Precipitator HV Supply Hacking

This forum is for specialized infomation important to the construction and safe operation of the high voltage electrical supplies and related circuitry needed for fusor operation.
Rex Allers
Posts: 379
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Precipitator HV Supply Hacking

Post by Rex Allers » Thu Aug 29, 2019 10:28 pm

Looks like my thread about hacking/reverse-engineering of the Chinese Precipitator supplies is one that got lost in the forum move. I think I started it a little over a week ago ~8/21.

I see Mark Rowley's thread about success with Precip supplies survives, but I don't see his thread about these supplies then failing for him. I think the failure thread was more than a week ago.

When I post a long message to the forum I usually write it locally on my PC and then paste it into the forum message compose box, so I have a copy of some of my threads. From that I'm now reposting my thread using the same name:

Precipitator HV Supply Hacking

The message about finding out what the 16-pin chip actually is is new, as it was found out while the forum was moving.
Rex Allers

Rex Allers
Posts: 379
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Re: Precipitator HV Supply Hacking

Post by Rex Allers » Thu Aug 29, 2019 10:31 pm

I thought I would start a new thread about the circuits in the Chinese Precipitator Supplies. There has been at least one recent successful use of one of these in a small-chamber working fusor. There have also been reports of pretty fast failure in some of these.

There have also been speculations about these supplies for a few years with very little actual data.

A typical eBay listing title:
High Voltage Electrostatic Precipitator Power Supply 300W 400W 600W

To see what I could learn about these, I ordered two supplies:
one 30 KV 300 W
one 60 KV 600 W

I have spent some time reverse-engineering / hacking the circuits of these two supplies. I must admit that at this point I have not yet powered either one of these up. I just looked at what is on these boards and what they are doing.

After many hours, I think I have enough info and in barely good enough form to share. At this point only what I see, think I know, and know I don't know. I can't yet comment on suitibility, reliability or if they can actually come close to the rated voltage and wattage specs.

Hopefully my current level of information will help others to evaluate, speculate, fix problems, comment, or correct me.

At this point, I have created 4 zip files with the info I have created and some of my thoughts on what's going on in these circuits. I have placed these in my own web space for now. The files are several megabytes. If things are recieved well, and after any corrections or additions, I could put the info on the forum. Admins, let me know if that is desired.

--------------------
So here are links to download the 4 zips. You have to get the 4 full link paths correct or you will get not found, or no permission msgs...
In: www.xertech.net/precip-hv/ (just for info, you can't access the whole dir.)
total size: ~7.5M

1) ~940K
http://www.xertech.net/precip-hv/Precip-Gen-Info-01.zip

2) ~870K
http://www.xertech.net/precip-hv/precip-30-hack-01.zip

3) ~1.3M
http://www.xertech.net/precip-hv/precip-60-hack-01.zip

4) ~4.5M
http://www.xertech.net/precip-hv/precip-parts-01.zip

About the zip contents...

#1 Precip-Gen-Info-01.zip
Start with this. Contains a PDF doc that is an intro. It covers general information and also descriptions of circuit aspects that are mostly common to all the supplies. Also contains imgae files that are used in the pdf.

#2 precip-30-hack-01.zip
Details of the 30 KV supply circuits that I have found. '_notes' is a semi-random list of my comments about the circuits.

#3 precip-60-hack-01.zip
Details of the 60 KV supply circuits that I have found. '_60kV files_' is a brief description of the content of the files in the zip. '_notes' is a semi-random list of my comments about the circuits.

#4 precip-parts-01.zip
Is just a collection of data sheets for parts that are used in the supplies. If you want it, it should just save some time for finding them yourself.
------------------

Hope it helps. Let me know what you think. Especially let me know if you see anything I got wrong.
Rex Allers

Rex Allers
Posts: 379
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Re: Precipitator HV Supply Hacking

Post by Rex Allers » Thu Aug 29, 2019 10:34 pm

One open issue for the 60 KV supply is that it has a 16-pin chip with the markings ground off.

