Stalled REVOCCASES RCC-PWR1 / ATX PLUG-IN 400W DC-DC power supply (PICOPSU)

[REVOCCASES]

This project has been stalled in favour of the new 600W plug-in: GxR-ONE

Production - GxR-ONE ATX PLUGIN PSU / a (G)UNIQUE and (R)EVOLUTIONARY power solution / RTX3080 ready

G-UNIQUE and REVOCCASES have teamed up to bring you a (G)UNIQUE and (R)EVOLUTIONARY power solution for your SFF build. GxR-ONE - a 600W DC-ATX PLUGIN PSU Limited quantity now on stock! Technical Data: Input: 12VDC (11.8VDC~12.3VDC) max 50A PCBA overall dimensions (mm): 38H x 55W x 14D...

smallformfactor.net

 

[oOflyeyesOo]

Oh fancy! What all does the secondary board hold?

 

[REVOCCASES]

Oh fancy! What all does the secondary board hold?

Thats the AC-DC MEANWELL PSU we can probably combine with the plug-in.

 

[smitty2k1]

I'd be really excited if you bring this to market - but I have one question. Why run the PCIe power through the plug in board? That seems unnecessary. The 12v could just go through the Meanwell's additional outputs directly to the GPU, no? I think the Meanwell has a 5v standby to turn on/off with the rest of the system?

Maybe it's just an issue with using a mostly off-the-shelf solution from RGEEK.

 

[REVOCCASES]

I'd be really excited if you bring this to market - but I have one question. Why run the PCIe power through the plug in board? That seems unnecessary. The 12v could just go through the Meanwell's additional outputs directly to the GPU, no? I think the Meanwell has a 5v standby to turn on/off with the rest of the system?

Maybe it's just an issue with using a mostly off-the-shelf solution from RGEEK.

Thanks for your feedback. If there is enough interest in the community I can offer it here. I'm looking for MOQ 50 pcs.

True, you could make it work like that but most plug-ins do it like the rgeek. Like HDplex 200 or gunique. I think one reason is to make it easier to use for the customers. And not everyone is using that meanwell with remote on/off function which makes it a bit more flexible imo.

PS: the tests I did today showed one more reason: the board helps to keep the ripple within atx specs. So you don't have to rely on the AC-DC psu here.

 

[REVOCCASES]

Sooo, I finished my DIY electronic load tester today and could do a few tests...

Test setup
- AC-DC PSU: MEANWELL RPS-400-12 set to 12VDC output (well, 19.995)
- Variable electronic load connected to the 12V rail of the DC-DC plug-in PSU
- Measured on the 12V rail of the plug-in PSU: load (A), ripple and voltage drop (mV)

Test 1: RCC-PWR1 prototype

I tested the voltage drop and ripple at 10A / 20A / 25A load on the 12V rail. Ripple stays well within ATX specs (below 120mV). Same as voltage drop which also stays within the 5% ATX tolerance (11,40V to 12,60V).

Test 2: MEANWELL RPS-400 max load test

Voltage drop at 35A was only 60mV and ripple well below 120mV.

Test 3: I found a second-hand J-H PLUGIN160 for testing. I know it's not meant to power something very powerful but I wanted to see how far I could push it.

Result: you can draw up to 5.6A (67W) from the 12V rail before it runs into over current protection and switches off. That's a good result because there are safety mechanisms to protect the board and your PC from blowing up. It's really a good quality board but unfortunately not powerful enough for what I have in mind.

 

[Thehack]

Sooo, I finished my DIY electronic load tester today and could do a few tests...

Test setup
- AC-DC PSU: MEANWELL RPS-400-12 set to 12VDC output (well, 19.995)
- Variable electronic load connected to the 12V rail of the DC-DC plug-in PSU
- Measured on the 12V rail of the plug-in PSU: load (A), ripple and voltage drop (mV)

Test 1: RCC-PWR1 prototype

I tested the voltage drop and ripple at 10A / 20A / 25A load on the 12V rail. Ripple stays well within ATX specs (below 120mV). Same as voltage drop which also stays within the 5% ATX tolerance (11,40V to 12,60V).

Test 2: MEANWELL RPS-400 max load test

Voltage drop at 35A was only 60mV and ripple well below 120mV.

Test 3: I found a second-hand J-H PLUGIN160 for testing. I know it's not meant to power something very powerful but I wanted to see how far I could push it.

Result: you can draw up to 5.6A (67W) from the 12V rail before it runs into over current protection and switches off. That's a good result because there are safety mechanisms to protect the board and your PC from blowing up. It's really a good quality board but unfortunately not powerful enough for what I have in mind.

It shouldn't be turning off at 67W. I have load tested them for 150w.

 

[REVOCCASES]

It shouldn't be turning off at 67W. I have load tested them for 150w.

Well, I got it second-hand. Maybe there was already something wrong with it...

Did you test the 12V rail with 150W or the complete unit?

If you have a picture or something with your load test I can also put it here. I don´t want to publish something false.

 

[Thehack]

Well, I got it second-hand. Maybe there was already something wrong with it...

Did you test the 12V rail with 150W or the complete unit?

If you have a picture or something with your load test I can also put it here.

I use a regular mobo with 75w gpu, and use a killawatt. The 12v rail is 16A rated. But regardless, I hope it is just a singular defect.

 

[REVOCCASES]

I use a regular mobo with 75w gpu, and use a killawatt. The 12v rail is 16A rated. But regardless, I hope it is just a singular defect.

wow, 16A @ 12V this would be almost 200W on the 12V rail!

