D Class amps - anyone got their minds around these yet?

[Mars Amp Repair]

hello folks,
We've been repairing the D Class Mackie powered monitors as of late. It seems to us that these manufacturers haven't quite got this D-Class down yet.
We've got filters that blow up & just kind of typical switcher type issues, but don't ususally see these kind of problems until the amp is like 5-10yrs old.

It appears to us that perhaps these manufacturers haven't got down how to make this D Class work with a source that has on & off demands ie:voice, band instruments, etc.

Whatever it is Mackie has one model HD1531 I think that they don't even have a retro-fix for the issues or even a redesigned amp module to send us to put in. They've been giving a full refund to the customer & having them ship the speakers back. That's gotta cost them.

We see blown filter caps that blow up again when replaced (we received a Mackie in which a tech attempted this). We haven't been attempting repairs like replacing the blown caps and such as long as they're under warranty & Mackie will send entire boards. The hope is the new boards will have retro-fit redesign at some point.

We also saw some MarkBass amps with the D-Class with similar issues. The guy at MarkBass really didn't have anything to offer and suggested the customer ship it back east to the only authorized servicer...somewhere in Virginia I think.

Personally, I hope this D-Class trend never makes it, but from past experience of some 30yrs I never say never anymore. With enough money behind it and similar profits to be had, they'll make them work somehow...g

Anyone else having similar issues?

[Steve Conner]

I bought some commercial Class-D amp modules a few years ago, in an effort to develop a portable bass rig. I exploded the SMPS twice and finally the amp itself, putting 400 watts worth of DC on the speaker, at which point I gave up and went back to Class-AB.

5 to 10 years is a reasonable service life for a modern piece of switchmode electronics designed down to a price. That's about the time it'll take the electrolytics to dry out, as they're carrying several amps of RF current inside a hot chassis. If you don't replace them with the correct low-ESR parts, they'll fail again a lot sooner.

The designer isn't under any obligation to make the kit last longer than the warranty period.

[Chuck H]

5 to 10 years is a reasonable service life for a modern piece of switchmode electronics designed down to a price...

The designer isn't under any obligation to make the kit last longer than the warranty period.

Spoken like a true engineer

Of course you personally are using AB.

I do know a couple of people that have turned off to the Mackie products because of this failure. And I think that is the spirit of the OP. A reputable MFG doesn't get that way or grow their client list with products that can't be serviced and/or fail just outside the warranty period.

BUT, it's up to the buying public to refine the nature of available products. And we know how that ends up. So there by the grace go the engineers with the daunting task of trying to make chicken soup out of chicken sh!t while trying to hold onto some measure of personal credibility. Not easy.

[km6xz]

Except in low production volume pro audio there really is not as much call for AB or A amps, customers like the dirt cheap prices and low weight of newer tech. It works longer than they usually keep a unit before replacing it with one with a few more features. They are surrounded by such power topographies in everything from car system, cell phones, TVs, etc. Cheap is what people vote with their wallets for.
The gear is repairable but takes a different skill set and methodology than 30 year old consumer products so shops doing them have created a service system around it. Most tech are unwilling to work on a lot of mainstream gear now, that only opens an opportunity for others to fill. Every town has a old semi-retired TV tech or tube guitar amp tinker but the scale of work to be done is dropping while the slice represented by very old technology compared to the whole consumer electronics market is shrinking. Can you make money by depot maintenance on D-amps or switch mode power supplies? Yes, if you design your service around them, high volume at low per unit charges. But it is not something that a hobbyist wants to do and anyone still in the amp repair business really has to be a hobbyist to keep at a field that a lowering of return on investment each year, and a shrinking market.
I am setting up a shop here in Russia mainly because of two reasons, I want to get my hands dirty and play, and because there is a great growth potential. Lots of gear coming out of warranty and no place repair it or get parts. Since discussing this with a few people in the retail side, the word got out and techs or wannabe techs from other parts of the country(a vast area 11 time zones across) who want to join forces if I can get parts for them. Actually that was the plan to start with but my own service centers and extended warranty program. There are lots of new opportunities everywhere to make a living but a lot of them are not exciting or hobby oriented like tube guitar amps. Personally I like the new tech, just as much as the old stuff, and actually prefer designing, building and repairing surface mount over point to point. That is where the repair market it due to its dominance in volume of new purchases, and the lack of competition. Everyone who ever plugged in a tube thinks there are a master amp tech, as long as it is 50-30 year old technology. There just are not that many people willing to pay for work on gear that appears so simple, simple enough to have them try it themselves. That leaves 80% of the gear outside their specialty, and with little competition for those who do want to do it. The customers in this depot style work is not the end user but in contacts with brands, distributors.

