Hi I have an IBM 330-450DX2 with a dx2 processor at 50mhz. I have a dx4 at 100mhz running at 3.3v. On the board there is a vrm pin which is simply a jumper that lets the 5v through. I have been looking for a 5v to 3.3v vrm for this board and can't find it anywhere. Does anyone know where I can find it? Thanks.
I doubt it´s just a jumper with 2 pins. Any VRM will need a ground connection also. For lower resistance at higher currents there are multiple pins for input and output in many cases.
Do you have a picture available?
majestyk wrote on 2021-05-01, 10:55:I doubt it´s just a jumper with 2 pins. Any VRM will need a ground connection also. For lower resistance at higher currents there are multiple pins for input and output in many cases.
Do you have a picture available?
yes.
Well that´s 10 jumpers = a jumper block. Most of them are connecting 5v to the CPU socket some are ground connectors.
Finding the original (proprietary) VRM will be quite a challange.
A DIY VRM is an option if you can handle a soldering iron well.
The easiest way would probably be using an interposer socket that has the 3,3V voltage regulator onboard. These are offered on Ebay frequently.
I found this schematic to create a homemade vrm, but it is not the same as my board, it is a single row of 10.
http://john.ccac.rwth-aachen.de:8000/misc/lacuna_vrm/
You can use this schematic, but you will have to adapt the pinout.
First you must measure the voltages of your 20-pin connector.
1. Remove the jumper block and the CPU
2. turn on the machine
3. measure with a voltmeter between ground and the pins to determine which ones are the 5V input pins
4. measure between one of the +5V rails of the PSU connector and the pins to determine which pin(s) are ground potential (you will measure full 5V there)
5. If you´re lucky and measure with an ohmmeter (machine now turned off) between the remaining pins (that are NOT 5V and NOT ground) they are all connected = 0 ohms. Then these pins are 3.3V output pins.
I just made some similar VRMs here:
What´s wrong with this FIC 486 PAK-3 ??
I just measured bottom row 1 and 2 ground, 3 to 10 5 v. I guess I need to make a pin that takes all the 5v outputs and the top row comes out of 3 to 10 3.3v. pin 1 and 2 I leave continuity.
narcotic wrote on 2021-05-01, 15:00:I found this schematic to create a homemade vrm, but it is not the same as my board, it is a single row of 10.
http://john.ccac.rwth-aachen.de:8000/misc/lacuna_vrm/
Turned up this page while looking for something else, then found this thread when seeing if anyone already posted about it. So bump for attention to this little design that may allow quick and dirty VRM adapters for many a motherboard.
It further appears that because it's a low voltage differential design that that chip or it's siblings might work for Pentium regulators for lower core voltages or split core. However it may limit the usable CPU to under 200Mhz due to power dissipation ability. So while you might get split core MMX working on a P133 motherboard, you might only be able to go as far as P166MMX. (Worth investigating also are eBay DC-DC regulator modules that might regulate to 2.8V nicely from 12V with good efficiency)
I have a PC330 machine with similar voltage reg header as OP and might in time get around to adapting this design to use with that one. However, I have several other boards I could be running DX4 class CPU on and not so many 3.3V CPUs, so I am not feeling personal urgency to go anywhere with it for that purpose at the moment. Besides DX2-66 machines gotta be DX2-66ing to keep the dream alive right? 🤣
Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.
What about using an ATX PSU with an adapter and taking the 3.3v directly from there?
An option:
Re: Socket 1/2/3 Voltage Interposer Tweaker (Alpha, seeking testers)
With the pinout for that header a module could also be made.
tauro wrote on 2022-12-25, 17:05:What about using an ATX PSU with an adapter and taking the 3.3v directly from there?
Re: ATX to AT pico Adapter! + Fan Headers, -5v, and 3.3v (Released)
😀
tauro wrote on 2022-12-25, 17:05:What about using an ATX PSU with an adapter and taking the 3.3v directly from there?
Would the 3.3V provided directly by a rail on an ATX PSU ( switch mode ) be clean and stable enough for a CPU to operate properly ?
darry wrote on 2022-12-25, 18:24:Would the 3.3V provided directly by a rail on an ATX PSU ( switch mode ) be clean and stable enough for a CPU to operate properly ?
3.3V CPUs are usually specified at +/-10%, so 3.0V to 3.6V. The bad copies are rated at 3.45+/- 5% instead, so just 3.3V to 3.6V. If you don't have a CPU that requires "increased voltage" (e.g. most non-MMX P200 CPUs required increased voltage, and upping the voltage surely helps 5x86 CPUs to get to 160 or 180), powering it straight from +3.3V on the ATX supply is fine, as long as the caps in the power supply are not totally cooked or crooked. No matter what's the power source, local decoupling caps at/near the CPU sockets are necessary, too.
I have used a DC-DC step-down based on an XL 4005 that costs about €3.
If the motherboard has the option to use the 3.3v you only need to know where to solder the wires. Why do something new when it already exists and is so cheap and they are also efficient and offer the possibility of controlling the voltage.
For the power consumption of a 486, 5A is enough, tested with a 486 DX4 100.