I will repeat my test next week to make sure it wasnt my fault with the measurment.

PS: if it helps, the LED starts blinking if I draw more than 67W from the 12V rail and then it shuts down

 

[Thehack]

wow, 16A @ 12V this would be almost 200W on the 12V rail!

I will repeat my test next week to make sure it wasnt my fault with the measurment.

PS: if it helps, the LED starts blinking if I draw more than 67W from the 12V rail and then it shuts down

I think it may be a defective unit. I do have a newer design that is 30A rated.

 

[oOflyeyesOo]

@guryhwa what do you think about the meanwell units?

 

[REVOCCASES]

I do have a newer design that is 30A rated.

Sounds good, why you didn't mention this when we were discussing about your 200X? last time I asked you if you have something more powerful and you said we would have to develop one first if I needed 300W or more. Would you mind to share some test results of the 30A version here?

 

[Thehack]

Sounds good, why you didn't mention this when we were discussing about your 200X? last time I asked you if you have something more powerful and you said we would have to develop one first if I needed 300W or more. Would you mind to share some test results of the 30A version here?

The 200X is actually 30A. The issue with a 300W version is that for modular connectors, you need two boards, like HDPLEX, or an extra big one.

You can go with a cheaper solder on design, as well, but that is not my style.

Regardless, the primary issue with releasing 300W version is the lack of a supporting AC-DC.

 

[REVOCCASES]

The 200X is actually 30A. The issue with a 300W version is that for modular connectors, you need two boards, like HDPLEX, or an extra big one.

You can go with a cheaper solder on design, as well, but that is not my style.

Regardless, the primary issue with releasing 300W version is the lack of a supporting AC-DC.

I'm ok with a solder on solution because it safes space. So if the 200X can really deliver 30A on the 12v rail I would be fine to buy from you and make some modifications so that people can use it like my modified rgeek or the gunique. But I would like to see some test results or have a sample to test it myself.

Concerning AC-DC there are many choices above 300W that would be suitable if the case has a little bit more space available (like my MID1 for example). There are also many custom builds using the gunique with some powerful AC-DC units.

Would you mind to publish some test results for the community or send me a sample for testing? Some of the upcoming APUs and CPUs with integrated graphics can draw a lot more power then 200w under boost.

 

[REVOCCASES]

I made a size comparison of the different power supply solutions available which are capable of >=200W output.

some examples (not considering cables and connectors):

HDPLEX 400W Set internal mounted: 0,43 liter + 0,20 liter = 0,63 liter

FLEX 400W (non modular): 0,54 liter

HDPLEX 200W Set internal mounted: 0,43 liter + 0,05 liter = 0,48 liter

RCC-COOLMOD1 & RCC-PWR1 400W Set internal mounted: 0,55 liter + 0,04 liter = 0,59 liter

If you would use the RCC-COOLMOD1 to build a nice and small external brick and combine it with the internal RCC-PWR1 400W ATX PLUGIN you would make up lots of space inside your case for e.g. HDDs, Coolers, Fans, Pumps, etc... just think about the pros and cons for a moment ... ?

 

[Yugo555]

Would you mind to publish some test results for the community or send me a sample for testing? Some of the upcoming APUs and CPUs with integrated graphics can draw a lot more power then 200w under boost.

I have a doubt about that. They will have Zen2 architecture with 35w and 65w stock TDP variations. Even with the overclock they will hardly pass over 100w-150w. Since side components doesn’t effect the consumption significantly, 200w PSUs will mostly be preferred by users. Also I don’t see any reason to stress the APUs with heavy overclocks. APU cases are small and relatively less ventilated. If people are imagining to reach over 100fps at 4K resolutions in the recent games by only getting a 200-300$ APU, that is just a dream for now.

 

[REVOCCASES]

Even with the overclock they will hardly pass over 100w-150w.

OK, here is an example from Intel:

Core i7-10700 K

It has a PL1 of 125W but a PL2 of 229W with TAU of 56 seconds. This means during boost it can draw up to 229W for almost one minute. The Core i9 10900 K can even go up to 250W. If you adding some other components you end up with something like 300W even without dedicated GPU.

Back to AMD: they again calculating their "TDP" differently, a so called "65W APU" will also draw much more power than just 65W if its pushed. We´ll have to wait for some first reviews to see how much exactly but I prefer to have a little overhead.

If we see something like Intel is doing, I would prefer to have a more powerful power supply:
Core i7 10700F, "65W" CPU: Boost power consumption = 224W for 28 seconds

If someone is using this kind of CPUs / APUs with a 160W or 200W PSU because they think its just a " 65W" CPU/APU it´s likely to go into OC protection during boost states and shuts down.

 

[Yugo555]

No no, don’t take that intel CPU as a reference. Noone builds any computer with intel’s integrated graphic CPUs. There are some project threads in the form where users share their power consumption tests with APU builds. Especially Velka and Lone’s threads, you can find some references. (Btw APU refers only AMD’s CPUs with integrated graphics).

 

[REVOCCASES]

No no, don’t take that intel CPU as a reference. Noone builds any computer with intel’s integrated graphic CPUs. There are some project threads in the form where users share their power consumption tests with APU builds. Especially Velka and Lone’s threads, you can find some references. (Btw APU refers only AMD’s CPUs with integrated graphics).

Well, lets wait for first reviews of the next gen APUs. I´m pretty sure we will see similar power peaks during boost states. Besides, this power supply is not exclusively meant for APU builds. It also comes with PCIE power connector for graphic cards.