Mouse not detected and Windows 98SE Freeze issues

[jj mac]

Thanks MEB for that useful information in view of which I have decided to
replace the existing 4 modules (total 40MB) with 2 modules each of 32 MB
(total 64MB).
I would ,however, like to use 100nsec SDRAM as a replacement for the
existing EDO 70nsec DRAM which is now difficult to come by and more expensive
than SDRAM.
Would it be permissible to make that substitution?

You enquired if I had yet found the manufacturer of the motherboard. I
could not find any reference numbers or the maker’s name on the motherboard.
I scanned the computer with "ctbios" which recorded that the bios make was
AWARD Modular Bios v4.50G, Award id string 01/05/96-i430FX-2A59CF54C-00 AND
the Chipset is i430FX //Intel Triton-FX. There was some other information
which I could not make out as it is in German.
Regards and thanks again.

--
JJ MacA


"MEB" wrote:

> Oh, that means a maximum of 32megs PER SLOT on a four slot board... that
> also means no single module can exceed 32megs.
>
> --
> MEB
> a Peoples' counsel
> --
> _________
>
>
>
>

 

[Franc Zabkar]

On Thu, 11 Sep 2008 14:36:01 -0700, jj mac <jj
mac@discussions.microsoft.com> put finger to keyboard and composed:

>PIN 2 joined to PIN 3 NO flash noticed but Status of Pin 2 RXD changes to
>OK , back to Bad when the link was broken.

That means the Rx and Tx pins are both OK.

>Pin 4 joined to Pin6 No Flash noticed but status of DTR (pin 4) and DSR
>(pin6) alternates from + to - in tandem. When separated voltage between
>pins alternates + and - 10 V
>
>Pin 7 joined to Pin 8 Status of DTR,RTS&CTS changes in the following
>sequence:
> No Flash noticed DTR
>+ - + -
>
> RTS + - - +
>
> CTS + - - +

That means the port is working correctly. The test uses an output pin
to drive an input pin, thereby testing both pins.

It appears that your mouse or COM port are failing intermittently. I'd
switch your mouse to the other COM port. If you still have failures,
then I'd suspect your mouse.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Franc Zabkar]

On Sun, 14 Sep 2008 11:49:01 -0700, jj mac <jj
mac@discussions.microsoft.com> put finger to keyboard and composed:

>I could not find any reference numbers or the maker’s name on the motherboard.
>I scanned the computer with "ctbios" which recorded that the bios make was
>AWARD Modular Bios v4.50G, Award id string 01/05/96-i430FX-2A59CF54C-00 AND
>the Chipset is i430FX //Intel Triton-FX.

The "F5" in the "2A59CF54C" string tells you that the manufacturer was
Fugutech:
http://www.wimsbios.com/awardnumbers.jsp
http://www.wimsbios.com/biosupdates/fugutech.jsp#awardnumbers

I think this may be your board:
http://www.motherboards.org/files/manuals/89/M506.pdf

This is the M507 which appears very similar:
http://motherboards.mbarron.net/models/pcchips/m507.htm
http://th2chips.freeservers.com/m507/index.html

Here is a photo of the M507:
http://motherboards.mbarron.net/models/pcchips/m507p.jpg

Notice the two square "write back cache" chips at the bottom right.
These are fakes. You can probably pry the covers off the chips with a
knife and all you will see will be empty space. However, the Celem
Cache Test is telling you that you have genuine L2 cache, which makes
me wonder whether the same BIOS was used in a different board, or you
may have a COAST (Cache On A STick) module which plugs into the
connector to the left of the fakes.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[jj mac]

Thanks for interpreting the results of the Celem Cash Test. The tabulated
results appear to have got scrambled in transmission but you were able to
gleam enough info. therefrom.
As stated in my reply to MEB I have decided not to exceed 64 MB of RAM but I
would like to substitute 100nsec SDRAM for the existing 70nsec DRAM, if that
is permissible. However, since Windows Diagnostics, after running for 15
hours, gave the RAM a clean bill of health I may put off doing so
unless/until I can find what is causing the lockups. Since reinstalling
Windows I have not added any additional programs so it could hardly be caused
by the Computer running out of memory.
I have run SeaTools for Dos v1.10 PH . Both the short legacy test and the
long “Read Scan Test†(which I ran twice) passed. So the HDD appears to be
OK.
I am not too concerned about a possible fault in the COM1 port so long as I
can use a wireless mouse. Provided of course that any such fault is not part
of the more serious lockup problem.
Did “Com Test†results show any fault in COM1 Port.? My post of 9/11/2008
at 2.35PM PST refers.
Thanks again for your continuing help which is much appreciated.

