Built a custom D8B external power supply

[RJH_MUSIC]

Yes perfect sense. Gotta thank for for all of the help. That is the power supply I bought, but I could not get the voltage up past 15.33v.

[RJH_MUSIC]

I just posted these photos in the other thread for new OS. Been a long road but I finally finished the fabrication putting all of the parts from the existing CPU into a new 2 space rack unit CPU.

[RJH_MUSIC]

I confirm today by going to trc's website that I did indeed by the incorrect 5 volt power supply and that's why I was having trouble. I bought the RS 100 - 5 which is not a switching power supply so it could not supply +5/-5 sense. To make matters worse the unit that I bought was a hundred Watts apply but only 16 amps which would replace the Aztec power supply which is rated at 22 amps. I poured over the schematics from the DAV today and realized that predominately everything in the console uses + 5. Here are a few things that use plus 12 to fader Motors being one of them and yes the +/-16v are all the analog audio. So I have a feeling the new 5 volt switching power supply I bought That's rated at 30 amps Max, will do the trick. Interesting Lee enough if I were to explore the route that you're taking with all new power supplies I think I could still fit them in this box. As I mentioned in the other post and most difficult part of the fabrication on this was the backplane because I had to mount the existing video card horizontally because none of the current low profile cards would work and I still haven't tested the modem or the mini card. But the OS does boot to working software. The second hardest part of this was cutting a perfect hole in the front grill to mount the d8b power switch. Men even if I wind up using all the original d8b Power Supplies except for the 5 volt the win is did you get a smaller footprint in a 2-unit rack case and much better Cooling. Lastly I'm going to bring a 50-foot cable through the back of it so that I can locate the physical unit far away from the console which should also help noise.

[RJH_MUSIC]

And I forgot to mention that the only thing I have left to mount in here is the Transformer. I would like to explore what you're doing with separate power supplies and maybe I'll tackle that after this is finished.

[doktor1360]

Have you tried this card? Just use the bezel from the box your housing it in...

https://www.newegg.com/ati-radeon-7000-32mb-agp/p/N82E16814102695

[RJH_MUSIC]

I thought I did but I do not recognize the s video output so maybe mine was different. I will have to try it when they are back in stock.Thanks.

RJ

[doktor1360]

I thought I did but I do not recognize the s video output so maybe mine was different. I will have to try it when they are back in stock.Thanks.
RJ

Unfortunately RJ, like a dumb-ass I didn't look at the New Egg link when I posted it - apologies. They're not gonna get that card back into stock, it's old as the hills (like me) and getting harder to locate almost by the day. That being said, here's two I found on eBay that would turn the trick - I'm just not sure of their footprint, but they sure look low profile on the back plane...

https://www.ebay.com/itm/143142569293?epid=1101792203&hash=item2153f6454d:g:8acAAOSwKxVcbxZN
https://www.ebay.com/itm/304149351837?epid=3012051770&hash=item46d0b6cd9d:g:HBEAAOSwTpdhQ5X~

They're both 'genuine' ATI cards... Diamond, Sapphire & HIS work as well. I didn't do a search for those, however what I did query quickly verified that they're out there, it unearthed these two examples. Personally, I have ATI & Sapphire Radeon 7000's installed in my D8B / HDR respectively...

All the best, keep the thread updated if you will... \m/

[arjepsen]

Ok, a few more findings today:
I believe I got the +/- supply figured. There is still some noise/hiss from each channel. It adds up quite a bit the more channels I turn up - which of course is somewhat normal, and if I compare it to my brothers console, I actually get less noise with the new supply.
One thing that bugs me though, is that there is some high-pitched whine in the noise - which most likely comes from one of the supplies. From reading on forums I get the impression that smps's (switch mode power supplies) are potentially quieter and much more effecient that linear supplies, but there are specific issues to be aware of.
Here's what I read about it:

You must load an SMPS with at least 20% of the rated load or it will modulate the output at a low frequency. For smaller SMPS this frequently manifests itself as a whistle in the 1-3kHz range. For larger supplies, it can create massive noise at a wide range of frequencies. And note that at some point you can't just add a load resistor because it will need to be huge. So you need to pick an SMPS that supplies the right current to begin with. It should be selected for at 80% of load. SMPS work better when loaded. Even to 100%.

So this might be giving an issue with the 5 and 12V supplies.
The 5V supply that Ralph Weritz recommended in his pdf is an 80W 16A meanwell (a brand that people seem to recommend). When I measured the idle current drawn from the 5V supply, it was around 2,6A, so that's 5 * 2,6 = 13W.
20% of 80W is 16W, so we're 3W below. I might have to get a less beefy 5V supply.

However the 12V supply might be worse. I haven't checked the schematics thoroughly, but one thing it definately does supply, are the fader motors. In some simple tests I made, I could draw up to about 3.4A when really giving them a workout - let's assume it could peak around 5-6A. However, when not using the faders - no real current was drawn.... so this might be the culprit of the whiny noise.
On the power distribution board in the desk, there are a few "spade" connectors. One goes to chassis, and then there's Agnd (analog ground), Dgnd (digital ground) and Mgnd (motor ground).
I've connected Dgnd and Mgnd, as described in the document by Ralph, but since we're still using the audio circuits, this might not be good. Need to investigate this more.

