Inside a compact class D amplifier

I never cease to be amazed at the lack of a massive heavy transformer and heatsinks forming part of the case. I made a HV source for calibrating old mutimeters and reforming caps a while back - sine-wave generator feeding a cheap 120W AliExpress class-D amplifier driving a pair of 380V-6V transformers. Here are some pics if anyone is interested.. https://drive.google.com/drive/folders/1Az-rTFB6dvECvZ2dCITVOYaYa44xc_0f

Gordo

Most 19" rack things (servers, switches) are front-to-back coolers. Wasn't always that way 20 years ago but is now (some switches go in the back side of the cabinet so are back-to-front). My assumption here is that, given the "filter" on the front that the fan is drawing air through that foam, across the heaters, and out the back.

Peter Laws

Class D amplifiers exist in the limbo between analog and digital and as such I find are great ways to start internet arguments between audiophiles. They generate much popcorn worthy entertainment in comment form.

that guy

I'm not sure how hot they will get under full load. But it always makes sense to have good ventilation and a clean environment for stuff like this.

Big Clive

I wonder what the heat output is from a rack full of those at full blast?

Phil Collins

I think the XLS range is prosumer, but it does for me. Nobody misses old crates.

Mike Page

I've not tested the unit due to the lack of suitably rated speakers. You've made me realise that the temperature sensor on the amplifier heatsink might just be for fan control if it doesn't signal an ovetemp state too. The first digital amps I used on an installation were Crown and they were night and day to the old ones with massive toroidal transformers in them.

Big Clive

Sorry, I'm travelling with work now. I did look them up and they were very standard dual and quad low noise op amps.

Big Clive

Sorry, to clarify at least in PC circles the fan cage is on the output side (presumably so the fan's not trying to tear itself apart) and that airflow doesn't look like out would be too bad.

nowt

Thanks, Clive. I was looking forward to this one. I've a Crown XLS 1502 as a jobbing amp and it's great. I didn't plan to get DSP but since it has it, I use it. (Would be even better with delay for time-aligning.) The fan usually stays off making it very friendly. But the price has shot up and I might need to get a 2 channel amp for an install. Is the Pulse unit's fan annoying?

Mike Page

Do you have a picture of the balanced input circuit board as well? I'd be interested to know what type of opamps they're using.

I can never quite get over the teeny weeny PSU caps on Class Ds. Teeny weeny everything in fact. Those 15 vias caught my eye though.

Mike Page

It's common to have an agreed direction for racked equipment. Otherwise the air risks going in circles getting hotter and hotter. Front in back out may well be the default.

Mike Page

Back in the day at Megger they used a vacuum chamber to emulate altitude for special tests. I remember babysitting one such test on an OTS - oil test set. The mains supply was taken into the chamber over a ribbon cable. The pump and seal together managed to keep enough vacuum. I can't remember the altitude, but my guess is 2000m. It's higher now. There are so many places higher than that.

Mike Page

It probably is. I haven't powered it without a load, even though it's probably safe to do so.

Big Clive

Very clean design.

Bill Kerr

Not just thin wire, but also the dreaded IDC connectors that just cut into the side of the wires. It was the curse of the Best Boy lights by PRG where they used ribbons for motor connections.

Big Clive

The individual conductors in those ribbon cables tend to be fairly small. Even with 6 of them for + and -, it seems odd to me to use the ribbon cable for power in a device that may consume 350W. Is it actually odd? Or is it normal in newer equipment?

Charles

Thoughts on the unidentified chips: The comparators should be used in conjunction with a triangle or sawtooth wave generator which converts a voltage level into a PWM duty cycle. Are they connected in such a way that you are able to infer that kind of relationship?

Raven Luni

I think the fan is blowing out and the foam is dust filtration?

nowt