[g1]

Sounds to me like we are speaking more of the shortcomings of SMPS than of Class D amps in particular. I think the big variances in loading of the supply and also the environment (club AC voltage etc.) are big factors here.
I recall early attempts at SMPS for amps such as GK 2100 & MB series, and Carver 2.0 having lots of failures.
When the GK2100 first came out, there were many failures and the factory came out with pages of mods. for them. Then they released the MB series, again with SMPS and eventually (after lots of failures) they came out with a retrofit PT and linear supply. The newer MB's were no longer equipped with SMPS.
The Carver 2.0 had lots of supply issues and I remember hearing of lots of Clair brother mods. for them so they could be reliable in pro applications.
Since these early attempts at SMPS usage there may be many examples of ones that did work reliably, but I am not aware of them and don't think power amps are a good fit for switch mode power supplies.

[Steve Conner]

It works longer than they usually keep a unit before replacing it with one with a few more features.
...
anyone still in the amp repair business really has to be a hobbyist to keep at a field that a lowering of return on investment each year, and a shrinking market.

These are all perfectly good points. All reasons why I don't build or repair guitar amps for a living.

My day job is designing microprocessor- and DSP-based industrial and scientific instruments. This is not a very cost-sensitive market: custom instruments can sell for tens of kilobucks each, because industrial customers can have test and measurement problems worth millions.

At that price point, you just use high-quality parts and a conservative, worst-case design: there is no point in nickel-and-diming the customer, with whom you're trying to build a mutually profitable long-term relationship, over a couple of capacitors.

If I built amps the way I built instruments (and I pretty much do) then I would never sell a single one. My relationship with tube amps and old analog stuff is similar to those crazy beardy guys who do Civil War reenactments at the weekends.

[km6xz]

The key to any business is to price based on the value of the problem as perceived by the customer. If someone builds the most bullet proof unit in existence and it solves problems the customers do not value, the unit is destined to be a failure.
In some small niche markets a $5000 one-off guitar amp can sell, to a small handful of people because very few people value their problem at $5000. It is all priorities. If someone has it as a priority, above all others, they can find a way to get it regardless of income. Few people would balk at spending over $5000 for a car because they value the mobility problem greater than, say, a $5000 TV or guitar amp. They ARE spending a lot on some problems to be solved. So finding a problem to solve for customers who value highly their problem to go away, and you can make a good living.
The specialized test instruments is one of those areas with highly motivated, deep pockets problem owners. Guitar amps and repairs is not one of those areas for the vast majority of guitar players.
When the first practical affordable digital tape decks came out(ADAT, at $3,500 each), they were so cheap compared to the alternatives that every musician upset because labels ignored them put 3 of them on a credit card. Alesis could not build them fast enough. That was the price point at which a large portion of home recordists could, at the time, see that their problem they valued was worth more than $3,500 ea to go away. Now, 10 times the performance would be needed for someone to value their problem worth $3,500-$11,000 if it disappeared. For me it was a room full of Studer 2 in decks and 1/2 mil consoles in 3 rooms back when the problem was big enough and valuable enough to invest that much. Now, labels do not value production much, so they only pay 1/8 of what they did in the 70s and 80s.

[gbono]

Just got home from the 2011 NAMM show. I noticed that Loud (who owns Mackie) also have a class D amp line for their Ampeg bass amp line. Any known problems with these models PF350/500?

I'm also assuming the capacitor problem on the Mackie line isn't due to counterfeit parts?