Anyways, just wanted to update on my findings

[RJH_MUSIC]

Interesting news. I am still waiting for the +5 VDC supply i purchased with sense to be delivered. SO let me describe what happened when i built up a power supply based on Ralph's post and instructions for supplies. Maybe my problem was not related but I assume it was due to the +5 not having sense and being way underpowered. I posted a picture of the power supply I built early in this thread. I took a 1 RU case and mounted the three power supplies recommended by Ralph plus one more, a +48 VDC for phantom power. when I turned it on, everything came alive so it was very promising. I then connected the Data cable and botted into 5.1 OS. The first thing I noticed was that in the software, the faders were jumping all over the place but the hardware faders did not move. I then tried to turn pan-pots and got no reaction but the mixer screen in the GUI should pots jumping around as well.

The only thing I added to Ralphs setup was a fuse board inside the case with 4 fuses, one for each supply rated at 6-8 amps, I connected the 2 purple wires to to -12 and the 2 yellow wires to +12 on the Meanwell. I connected the 4 orange and 1 orange/white (sense) to the +5 terminal and the 4 black and 1 Black /white to the -5 terminals of that supply. I originally thought my problem came from the +/- 15V supply but could not have been because that is the +/-16V which is analog audio, not lights and fader motors. But i had not yet built up the vellemen kit. Is it possible I had a loose connection somewhere that caused the problems I had? SO later, i looked at the original Aztec LPS-152 and noted that it's output is 22 amps, as where the meanwell only outputs 16 amps. And I thought that was the problem, which is why I purchased the 30 amp SMPS power supply with sense.

So are you thinking my problem is unrelated?

RJ

[arjepsen]

Interesting news. I am still waiting for the +5 VDC supply i purchased with sense to be delivered. SO let me describe what happened when i built up a power supply based on Ralph's post and instructions for supplies. Maybe my problem was not related but I assume it was due to the +5 not having sense and being way underpowered. I posted a picture of the power supply I built early in this thread. I took a 1 RU case and mounted the three power supplies recommended by Ralph plus one more, a +48 VDC for phantom power. when I turned it on, everything came alive so it was very promising. I then connected the Data cable and botted into 5.1 OS. The first thing I noticed was that in the software, the faders were jumping all over the place but the hardware faders did not move. I then tried to turn pan-pots and got no reaction but the mixer screen in the GUI should pots jumping around as well.

The only thing I added to Ralphs setup was a fuse board inside the case with 4 fuses, one for each supply rated at 6-8 amps, I connected the 2 purple wires to to -12 and the 2 yellow wires to +12 on the Meanwell. I connected the 4 orange and 1 orange/white (sense) to the +5 terminal and the 4 black and 1 Black /white to the -5 terminals of that supply. I originally thought my problem came from the +/- 15V supply but could not have been because that is the +/-16V which is analog audio, not lights and fader motors. But i had not yet built up the vellemen kit. Is it possible I had a loose connection somewhere that caused the problems I had? SO later, i looked at the original Aztec LPS-152 and noted that it's output is 22 amps, as where the meanwell only outputs 16 amps. And I thought that was the problem, which is why I purchased the 30 amp SMPS power supply with sense.

So are you thinking my problem is unrelated?

RJ

Well, first: I did another measureing on my brothers desk - which is fully loaded with analog tape cards, and four fx cards.
The 12V line measured about 5A peak during a fader calibration.
The 5V line measured about 8.5A....
Anyway, the info I quoted earlier states that smps's work best under 80% load, and you shouldn't be too afraid of hitting 100%. 12 * 5 = 60W, 5* 8.5 = 42.5W. However, I wouldn't be surprised if the 5V line could shoot up higher, depending on what the desk is doing, so I think I might get a couple of 75W supplies for those two.
There's still the tricky part of loading the 12V supply enough - I'm investigating a "current shunt" circuit that might be able to make sure that the 12V supply is loaded just enough constantly.

Back to the problem you describe:
From my current messing about, I've experienced the issues you describe quite a lot. The faders and pan-pots are powered by the 12V line - and with some chips actually converted to a regulated 5V.
Now the groundings are very important (especially in regards to using smps's). In Ralph's document, he shorts the Mgnd spade to the Dgnd spade on the power distribution board - but he also is using the supplies inside the console. It seems you are connecting them back at the cpu rack? The Mgnd (motor grounds/faders ) needs to be referenced to Dgnd - otherwise you will experience the unstable behaviour you describe. I think normally the Mgnd has a connection to the rack chassis. In my case, I'm trying out using a 100R resistor between the two spades in the desk - in order to try to seperate them a bit, and hopefully this can help to have less noise from the 12V creaping into the audio circuits.
To make it short: make sure that your supplies have a common ground point (BUT only one!).