculative talents approach anything short of amazing.

2. Start your own thread entitled "Armchair Engineering" so you both can impress each other with how much you know about products you haven't gotten to market. You can send private messages to each other with high fives and such.

My engineers are speculating at the moment also. But they will bear fruit in no time at all.

I do apologise. Switch Mode Power Supplies and Class D amplifiers are a subject that interest me. When this amplifier was mentioned here and elsewhere the first reaction of most was to point and larf at the claimed power ratings. I don't really have any problems with that, the claimed ratings, it is obviously designed to do what you claim.

Elsewhere others have asked questions about bass use, using the separate channels for different sources and damping factor. I have got another post in moderation here at the moment. I think I'm being checked so I don't post rude pictures. I've tried to answer those questions on the basis of what I understand and can see about the product.

I don't mean to frighten people off I'm just suggesting where there might be issues that they may need to consider. I don't think I've managed to do a 'sky is falling down'. Just suggested that for some applications mileage may vary. You, as the designer, will know better.

It may be the case that you will get customers experiencing some 'issues'. If they know about the possibility in advance and how to approach and solve the problem then that's a happy customer, or at least one who might not throw their hands up in despair and slate the product.

I suppose it might look like I am showing off but I have spare time and some tools and know a bit. If I am wrong then I am wrong. If I am partially right then there are caveats that people may want to be aware of. Since Peavey is letting people 'try' before committing and the amplifier is going to keep most people happy then that should be cool.

Uhm

1) No. I could probably roll my own, although it would be a pipe dream.
2) As suggested not intentionally. If I am or do go overboard then slap me.
3) Like I say, I know some things.
4) I didn't realise there was a competition.

Oh, it's not a resume. I'm unemployable.

You're up to a B+ due to points gained for having a great attitude. If I graded on the curve, you would have an A, but I don't. Please accept the B+ as pretty good considering you have no way of factoring in the custom semi's. We should have left the Peavey part numbers on the main bits visable. That would have avoided a lot of this. Our bad.

Oh, go on.. I'm trying to get my grade average down here. The HRT must be working

Dremmeling off the Peavey part numbers is an impressively cunning triple bluff. That is really going to mess with El Presidente's head.

Ref post on...

http://www.talkbass.com/forum/showth...50#post8839150

Without knowing the circumstances maybe you need a new piece of cardboard.

Oh look. Stage left beckons.

Good luck with the product. Respect and high double fives with hand jives.

If you are referring to Chunkylover73's overheating, he says the fan wasn't working. If true, no amount of cardboard redesign will help.

You just lost your B+. You are at B- now.

I thought I exited stage left. An invite for an encore.......?


Oooooops HRT wearing off.

Why deliver product with fans that don't work? Quality control gone wrong?

Hows about we take a serious look at surge voltage rating of capacitors and then move on to ripple current ratings and lifetime multipliers/dividers and then move on to transient thermal impedances of semiconductor devices?

Let's go FMEACA and determine whether your product warranty matches up to the build or the figure has been calculated including expected returns. Not that the rest of them don't do it. You pay your money you take your chances?

Why are you not using FR4 in the product, I forget the name of the compressed 'thin' resin paper stuff? What's with the one ounce copper? That's a lot of weight flopping things about the place with little to no supporting points, which don't really do much anyway, cracking tracks and joints.

Looks like an unreliable piece of dirt to me. Care to argue against that opinion?

Ouch....! Like it was only a suggestion.

Oh Dear. There you go. I have just made an effort to warn people off buying this product.

Do you actually work for Peavey or are you out to rubbish the product yourself and giving invitations to help?

I stopped the speculation long time ago as JD doesn't give out any useful information but I support Mortyboy constructive messages. And to reply to your points:

1- I want to buy one but it's not available in Spain nor in France... More than one year after announcement, it's been long

2- when I engineer something, I actually like to do it on a confortable armchair

PS: Are you still hiring the same German engineer to copy IPR ? You know the one who worked for you on RMX and who is owner of a company name starting with A

Am I the only one to think your attitude becomes annoying with these notes and the way you never reply to answers ?