--
JJ MacA


"Franc Zabkar" wrote:

> On Fri, 12 Sep 2008 08:40:01 -0700, jj mac <jj
> mac@discussions.microsoft.com> put finger to keyboard and composed:
>
> >I have now run the Celem Cache Test.
>
> >The peaks are at 2.8kb(3500) , 11kb – 260kb (800) & 400kb – 2580kb and
> >beyond (530) I am not sure how to interpret the CCT axis.
> >
> >Do these results mean that L1 cache (CPU) is 2.8KB, L2 Cache (Motherboard)
> >is 260KB & System RAM is 2580+ KB (the graph continues beyond 2580kb) ?
>
> L1 cache appears to be about 8KB, L2 cache = 256KB. The specs for a
> P120 suggest that this CPU should have an 8K data cache and 8K code
> cache, so I guess your results are OK. System RAM is 2580+ KB, as you
> say.
>
> The test works by "growing" a program until it fits into the amount of
> memory being tested. For example, if a program can fit completely into
> L1 cache, then it will run at L1 cache speeds. If it overflows L1
> cache, then its execution speed will start to degrade until, at double
> the cache size, the program loses all benefits associated with the L1
> cache and now runs at the speed of L2 cache. Similarly, when it grows
> to be twice the size of L2 cache, then it runs at the speed of main
> memory. This means you should see a plateau until around 256KB (+L1
> cache), then a gradual drop to the next plateau at 512KB (+L1 cache).
>
> >The question is can I install more than 64MB of RAM on a motherboard based
> >on Intel Triton i430 FX chipset with Pipeline Burst Cache.
>
> The results are showing that L2 cache is currently enabled and
> working. The Celem Cache Test won't tell you if L2 cache will be
> completely disabled after you increase the RAM beyond 64MB. You will
> need to run the test again afterwards to confirm or disprove this.
>
> - Franc Zabkar
> --
> Please remove one 'i' from my address when replying by email.
>

 

[Franc Zabkar]

On Sun, 14 Sep 2008 13:55:00 -0700, jj mac <jj
mac@discussions.microsoft.com> put finger to keyboard and composed:

>... I have decided not to exceed 64 MB of RAM but I
>would like to substitute 100nsec SDRAM for the existing 70nsec DRAM, if that
>is permissible.

I am confused about your RAM modules.

If this is your board ...

http://th2chips.freeservers.com/m507/m506_507.jpg

.... then you have 4 x 72-pin SIMM slots, not 168-pin SDRAM slots. SIMM
slots are usually populated two at a time, using identical pairs of
modules, ie 32 bits each for a total width of 64 bits.

See http://pinouts.ru/Memory/Simm72_pinout.shtml
and http://pinouts.ru/Memory/DimmSdram168Unbuf_pinout.shtml

You may also have 2 x fake "write back" cache chips (at bottom RHS)
plus two genuine ones above them.

Just FYI, here is a different PCChips motherboard with two fake cache
chips (and no genuine cache):
http://motherboards.mbarron.net/models/486pci/m919v1p.jpg

Notice that the second transistor in the CPU's Vcore power supply has
also been deleted (not that connecting two transistors directly in
parallel is a good idea anyway).