[km6xz]

The problem with early adopters of new technology is that the designers do not have as much experience yet and have not encountered and designed out the problem issues. When solid state power amps started becoming available, they quickly earned a reputation of being temperamental and unreliable. They were. Big amps were even more unreliable, hence the nicknames such as "Phlame Linear" 400 and 700 watt amps. Yet these were an order of magnitude better than the generation before. Now, BJT power amps are very reliable, and with better sonic performance. Pro and MI audio designers are way behind the learning curve compared to other ares of consumer electronics which lives on switch mode power supplies and Class D amps. They are getting better and the next generation will rival the best AB transistor designs. The voltage the switching supplies can generate is steadily increasing so big amps and even tube amps weighing less than 1/2 current amps will be the norm in a couple years.
How are the IPR Peavey 7 lb Class D amps holding up? I would guess better than the old Deca series. These new amps sure look appealing...lots of power in a rack that can be picked up with one hand.

[gbono]

I have to agree with you that the MI designs, using PWM supplies, are behind the curve. My experience with consumer class D audio designs have been with Ice Power from B&O and BASH. No issues but I assume these robust designs were/are to pricey for MI budgets. As usual, most of the backline bass amps at the NAMM show were class D (Ampeg) kit while the guitarist are still all tube. I wonder how many decades will need to pass before tubes are yet again obsolete (oops maybe that's blasphemy here).

[Chuck H]

Nah. Before I was rolling my own amps I had been heard to say "If they ever come up with an amp that sounds like tubes, but doesn't have tubes in it I'll never look back." That's not my current attitude on the subject though. But I can see amps that sound acceptibly similar to all tube/transformer designs that only have one or two preamp tubes and perhaps NO transformers becoming a reality in the near future. Tubes may still be needed for specific parts of the circuits to get the right effect but all other functions will be handled by SS devices. I must admit that my electronic knowledge hasn't kept up with some design trends. But it's not really my goal either. I can still make a great amp with tubes and transformers. But good on anyone wo develops affordable, reliable great sounding amps of any design.

[km6xz]

As there has been a race or power....just look at what big stadium shows in the 50s and 60s used, a couple hundred watts or 5000 seat, now any decent disco has 50,000watts(all to get 1-2 octaves lower) for a capacity of 250 people....tubes just can't keep up due to transformer limitations. If load matching could be done without 400 lb transformers there are excellent tubes for AF/AF AB operation the 300-50,000watt range. The one area of tubes that are still being developed and improved is very high power AF and RF tubes, over 3,000 watts in plate dissipation. They are very reliable and last a very long time running continuously as very full rated power. 50,000watt tubes are the cheapest tubes or power to cost ratio.
Tubes will likely be relegated to sound generators and not broadcasters(as in providing the sound coverage from a stage) for DIY'ers. For commercial products which can produce reasonably priced complex circuits the future does not look too bright for tubes. Tubes are in poor supply, the factories still in it can't deliver reliable consistent numbers or specs. Tube transfer functions are getting closer in modeling so there are a lot of options for the tube sound.
Here is an image of an amp I built using 3CX300 exterior anode tubes, running a very cool 200 watts but with 600 watts of plate dissipation capability . It is running conservatively for CCS service in Hi-Fi application. The same tubes were used in a bass rig, with blower cooling instead of clear passive oil cooling in this amp ran a very cool 350 watts but my transformer was the limiting factor. If a larger transformer was available it would have run 600 watts easy before clipping.
Here is a close up of the tubes in their oil chamber before the oil was filling the glass column.