What will you think if I give you a C- to have not included a 2 dollars heat sink on top of these Mosfets ? Even with a fan problem and given the efficiency of this class D design, if a heat sink was here, the amp will run fine into 4ohm with signal music...

But I already know what you will reply: the amp works perfectly without heat sink and heat sink adds cost for an abnormal operating mode. Yeah, but the issue is that a reliable design is the one that takes care of the abnormal conditions. Otherwise, no amp will have short circuit protection...

lol Morty.

But the more important question is whether or not you can do better given the same constraints?

If it's selling for $300, then the actual cost to Peavey is probably going to be in the $50-$75 range. Maybe a bit higher if they're expecting to sell larger quantities.

Implementing everything you could possibly think of would dramatically raise the price of the product to the end user....that's why brands like Crest exist - where people are willing to spend extra on other areas of performance.

hey art, i just realized that i didn't thank you for some of your posts several back. so...thank you.

all, without the more traditional heat sinks, do you think this amp will have good durability, or may it have some burn up issues over time? or, is the thermal protection sufficiently good to keep it in the durability region for many, many years?

MBentz,

you are right... to a certain extend. In all designs, compromises must be done, no need to argue, but problems appear when you go too far and cut too much corners...
For example, relying on a AC Mains breaker to protect a mosfet or not using a heat sink on output are things not done in a correct way, whatever JD will say.

Until I will bench it, I will not have a definitive opinion on this IPR thing but so far, I'm dubious. Hopefully for IPR users, I will be wrong and I will be happy to finally see a cost effective class D design in the market

Sigh ....... You missed so much, I had to drop you down to a C. The other items you missed (including a major fundamental in your simulation above) got you a D. Attitude dropped you to an F-. You are now disqualified to render objective design opinions.

Yes, I know, we have to repeat this scenario in the other forums you joined recently to comment on the IPR and to promote your agenda.

Noooooo, our cost is much much higher than that. We actually make very little on this product. This is one of those, you had better sell a lot of them models.

Arthur.... Remember Ivan's report on the thermal performance? If you just have to have heat sinks, you pay extra and buy the Crest E-Lite 1800. I gave it heat sinks because we are charging more for it and I have to do things to justify the extra $$.

Care to share with us some of the commercially successful designs you have out there? It's only fair that you give me something to pick apart and comment on based on a lot of assumptions.

Sick,

How does your design handle over current? Linear limiting? (prolly not) Is OC limit really thermal plus hiccup mode only on gross over current? Red light comes on to spank user and put user in time out corner until power is manually cycled?

For non-clocked designs, aren't both phase shift and every form of hysteresis besides plain vanilla patented? For a given idle frequency, it is very difficult to get as much BW out of clocked designs as with the self-oscillating kind (perhaps some very fancy non-linear DSP control could even things up, but I've never seen it reported).

I think you mentioned that the rail voltages are rather low (<<100V), so perhaps you just switch in the near MHz range so sampling effect limits on BW is not an issue. I wonder how much carrier appears on the outputs. Perhaps it is nulled at idle and maxes out at some fraction of full load. Do tell.

Btw, have you been at Peavey long enough to remember Sondermyer (sp?) or Attwood?

By the way, in the Solar power / inverter biz, we use sm MOSFETs soldered over a sea of vias (>60 per FET) through a four layer 4-2-2-4 FR4 PCB. Sil pad goes underneath to provide isolation to the heatsink. Thermal performance is surprisingly good. This is bought at the price of not so nice injection of common mode EMI currents. -- a.s.

"How does your design handle over current? " Very well thank you.

The red light time out only activates in the event of a DC fault on the output. Then the supply recycles. There is no spanking involved.

Rail voltages are just right for the rated power.

There is some carrier leakage on the output.

I've known Sondermeyer since he joined Peavey during the very late 60's. I've been here about 14 years, so yes I know and very much respect Attwood.