From looking at the circuit I'm pretty sure it must be a microcontroller of some kind but I haven't found a chip that seems to match the pin functions I have sussed out. On the elimination side, 16-pin micro chips seem to be prety rare, and 5V power on pin 6 may help ID too.

I'm hoping someone might have good microprocessor cred and recognise what this chip might be.

Here's an image I made. I think the analog pins are configured as inputs that way. The unmarked pins should be 5V-active digital in or out pins.

Let us know if anyone has a guess of what it is.
16-pin-chip.png
16-pin-chip.png (11.15 KiB) Viewed 1616 times
Rex Allers

Rex Allers
Posts: 379
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Re: Precipitator HV Supply Hacking

Post by Rex Allers » Thu Aug 29, 2019 10:41 pm

On the 60 KV supplies, I figured out what the 16-pin chip is. (The manufacturer grinds off the markings.)

Today I got my replacement for the 60 KV supply I had ordered that arrived with a broken HV output wire at the transformer. Looking at the top of the 16-pin chip, I saw some hints of numbers. With a tiny bit more sanding to take off the tops of swirls, a bit of water to blend and coat, and tipping for best light under a microscope -- I think I have it!

Saw something like this...

Code: Select all

-----------------
      15W408AS
 STC
      1824   ???
-----------------
I wasn't sure if this was just a partial or if I interpreted completely accurately, but after some web searching, it looks to be a complete part number and makes sense.

As I suspected, the chip is a microprocessor. It is made by the Chinese company STC. The part number (STC15W408AS) is a variation of the old, long lived 8051/8052 design and is in their 15 series.

Here's the web page for this part's info...
http://stcmicro.com/stc/stc15w408as.html

In the middle of that page are links to...
General Overview: STC15W408AS_Features.pdf
Data Sheet: STC15F2K32S2.pdf

As I suspected the 4-pin header near the chip could be used to load firmware onto the chip. There are tools for development if you follow links on the stcmicro pages. I haven't tried to install them and read the code from the chip. It may be locked or encrypted and even if clean, it would be a big task to disassemble / decompile.

So not sure if this new info helps much to more fully understand the supply operation, but at least we know what the chip is.

Here is an image of the chip pin-out from the datasheets that I have modified by adding my deduced functions (about the 60 KV supply) for each pin, outside the vertical lines that I also added.

Micro pins usage.png
Micro pins usage.png (62.78 KiB) Viewed 1616 times

Looking at this new blending of info, the pin-14 signal which I have called 'Switcher Osc Mon', could possibly be configured as an interrupt, occuring at the switcher Osc rate. (More guessing.)
Rex Allers

User avatar
Mark Rowley
Posts: 444
Joined: Sun Dec 29, 2002 5:20 am
Real name: Mark Rowley
Location: Sacramento California
Contact:

Re: Precipitator HV Supply Hacking

Post by Mark Rowley » Fri Aug 30, 2019 12:04 am

Hi Rex
I noticed the same as well. I have a rough copy of that post which I’ll put here:

———————————————————————————————-
18Aug19

Well, it happened...
The 60 just took a nose dive. I’m reluctant to say it was a flyback failure as I barely taxed the supply. I’m thinking it was a cheap overseas MOSFET or similar. I’ll dig into it over the next few days. Maybe beefing up the driver won’t be too difficult.

This how it went down. I just got set up for a run and slowly inched it up to 15kV with about 25mTorr of deuterium. There couldn’t have been more than a 1.5mA draw and suddenly no output. No spark, pop, pressure disturbance, notable current increase, etc. Both ready lights remain on and no overload relay “clicking” present.

The supply has never been pushed close to its limit (which was the plan for later today). The only mentionable run it had was last night (10 minutes at 3.5mA /25kV / 40mTorr / 346cpm neutrons with the small SNM14 tube.

