PROJECT: Complete RT3000p amp rebuild

ChrisD06

I've been working on this for a while now.

Essentially, I will be remaking and improving the original Polk schematics of the RT3000p 14-0017-0000 Rev B board, custom printing a new PCB, and installing all brand-new components to it. This board will also work with the PSW1200, which has only slight modifications to the design.

The end goal is to have a PCB design that can be printed by anyone + a list of components with their respective locations, and it will also be more reliable and easier to work on than the original. It will also use currently manufactured components which are 100% perfect replacements for the obsolete ones, which there are only a couple of.

I currently have the entire list of components, and the needed files to work on remaking the schematic. I'll copy over the original schematic in Kicad tomorrow and work on improving upon the design.

I'll then work on making a new PCB which will be double-sided (instead of the original single-sided board). This way, I can eliminate the need for all jumpers used except the D23 jumper (which is used for the PSW1200/RT3000p change), but also hopefully make the transistors mounted differently.

Currently, the transistors are bent at a 90 degree angle, and screwed into the heatsink. THE ONLY THING holding the amp board on is the solder joints (unleaded solder) on the transistors. I'm hoping to AT LEAST make the joints on the other side of the board, using leaded solder. I may also use the 4 extra threaded holes in the heat sink for some standoffs for extra support.

This should improve reliability greatly, but also let the transistors sit even more flush on the heatsink for better cooling. Also, because my model will be custom printed, all boards will support 4 filter capacitors.

I will be posting updates as I progress.

F1nut

I like it!

ChrisD06

Thanks, F1. I will need to get my hands on a PSW1200 at some point in order to confirm the minor changes on the amp board, as Polk's schematics appear a little unreliable.

Hopefully I can eventually work on new boards for the RT2000i/p, RT1000i/p, SRT, CS1000p, etc. I think this will be really fun, would be great when I apply or electrical engineering in university too!

ChrisD06

Okay so its been over 2 months and I've not provided an update, but I have a good excuse.

I'm completely reworking how I'm going to go about this. My plan for the amp board still stands and the schematic is halfway done, however I won't claim it works for the PSW1200 as I do not have one on hand and I'd rather physically see the differences between it and the RT3000p board than assume and hope.

I will be also making a custom crossover board for the RT3000p, and the RT35s.

Finally, I will be making a custom amp/crossover for the CS1000p as well.

Here are my goals:
1. Longevity - I expect them to last a long time without degradation, I will provide a parts list along with the PCB files once finished, and will offer a rebuild service as well.
2. Better thermals - I find the transistors get needlessly hot in the amps, so I will be ensuring they are perfectly flush with the heatsink and using a thermal paste such as Arctic MX-4 or Thermal Grizzly Kryonaut. The OEN stuff is crap and has barely any thermal conductivity.
3. Open-source - Because all documentation/files for these will be available, anyone can improve them in any way they'd like!
4. Improved sound quality - This stems from the crossover upgrades however bumping the capacitance slightly in the subwoofer amps should provide more punchy bass (faster attack) and also reduce AC noise with better filtering. Also, the transistors running cooler should allow them to provide more power without throttling to protect themselves.
5. Compatibility - The new boards will not use a different plate, transformer, or heatsink. It will just be new PCBs with 3D-printable parts (some) that can be assembled.

Anyhow, thanks for reading my over-ambitions. Don't expect another update for at least a month, I will need to juggle this project on my free time when I'm not working (just got a new job today) and between my last year of high-school.

ChrisD06

UPDATE!!!!!

I am working on an RT3000p crossover board. I have all of the parts picked out, and am just doing the PCB design. I have the schematic already and it's just a matter of making it as compact as possible while giving the resistors breathing room. I also need to find the thread spacing for the screws on the OEM board.

My goal was to aim for supporting the full 500W Polk claims these can handle, however the resistors simply won't handle it. The capacitors and inductors will all be able to easily handle the load however.

I am using Mundorf ECap Raw 150uF, four of them, in parallel for the 600uF 80Hz HPF (the best electrolytics I can find, the film ones are insanely expensive), ClarityCap CSA for the caps (tweeter cap is bumped to 9.1uF) and the inductors will end up being Jantzen Audio ones. L2's DCR will be 1.38, which is 10% higher than the 5% tolerance polk has for the 1.25 listed value.

Unless someone knows of a 3.0mH, 1.25 DCR inductor I could use? Air core or laminate, I don't care. @VR3 you're a smart cookie, I'd like your input.

I might not wait until summer to fully build these. I might do it around Christmas. I'll be using Cardas Quad Eutectic solder as Jesse seems to recommend it alot and I might as well go for the best I can here.

VR3

Hifi collective carries all of the available sizes Jantzen makes to hit your target

To increase resistor power handling, take 6 resistors or more in parallel and then bundle them around a dummy resistor in the middle.

ChrisD06

VR3 wrote: »

Hifi collective carries all of the available sizes Jantzen makes to hit your target

To increase resistor power handling, take 6 resistors or more in parallel and then bundle them around a dummy resistor in the middle.

You're the best. The resistor idea is really neat but I think I'll just use one on the first design to keep costs down. I could always revise the board to support the MRC-50 down the line or even to support 6 parallel resistors and Sonicap Gen I or something.

I'm currently going for a bang-for-the-buck design. It involves some paralleled capacitors, some slightly off values, and an electrolytic 80Hz HPF, but it'll be damn good compared to stock and the differences over a balls-to-the-wall design won't be as big as OEM to my current one. I would go all out but I couldn't afford it anyhow LOL

VR3

You can pick up 6 Jantzen mox for the cost of one mills resistors and the performance is as good or better, imo

ChrisD06

VR3 wrote: »

You can pick up 6 Jantzen mox for the cost of one mills resistors and the performance is as good or better, imo

I'll check them out. Do I need a dummy resistor, couldn't I just connect 6 in parallel to the traces or no? I'm trying to get a super clean and serviceable layout.

