Problems burning Kickstart

[Pat the Cat]

Originally Posted by idrougge

Do you really have 240 V in the UK?

Depends what the local substation is set to... More so in suburbs and rural districts.

Usually reads about 238 on a meter, in such locations.

The EU directive was carefully fudged, 230V +- 5% covers 220-240 V anyway, so nothing technically changed except the paperwork.

[xArtx]

That cheap eBay eraser is designed for a 120 Volt supply,
so I wouldn’t be surprised if there are issues if your mains is 240V.

[xArtx]

You get what you pay for lol



The blue wire is cut on purpose to replace the mains cable,
but the brown wire from the tube was disconnected when it arrived in the post.
Perhaps from part of the drawer sliding over the wiring under the PCA, so I moved all I can to the top side.

Of course it does work with a 240 Volt supply. I don’t know what that will do with regard to the life of the tube,
but at the very least, you couldn’t meaningfully compare your exposure time with someone with a different mains supply.

[nogginthenog]

Originally Posted by xArtx

You get what you pay for lol

Of course it does work with a 240 Volt supply. I don’t know what that will do with regard to the life of the tube,
but at the very least, you couldn’t meaningfully compare your exposure time with someone with a different mains supply.

Other erasers cost $$$. Everybody is selling this one!

Not seeing a voltage regulator chip on that PCB. Hopefully there is a zener in there... Someone should send one to BigClive for a teardown lol.

p.s. shouldn't need to say this but UV is dangerous for your eyes!

[xArtx]

To find out what? It definitely does nothing to sense the mains supply voltage.

[xArtx]

I had another look at it today. Here’s a 15 second video winding the supply voltage
between 100V and 240V a couple of times, and then lastly from 110V down to 0V.

[ Show youtube player ]

I’m familiar with the drive circuit it’s a ZVS driver with a full wave rectifier made of
the four diodes to the left of the board. The driver actively drives it’s output in both
directions to produce a new (closer to triangle) wave at about 25kHz which would
vary between units, and it’s covered by the resonant frequency of the toroid.

Once the driver is oscillating, and the lamp has struck, the frequency of the driver
doesn’t change with varying supply voltage but the amplitude does.

[idrougge]

Is the xsurf_atapi.device romable and if so, is it autobootable?

[Irl]

Originally Posted by idrougge

Is the xsurf_atapi.device romable and if so, is it autobootable?

Hope this helps.. http://eab.abime.net/showthread.php?t=69583 http://eab.abime.net/showthread.php?t=50061

[idrougge]

I hope it's not at that level of complexity.

[Irl]

Originally Posted by idrougge

I hope it's not at that level of complexity.

I may be misunderstanding your goal then. If you are hoping to insert the device driver into the ROM, you will need to consider many factors. Is it 'pure' or not, will it fit in ROM or need to move something to workbench to make it fit, etc. I have not done any ROM modifications, myself, which is why I only linked to other discussions.

[idrougge]

My reasoning is that since the Buddha (from the same manufacturer as Xsurf) has its device driver in ROM, hopefully XsurfIDE (which is an old-style BindDrivers file) can be merged into a ROM. However, I'm not enough of a guru to know if autobooting requires more than just a device driver being part of the ROM, or if it is even "bound" if it is.

[kneehighspy]

i have been trying to get some kickstarts burned for one of my a1000s, but really don't know or understand why they arent working.

i purchased a 2.0 and a 3.1 from cloanto and they are 42 pin eproms so i am sure they are 27C160 and i have them in my automatic kickstart switcher (with pins 1 and 42 outside of socket) i purchased from keir on amibay and they work great.

i read them with my minipro tl866A with the adapter from tuxbar on amibay, and they read fine.

they dont work after straight reading, but once i do the bit flipping, they will run fine in winuae under amiga 1000 emu loading them into the emulator.

take those same working roms, put them in the multikickstart adapter in one of my 1000s (after burning them to 27C400), they don't work at all.

i take those same burned 27c400, read them with my eprom burner, load them into winuae and they work fine.

tried all sizes from 256k to 1024k and nothing works, any ideas? clueless here

thanks guys!

//kneehighspy

[kgc210]

If they work in WinUAE but not on a real Amiga sounds like they need byte swapping.

When you build them in Remus there is an option to byte swap, or your software should have the option to do so.