Notice also that this board has both SIMMs and DIMMs.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[MEB]

"Franc Zabkar" <fzabkar@iinternode.on.net> wrote in message
newsrsqc49a7o9ii875cagq0k3te1b0jd2117@4ax.com...
| On Sun, 14 Sep 2008 11:49:01 -0700, jj mac <jj
| mac@discussions.microsoft.com> put finger to keyboard and composed:
|
| >I could not find any reference numbers or the maker's name on the
motherboard.
| >I scanned the computer with "ctbios" which recorded that the bios make
was
| >AWARD Modular Bios v4.50G, Award id string 01/05/96-i430FX-2A59CF54C-00
AND
| >the Chipset is i430FX //Intel Triton-FX.
|
| The "F5" in the "2A59CF54C" string tells you that the manufacturer was
| Fugutech:
| http://www.wimsbios.com/awardnumbers.jsp
| http://www.wimsbios.com/biosupdates/fugutech.jsp#awardnumbers
|
| I think this may be your board:
| http://www.motherboards.org/files/manuals/89/M506.pdf
|
| This is the M507 which appears very similar:
| http://motherboards.mbarron.net/models/pcchips/m507.htm
| http://th2chips.freeservers.com/m507/index.html
|
| Here is a photo of the M507:
| http://motherboards.mbarron.net/models/pcchips/m507p.jpg
|
| Notice the two square "write back cache" chips at the bottom right.
| These are fakes. You can probably pry the covers off the chips with a
| knife and all you will see will be empty space. However, the Celem
| Cache Test is telling you that you have genuine L2 cache, which makes
| me wonder whether the same BIOS was used in a different board, or you
| may have a COAST (Cache On A STick) module which plugs into the
| connector to the left of the fakes.
|
| - Franc Zabkar
| --
| Please remove one 'i' from my address when replying by email.

Franc's pick for the board appears to reflect the BIOS string you found,
which shows the four 72 pin SIMM sockets per the manual and the page to
which he directed..

2 32meg 72 pin SDRAM may work, however, I never had any personal experience
with the FX board, so it MAY not properly address the SDRAM [it was the next
step in the memory type cycle]. Make sure that you turn off EDO and/or other
settings [just a check to make sure there are none] in the BIOS if it has
those settings and check the board jumpers BEFORE attempting a full boot...
If you already have a SDRAM module {any size} try it first before you
purchase any...

http://www.proprofs.com/mwiki/index.php/Understanding_RAM_Types:_DRAM_SDRAM_DIMM_SIMM_And_More

--
MEB
a Peoples' counsel
_ _
~~

 

[stalin12]

I am not too concerned about a possible fault in the COM1 port so long
as I
can use a wireless mouse. Provided of course that any such fault is not
part
of the more serious lockup problem.
Did “Com Test†results show any fault in COM1 Port.? My post of
9/11/2008
at 2.35PM PST refers.
Thanks again for your continuing help which is much appreciated.




--
stalin12

 

[jj mac]

Franc
I am pleased to have your confirmation that the ComTest passed and to be
able to report that my serial mouse has started to work again. Before
carrying out the ComTest I had removed the COM1 port from the Computer
unplugging the connector from the motherboard. I did so to check that there
was no obvious defect in the ribbon cable and the pins in the port socket.
Having replaced the port I carried out the COMTESTS. I then connected the
serial mouse without disconnecting the wireless mouse which I had been using.
I was surprised not to get a message to tell me that Windows did not detect
a mouse. The mouse pointer appeared on screen but the mouse, however, could
not move or control it. I concluded that the fault was still there but after
receiving your confirmation that the COM 1 port was OK I reconnected the
serial mouse but this time I disconnected the wireless mouse. I was
pleasantly surprised to find the serial mouse worked. Furthermore the
Computer lockups have also greatly reduced. It would appear that the
connection of the ribbon cable to the motherboard had developed a fault which
was corrected by the simple action of removing and replacing the ribbon plug.
It is now obvious that the wireless Mouse connected to the USB2 port is the
main culprit causing Lockouts. The USB2 5Port Adapter Card which I installed
a few months ago might also be implicated. I hope, however, to be able to
retain the USB ports. Would moving the card to a different slot or
increasing the RAM help?

I can confirm that my motherboard is a Fugutech M506 board with 4no.72 pin
Simm slots at present containing 2 no. 4M modules of EDO Dram & 2 no. 16M
ditto. If I had done my homework properly I should have known that SDRAM
was not supported on this board.
I am now thinking of replacing the two 4M modules with two 32M modules
(total 64M). In the first instance I will leave the existing two 16M modules
(total 32M)
I will remove them if there is a problem. If they can remain I will have a
total of 96M DRAM, a big improvement on 40M.

With the mouse now working and the lockouts occurring less frequently the
problems with this computer are now largely resolved. Be assured that your
patience and help with the running of test programs and explaining the
results is very much appreciated.