—————————————————————————————-

29Aug19
So here’s the update since that last post:

Well that was an easy fix. At least for the 60kV unit. As expected, the MOSFET driver chip (IR2153) was toast.

Interestingly, I replaced the chip a couple days ago and it still didn’t work. In fact it began to blow the 5a fuse which it didn’t do before. So I once again checked everything and found all to be ok. Ended up being a bad replacement chip.

I’m going to try another “push to the limits test”, but this time with the original voltage control pot. As indicated by others, the added capacitance of the 8’ extension lead could’ve been the culprit. For safety, it’ll be a simple task to add an acrylic extension to the control.

Mark Rowley

Rex Allers
Posts: 379
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Re: Precipitator HV Supply Hacking

Post by Rex Allers » Fri Aug 30, 2019 1:30 am

Mark, interesting info that replacing the IR2153 chip fixed your 60 KV supply. Maybe your long extension wires for the V pot are involved. I'm thinking maybe not so much extra capacitance but acting as an antenna. Spikes or surges on the HV output into the fusor load might cause an EM wave that couples back into that extension wire.

Spikes on the input side of the chip probably could blow it if outside the normal 5 V signal range.

Just a speculation.
Rex Allers

User avatar
Mark Rowley
Posts: 444
Joined: Sun Dec 29, 2002 5:20 am
Real name: Mark Rowley
Location: Sacramento California
Contact:

Re: Precipitator HV Supply Hacking

Post by Mark Rowley » Fri Aug 30, 2019 2:23 am

Bob Reite's supply failure sounds somewhat similar and also employed the V extension (i think). If the chip fixes his then we at least know where a trouble spot is.

Mark Rowley

User avatar
Mark Rowley
Posts: 444
Joined: Sun Dec 29, 2002 5:20 am
Real name: Mark Rowley
Location: Sacramento California
Contact:

Re: Precipitator HV Supply Hacking

Post by Mark Rowley » Thu Sep 05, 2019 5:32 am

Today I fired up the 60kV unit after replacing the faulty driver IC and re-installing the original voltage control potentiometer. Rather than using the wired extension, I added a 7” lucite control rod (pictured). It’s still not conclusive the extension is what killed the IC, but best to try another run without it. After filling up the deuterium syringe reservoir, I initiated a non-stop run at approx 30kV, 3mA, 35mTorr deuterium pressure. This maintained for about 35 minutes without any issue. In the next day or two I’ll bump up the secondary voltage control to achieve a 10kV starting point and give it a go at 35kV with some higher current.

Mark Rowley
Attachments
7704BF13-736A-40BC-9791-A274E04EA39F.jpeg
7704BF13-736A-40BC-9791-A274E04EA39F.jpeg (130.49 KiB) Viewed 1358 times

Cristiano_Machado
Posts: 35
Joined: Fri Aug 24, 2018 1:03 pm
Real name: Cristiano Machado
Location: São Paulo - Brazil

Re: Precipitator HV Supply Hacking

Post by Cristiano_Machado » Tue Sep 10, 2019 1:34 am

Hello Mark and Rex,

I received my 30kV unit today and I was wondering what is the use of these white connectors? Can I use them to measure the voltage output? I saw your schematic and I understood it can not be used to control the voltage.


30kv hacked.jpg

Rex Allers
Posts: 379
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Re: Precipitator HV Supply Hacking

Post by Rex Allers » Tue Sep 10, 2019 2:21 am

Cristiano,

If you downloaded the zip files of my hacking efforts, you should be able to see from the layout circuits that those connectors are just in parallel with the red and green LEDs that are close to them, so not of much use unless you wanted to remote the LED signals. Not much information contained in them.

As I have mentioned earlier, none of the circuits in the 30 KV or 60 KV supplies actually monitor the output voltages. The secondaries of the transformers and their internal full wave rectifier diodes are floating with respect to the circuit boards.
Rex Allers

Post Reply