[Vot]

Originally Posted by kneehighspy

i have been trying to get some kickstarts burned for one of my a1000s, but really don't know or understand why they arent working.



i purchased a 2.0 and a 3.1 from cloanto and they are 42 pin eproms so i am sure they are 27C160 and i have them in my automatic kickstart switcher (with pins 1 and 42 outside of socket) i purchased from keir on amibay and they work great.



i read them with my minipro tl866A with the adapter from tuxbar on amibay, and they read fine.



they dont work after straight reading, but once i do the bit flipping, they will run fine in winuae under amiga 1000 emu loading them into the emulator.



take those same working roms, put them in the multikickstart adapter in one of my 1000s (after burning them to 27C400), they don't work at all.



i take those same burned 27c400, read them with my eprom burner, load them into winuae and they work fine.



tried all sizes from 256k to 1024k and nothing works, any ideas? clueless here



thanks guys!



//kneehighspy

When the chips are installed in the a1000 are the two extra address pins floating or tied to vcc or gnd

[kneehighspy]

Originally Posted by Vot

When the chips are installed in the a1000 are the two extra address pins floating or tied to vcc or gnd

they are floating, pins 1 and 42 are outside the multikickstart adapter headers. these are roms i purchased from ebay, they say they are official cloanto roms with serial #s, but i am sure they are just copies of 2.0 and 3.1 respectively burned on 27c160 eproms.

when i read these ebay roms and save them with my burner software, i can load them into winhex, byte flip them and they work fine in winuae.

then i try burning these same rom files (byte flipped, working in winua) onto 27c400 eproms, they do not work in my a1000 multikickstart adapter.

i am gonna order some 27c160s and see what happens, but i have read others with working 27c400s in 1000s with multikickstart adapters.

thanks for the help.

//kneehighspy

[r.cade]

I have an Amiga 3000, board version 8.9. It has original 2.0 ROMs in the standard U180/U181 (no ROM tower).

I burned new 3.1 ROMs to brand new AM27C400-120 chips (doubled up). I went through the word split and byte swap routine correctly, and I even reversed the process with the files and made a working ROM for WinUAE, no problem. However, when inserted into the A3000, red screen. I adjusted the ROM speed jumpers, no change. Red screen.

To test that I did it properly, I read in my 2.0 ROMs (with exact same settings, as 27C400) and then just wrote it directly back out to the EPROMs. Still, red screen. I took these ROMs and reversed the byte swap and word split and they work in WinUAE- the files and burning process seem to be fine. They verify 100%.

What is the problem? Will AM27C400's not work in a stock A3000?

[jbilander]

Originally Posted by r.cade

What is the problem? Will AM27C400's not work in a stock A3000?

Check this post:

http://eab.abime.net/showpost.php?p=1289346&postcount=9

[r.cade]

Originally Posted by jbilander

Check this post:

http://eab.abime.net/showpost.php?p=1289346&postcount=9

Yes, these are replacing the normal ROMs.

Anyway, I bought the AMD 27C400 95ns version (faster than the 120ns I had before) and they at least boot (sometimes) but are still unstable, at any ROM speed settings.

So, my A3000 needs ever faster EPROMs, or else these AMD's from China are all junk. I'm willing to believe either one is true. They all burn and verify fine, just don't work reliably in a real A3000.

EDIT: I tried the 95ms ROMs in my other A3000 (9.x board) and they work fine. The 120ms still do not, though. So I guess the board or board revision is also sensitive to the EPROM speed.

[quantum8]

First time attempting to burn and I'm getting stuck at the "Perform a swap BYTE operation on the buffer contents" step.
I've got a tl866 and using xgpro software with the 27C400 device loaded. How do I perform a byte swap? The load options say binary or intel. And there's a section for load mode where I can select Normal, load low byte of word, high word of dword, etc but I don't think that's it.

Does anyone have an idea or should I try using other software?

[demolition]

Originally Posted by quantum8

First time attempting to burn and I'm getting stuck at the "Perform a swap BYTE operation on the buffer contents" step.
I've got a tl866 and using xgpro software with the 27C400 device loaded. How do I perform a byte swap?

I don't know that particular software but I know that dd (*nix tool) can do it for you like this:

dd if=original.bin of=swapped.bin conv=swap

In Windows, you can use Hex Workshop and do it like this:
Tools->Operations->Byte Flip
and then 'treat data as: 16-bit'