After that, I turned to another direction for massive power for guitar and bass rigs, a stasis style amp. That was working from the concept of creating a large power amp using AB MosFETs, 1000 watts in the first version. The idea was to never allow the amp to clip or run into non-linearity. The signal source was a high speed comparator which had multiple servo loops. The input signal was feed into both the large power amp and a reference amp. The reference amp was a small tube amp, class A with a 6V6 running single ended into a small OT driving a load resistor with series inductor to simulate a speaker load( later version actually have a 6AQ5 output at less than 2 watts driving a real 5" speaker inside a damped box). This amp had no negative feedback applied, running wild. A sample of that signal was phase corrected and applied to the comparator. The Multiple loops essentially allowed a feedback loop to make the large amp clean, and the difference between the input signal from guitar compared to the output of the large amp, to generate a difference signal, with the out of the reference amp providing the input to the comparator. After a lot of tweaking the amp sounded like a AC 30 on steroids, 1000 watts with the same transfer function as a bare bones little reference amp. The idea was to be able to plug in a small module that was a low power version of different classic tube amps, such as a Champ, AC30, Bassman etc.
A smart company could sell personality modules or simple enough interface to allow DIY'er to create small plug-in reference amps using simple to change parts sections and values without having to deal with expensive transformer or power tube changes.
While looking for the liquid cooled amp I came across a photo of my home workbench at the time, all 17 feet of it:

My old shop had 15 such benches before I gave the business to the employees and moved here.

[Chuck H]

It's a valid idea. I've been penciling a (more pedestrian) similar idea. Low power clone modules run into a MOSFET power amp. Nothing like a thousand watts. These were still to be individual amps for players so I was going to go 200 watts. The modules would be the barest versions of classic designs with the exception of the output tubes which would probably be either single ended small bottles (6v6,el84) for "Champ", "AC4" or Kalamazoo or whatever and dual triode low mu 12ax run AB for push pull amps like Marshalls and such. The module signal would run an active load and be padded for an effects loop and/or run through a digital effects processor, then on to the SS power amp. The modules could be built DIY or purchaced and switched out by owners. I was going to call the venture "Analog modeling"

[km6xz]

Hi Chuck
The reference amp front end for a larger amp was the first attempt about 20 years ago when high power MosFETs became available but there was a clearly decernable character heard of the solid state amp, regardless of signal source. That is when I decided to use force the power amp to use the transfer function of the small amp, in using it in a high speed loop so the sound of the amp was nulled out by the comparison to the reference amp. It REALLY did sound like a AC30 even at low volumes. The 6AQ5 reference amp was not quite like any classic amp in sound, the small 5" speaker as a load had a lot to due with that but it broke up in a very pretty way, and is still being used gigging by the person I gave it to. A lot of bass players use tube front-end and solid state since they need much more power than guitarists but a lot of players would want the reserve power if they had it.
Personally I am tired of too much power on stage and what it does to vocals and clarity of the whole system since no mic can escape being blasted with stage sound. Even with ear monitors vocals are much poorer than when stage volume was a lot less. Concerts are often painful to listen to, and seem more of an ego trip for house mixers than entertainment of the audience. No one in the performance is in charge of what the audience actually experiences. No one comes out to the audience and bothers to notice if it is painful or being enjoyed, and if they did they would not have the authority to correct the situation.

[Chuck H]

+ to that. Still, 200 watts for a good clean sound isn't too much. The point is to have enough plus a bit. I wouldn't know enough about implementing a difference circuit. My plan was to simply recreate a suitable low power copy and amplify it with brute force. I can understand how some nuances of "feel" could be lost. Tone is so subjective though and a clean reference amp would be the goal for the power. I wouldn't be attempting to make a single ended tube sound like an AB circuit, I'd be using small AB tube power amps. With recent technology it might even be possible to do the whole thing without power transformers and only small output transformers. Output transformers are only because I can't concieve a way to load the power tube plates without implementing unwanted voltage sag with current differences at that point in the circuit. I do think it's possible. I just don't know how to do it. Also, the modules would be limited to power tubes (small ones), a phase inverter tube (for AB circuits) and the only other tube functions would be those in the preamp where tube behavior can't be accurately replicated with SS devices. Of course there would be an overall difference in tone between a classic AC30 and this modular system, but the goal would be to make those differences small enough to be acceptible while still providing a system with more power, volume versitility, modern features for effects processing and lighter weight. I thought it might even be fun to have space for dual modules for channel swiching. How about a clean channel BF Fender module and a wide open plexi Marshall for leads?!?