Regards
--
JJ MacA


"Franc Zabkar" wrote:

> On Sun, 14 Sep 2008 13:55:00 -0700, jj mac <jj
> mac@discussions.microsoft.com> put finger to keyboard and composed:
>
> >... I have decided not to exceed 64 MB of RAM but I
> >would like to substitute 100nsec SDRAM for the existing 70nsec DRAM, if that
> >is permissible.
>
> I am confused about your RAM modules.
>
> If this is your board ...
>
> http://th2chips.freeservers.com/m507/m506_507.jpg
>
> .... then you have 4 x 72-pin SIMM slots, not 168-pin SDRAM slots. SIMM
> slots are usually populated two at a time, using identical pairs of
> modules, ie 32 bits each for a total width of 64 bits.
>
> See http://pinouts.ru/Memory/Simm72_pinout.shtml
> and http://pinouts.ru/Memory/DimmSdram168Unbuf_pinout.shtml
>
> You may also have 2 x fake "write back" cache chips (at bottom RHS)
> plus two genuine ones above them.
>
> Just FYI, here is a different PCChips motherboard with two fake cache
> chips (and no genuine cache):
> http://motherboards.mbarron.net/models/486pci/m919v1p.jpg
>
> Notice that the second transistor in the CPU's Vcore power supply has
> also been deleted (not that connecting two transistors directly in
> parallel is a good idea anyway).
>
> Notice also that this board has both SIMMs and DIMMs.
>
> - Franc Zabkar
> --
> Please remove one 'i' from my address when replying by email.
>

 

[MEB]

Okay, so you're both jj mac and this entity apparently.

I see you have confirmed the Comm port issue to Franc, AND you realized the
SDRAM [which some have misconstrued in various articles and web pages and
which you apparently though was the same pin configuration] will not work
with your system {your's supports 72 pin FPM/EDO}, as you have now confirmed
the model for the board. Your intended memory installation [in your other
post] should now work. It is always best to use ALL EDO or FPM memory,
mixing can cause issues.

You also note that the issues likely occurred AFTER installation of the USB
adapter. You have also noted that two mouse [wireless and serial] installed
will not work. That is not an uncommon error/issue. The mouse holds a
special place and reliance in Windows/GUI systems...

It may be: you have corrupted the system with the installation/drivers for
the USB device. Remove ALL drivers and references to the device and USB, and
either re-install the driver OR try the Maximus Decim Universal driver AFTER
the complete removal of the other [Hub and subs]. Make sure to re-start
after each phase [removal and installation].

AFTER either re-installing the old/original driver or PREFERABLY the
Maximus Decim, try pointing the install process to the install folder for
the wireless mouse so it can find the INF for *only* the wireless {it is
likely needed} while retaining the Universal driver. Remember you may need
to also remove the old serial mouse entry, if necessary via Safe Mode [ghost
entry, check for other issues while reviewing Device Manager and Add/Remove
in both Normal and Safe Modes].

--
MEB
a Peoples' counsel
_ _
~~
"stalin12" <stalin12.315ada1@news.win98banter.com> wrote in message
news:stalin12.315ada1@news.win98banter.com...
|
| I am not too concerned about a possible fault in the COM1 port so long
| as I
| can use a wireless mouse. Provided of course that any such fault is not
| part
| of the more serious lockup problem.
| Did â?oCom Testâ? results show any fault in COM1 Port.? My post of
| 9/11/2008
| at 2.35PM PST refers.
| Thanks again for your continuing help which is much appreciated.
|
|
|
|
| --
| stalin12

 

[Franc Zabkar]

On Tue, 16 Sep 2008 19:08:37 +0100, stalin12
<stalin12.315ada1@news.win98banter.com> put finger to keyboard and
composed:

>
>I am not too concerned about a possible fault in the COM1 port so long
>as I
>can use a wireless mouse. Provided of course that any such fault is not
>part
>of the more serious lockup problem.
>Did “Com Test� results show any fault in COM1 Port.? My post of
>9/11/2008
>at 2.35PM PST refers.
>Thanks again for your continuing help which is much appreciated.