VR3

You can use the old resistors as the dummy

ChrisD06

VR3 wrote: »

You can use the old resistors as the dummy

Right. But is it needed or do I just not connect it? Like could I just have the 6 resistors without a dummy at all? Also wouldn't them being in parallel lower the resistance or do I just use higher resistance ones to counteract that?

VR3

So as an example, 6 x 20 ohm resistors in parallel would be 3.3 ohms

The dummy resistor, you would cut the leads off.

I recommend using the dummy resistor in the middle as this keeps space between all the resistors and that dummy resistor will act like a heat sink of sorts and it is easier to keep everything neat and tidy.

You will have minimal heat build up going this route at all volumes

ChrisD06

VR3 wrote: »

So as an example, 6 x 20 ohm resistors in parallel would be 3.3 ohms

The dummy resistor, you would cut the leads off.

I recommend using the dummy resistor in the middle as this keeps space between all the resistors and that dummy resistor will act like a heat sink of sorts and it is easier to keep everything neat and tidy.

You will have minimal heat build up going this route at all volumes

Ah gotcha. Could I just put a spot for 10 of the 10W Jantzen Mox resistors and find the values of them to create 1.5, and 1.0 ohms respectively? It's gonna take up a ton of space but it'd be cool at least. That would enable 100W power handling for the tweeter assuming you had the speaker running at 500W and the tweeter were to draw 20% (generally the highest it will draw)

ChrisD06

I quickly drew up a schematic just to see how absolutely ridiculous it would look and I love it.

I'm thinking, if I went with 10 of these Jantzen resistors, I could put the inductors and capacitors on the bottom board, and then make a board on top for just the resistors (for better cooling and because there would be like 30 of them), and then stack the two of them and have an interconnect cable of high guage wire. It would look insane (and cool) and actually function pretty well.

VR3

Yep, you would have maximum power handling but zero driver protection

ChrisD06

VR3 wrote: »

Yep, you would have maximum power handling but zero driver protection

Oh yeah I'm aware. I don't think the SL6502 can handle 100W anyways.

First revision will just be Mills 12W. Also I'm really struggling to find an inductor at HiFi collectivie which is 3.0mH and 1.25 DCR within 5%. Maybe I'll have to use the stock one.

I'll work on the board this weekend if I have time. Will use super thick traces and keep them as short as possible. I'll also try to keep the trace lengths equal and the board will be as compact as possible.

VR3

Really with inductors as long as the dcr is within 10% you will be fine, imo

ChrisD06

VR3 wrote: »

Really with inductors as long as the dcr is within 10% you will be fine, imo

Awesome. I'll order everything from HiFi collective in that case whenever I'm ready to do this.

This is gonna be a big project, I'm pretty excited to hear the final results though.

ChrisD06

Holy crap, everything is within 5% here and DAMN its going to cost me close to a grand for this entire rebuild... I think I should switch the 1mH to iron core and the mills to the Jantzen Mox 10W.

VR3

Iron cores are nasty

ChrisD06

VR3 wrote: »

Iron cores are nasty

I have no real choice if I ever want this to happen. I can't shell out $25 per coil.

Switching the 1mH to iron core and the Mills to MOX, the price is now $114 cheaper. That's insane.

VR3

Best 114 dollars ever

VR3

Do it once do it right

ChrisD06

VR3 wrote: »

Do it once do it right

Good point. I'll see what I can do in that case.

ChrisD06

Okay folks, another update. The schematic is complete and I'm taking measurements for the RT3000p crossover now.

Here's an idea: external crossover. Hear me out.

So the idea will be that there will infact be 2 sets of binding posts on the speaker. One for the tweeter, one for the woofer. These will go directly to the drivers, no crossover at all.

The crossover will be mounted externally. The speaker cables will plug into the crossover and then the crossover will plug into the corresponding binding posts for each driver.

By doing this, you not only have a massive, high quality crossover with 500W power handling (or damn near close), but you also can truly bi-amp the speakers with external crossovers and amps.

What are your thoughts? Too overkill?

VR3

Lose the binding posts between speakers and crossover, use neutrik nl4

ChrisD06

VR3 wrote: »

Lose the binding posts between speakers and crossover, use neutrik nl4

What about disconnecting the crossover to use an external crossover? How would I do that?

VR3

exactly as you described

shopdog328

Reading your post with interest. Also have a pair of rt3000p’s with a fried amp. Thinking about just repowering them. Looking for a little advise. Looks like the MW8500 woofers are wired in series, are they at 8ohm impedance?
Thanks!

ChrisD06

shopdog328 wrote: »

Reading your post with interest. Also have a pair of rt3000p’s with a fried amp. Thinking about just repowering them. Looking for a little advise. Looks like the MW8500 woofers are wired in series, are they at 8ohm impedance?
Thanks!

Correct. Two MW8500 subwoofer drivers wired in series.

I believe each subwoofer is a 4ohm impedance, and the total load with both woofers on the amplifier is 8 ohms. Someone correct me if I'm wrong.

What's wrong with the amps? Could you possibly describe the symptoms?

shopdog328

Thanks for the reply and info

When powered on the speakers hum in an intermittent pattern. About 1 second off and one second on. Had one rebuilt about 15 years ago by Polk,been using the system for a downstairs HT very sparingly, bit now my kids are older and want to use it more.

Toying with picking up a Dayton plate amp or a Bergenger nx3000 amp.