The test showed that your COM port was working at that time. It may
still have an intermittent fault, though. You really need to switch
the mouse to the other COM port and test it for a few days.

If your motherboard is the same as the one in the photos I posted,
then it uses a UMC UM8663 Super-IO chip which incorporates a floppy
drive controller, two IDE hard drive ports, two COM ports, and a
parallel printer port. I believe the nearby UM8667 (?) chip has the
RS232 drivers for the two COM ports.

If the mouse works on COM2 but not on COM1, then I would suspect the
UM8667 chip or the internal cable between the motherboard header and
the DB9 connector on the slot bracket. If the mouse fails on both
ports, then I would suspect the mouse itself, or the UM8663 chip.
However, if the Super-IO chip is faulty on both COM ports, then I
would also expect faults in the chip's other functions, ie floppy,
parallel printer, hard drive. Since this does not appear to be the
case, then the UM8663 is unlikely to be faulty.

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[Franc Zabkar]

On Tue, 16 Sep 2008 12:43:01 -0700, jj mac <jj
mac@discussions.microsoft.com> put finger to keyboard and composed:

>Computer lockups have also greatly reduced. It would appear that the
>connection of the ribbon cable to the motherboard had developed a fault which
>was corrected by the simple action of removing and replacing the ribbon plug.
>It is now obvious that the wireless Mouse connected to the USB2 port is the
>main culprit causing Lockouts. The USB2 5Port Adapter Card which I installed
>a few months ago might also be implicated. I hope, however, to be able to
>retain the USB ports. Would moving the card to a different slot or
>increasing the RAM help?

I serviced my old PCs not too long ago. I was seeing GPFs a bit too
often, but after cleaning the dust off the heatsinks and fans
(motherboard and power supply), and after removing and reseating all
cables, connectors, memory modules and cards, the systems stabilised.
I also sprayed all connectors and slots with electronic cleaning
solvent, and gently scrubbed the card fingers with a soft pencil
eraser. It may pay you to go through your machine as well.

>I can confirm that my motherboard is a Fugutech M506 board with 4no.72 pin
>Simm slots at present containing 2 no. 4M modules of EDO Dram & 2 no. 16M
>ditto. If I had done my homework properly I should have known that SDRAM
>was not supported on this board.
>I am now thinking of replacing the two 4M modules with two 32M modules
>(total 64M). In the first instance I will leave the existing two 16M modules
>(total 32M)

AFAIK, 32MB modules would be electrically "double sided". Some
chipsets only see half the capacity of such modules. I don't know
whether yours is one of them.

>With the mouse now working and the lockouts occurring less frequently the
>problems with this computer are now largely resolved.

You should aim for zero lockups. Something is still wrong. I wouldn't
blame your USB card just yet, unless it has a VIA chipset, in which
case you're not the only one reporting problems.

See http://www.usbman.com/Guides/VIA Tips and Tricks.htm

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[jj mac]

When sifting through a pile of papers today I came across a Users Manual for
the Triton motherboard which I did not know I had. The manual states
thatâ€The mainboard lets you add up to 128 MB (32 MB in each socket) of system
memory via four SIMM sockets on the mainboard.†It also states that “all
SIMMs must be faster than 70ns all banks can use either 1-sided or 2-sided
SIMMs all banks are auto bank DRAM Type: Fast Page Mode or Extended Data
Output (EDO)â€
It appears that DRAM is now obsolete and virtually impossible to get. There
is a lot of EDO SIMMS for sale on EBAY but it is all second-hand and rated at
60ns which I take to be faster than 70ns. I am tempted to make a bid for four
x 32MB of 72pin EDO SIMMS currently on offer (closing date Tuesday 23 Sept)
for opening bids in excess of £4.99. If EDO SIMMS is EDO DRAM by another
name and if 60ns is faster than 70ns (which it must be) I will go ahead. If
only two of the four modules work it could still be worth it.
You guessed it. My USB2 card has a VIA chipset. It seems to be working
perfectly.
The USB wireless mouse is causing lockups but now that I have got the serial
mouse to work I do not need it. I have BookMarked USB.Com which could be a
useful reference if any problems with the card arise. Thank you.
I am taking your advice about carrying out a bit of spring cleaning inside
the Computer.
Thanks again for all your help and sound advice.
Regards

--
JJ MacA


"Franc Zabkar" wrote:

> On Tue, 16 Sep 2008 12:43:01 -0700, jj mac <jj
> mac@discussions.microsoft.com> put finger to keyboard and composed:
>
> >Computer lockups have also greatly reduced. It would appear that the
> >connection of the ribbon cable to the motherboard had developed a fault which
> >was corrected by the simple action of removing and replacing the ribbon plug.
> >It is now obvious that the wireless Mouse connected to the USB2 port is the
> >main culprit causing Lockouts. The USB2 5Port Adapter Card which I installed
> >a few months ago might also be implicated. I hope, however, to be able to
> >retain the USB ports. Would moving the card to a different slot or
> >increasing the RAM help?
>
> I serviced my old PCs not too long ago. I was seeing GPFs a bit too
> often, but after cleaning the dust off the heatsinks and fans
> (motherboard and power supply), and after removing and reseating all
> cables, connectors, memory modules and cards, the systems stabilised.
> I also sprayed all connectors and slots with electronic cleaning
> solvent, and gently scrubbed the card fingers with a soft pencil
> eraser. It may pay you to go through your machine as well.
>
> >I can confirm that my motherboard is a Fugutech M506 board with 4no.72 pin
> >Simm slots at present containing 2 no. 4M modules of EDO Dram & 2 no. 16M
> >ditto. If I had done my homework properly I should have known that SDRAM
> >was not supported on this board.
> >I am now thinking of replacing the two 4M modules with two 32M modules
> >(total 64M). In the first instance I will leave the existing two 16M modules
> >(total 32M)
>
> AFAIK, 32MB modules would be electrically "double sided". Some
> chipsets only see half the capacity of such modules. I don't know
> whether yours is one of them.
>
> >With the mouse now working and the lockouts occurring less frequently the
> >problems with this computer are now largely resolved.
>
> You should aim for zero lockups. Something is still wrong. I wouldn't
> blame your USB card just yet, unless it has a VIA chipset, in which
> case you're not the only one reporting problems.
>
> See http://www.usbman.com/Guides/VIA Tips and Tricks.htm
>
> - Franc Zabkar
> --
> Please remove one 'i' from my address when replying by email.
>

 

[Franc Zabkar]

On Thu, 18 Sep 2008 05:35:01 -0700, jj mac <jj
mac@discussions.microsoft.com> put finger to keyboard and composed:

>If EDO SIMMS is EDO DRAM by another
>name and if 60ns is faster than 70ns ...

Yes, you are correct on both counts.

Just out of curiosity, does your board have two fake "write back
cache" chips as in the photos I posted?

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.

 

[jj mac]

Franc, unfortunately the chips are hidden under the hard drive cage. With
the aid of a mirror I can get a glimpse of the inner chips in the bottom and
upper row. There are identification markings on the upper chip whereas I
cannot see any markings on the bottom chip. This is contrary to the photo of
the M506 board where markings are visible on the bottom row. On that
evidence I would guess that the bottom row could be fake but to be honest,
even if I were to remove the hard drives to get access and a clear view of
these chips, I doubt if I could say with any certainty which ones, if any,
were fake. Both chips that I could see appear to be identical with an array
of input/output pins on all four sides. They look like the real thing.
I am pleased to have your advice that EDO SIMMS and EDO DRAM are the same.
Incidentally the Users Manual which I found for the motherboard does not
refer to Fugutech or M506 at all, but I am sure that is what it is. It is
described as a "Triton mainboard with on board PCI IDE and Super Multi-I/O. A
high-performance mainboard based on the advanced Pentium microprocessor,the
PCI Local Bus and the intel TRITON chipset"
Kind Regards--
JJ MacA


"Franc Zabkar" wrote:

> On Thu, 18 Sep 2008 05:35:01 -0700, jj mac <jj
> mac@discussions.microsoft.com> put finger to keyboard and composed:
>
> >If EDO SIMMS is EDO DRAM by another
> >name and if 60ns is faster than 70ns ...
>
> Yes, you are correct on both counts.
>
> Just out of curiosity, does your board have two fake "write back
> cache" chips as in the photos I posted?
>
> - Franc Zabkar
> --
> Please remove one 'i' from my address when replying by email.
>