I never said they are bad because they were stolen designs. Just that they are stolen designs.
.
I would be very leery of the particular one we've been talking about though.
The missing jumper/switch block is the one used to set FSB.
That it is missing suggests that function has been moved into the BIOS. (Or the FSB is fixed.)
So now you have a little unknown company selling boards with custom/modified BIOSs.
I would not count on a lot of BIOS bug fixes or updates from a company like this, nor would I expect them to maintain BIOS downloads for very long once the boards go out of production.

PCBONEZ wrote:

I have several decades of experience following the motherboard industry and I know how it works. It's self evident that you don' […]
Show full quote

I have several decades of experience following the motherboard industry and I know how it works.
It's self evident that you don't know how it works when you ask for motherboard schematics.
Those get released for reference designs but never ever for completed designs used in production boards.
.
If it makes you feel better you can attach "probably", "most likely", "historically" or "standard practice" where appropriate to every thing I said.
Because, whether you like it or not, what I said -IS- what is most likely going on here.

Ofc i know that the schematics aren't released for production boards, and i don't think that you have any of that. That's why i don't understand why you are so sure that Gigabyte exact design (layout/components/everything else) was exactly cloned in everything for this new batch of boards by ITT (we aren't even sure that Hsing was subcontracted to produce boards for IIT), and you even are going so far as declare openly that they stole Gigabyte's IP 😮.

Most likely =! It actually happened.

PCBONEZ wrote:

Gigabyte has never licensed a mobo design to someone else - ever in their history.
Hsing Tech has been stealing designs for various things for decades.

Oh, so now you will tell me that you actually worked with Gigabyte and you are aware of all their deals, like what stuff they actually license and for what stuff they actually bought a license to use it?
About the rest, is possible. But again, no real proofs besides rumors and forum assumptions, except for the cache/chipset thing that happened 20+/- years ago.

In any case, Gigabyte never presented a public case against any other chinese manufacturer for IP steal, including Hsing Tech, so...

PD: I don't want to divert this thread more. OP already took his decision and he will buy one of these boards. Maybe he can post a review about it later, so we can judge if the board is worth or no. That would be a real solid proof, and no the f*cking assumptions that we are making here.

hyoenmadan wrote:

Ofc i know that the schematics aren't released for production boards, and i don't think that you have any of that. That's why i don't understand why you are so sure that Gigabyte exact design (layout/components/everything else) was exactly cloned in everything for this new batch of boards by ITT (we aren't even sure that Hsing was subcontracted to produce boards for IIT), and you even are going so far as declare openly that they stole Gigabyte's IP 😮.
.

I didn't say anything about Gigabyte's IP. Unless you mean something else by IP.

It's quite clear it's Gigabyte's design. Even the solder traces are the same.

If you read what I wrote earlier what I said was I find it doubtful that such a small company could afford a motherboard fab.

Hsing Tech is the logical choice for their OEM because Hsing Tech has never given a rat's patootie about licensing anything.
Sure, they -could have- done it themselves or OEM'ed from some small unknown fab.
But that would give me even less confidence about their quality.
Hsing Tech might be scum when it comes to principals and ethics but at least they have experience building motherboards.

I don't think Gigabyte gives a damn about designs so old they file them under "Legacy".
.

hyoenmadan wrote:

PD: I don't want to divert this thread more. OP already took his decision and he will buy one of these boards. Maybe he can post a review about it later, so we can judge if the board is worth or no. That would be a real solid proof, and no the f*cking assumptions that we are making here.

The topic is "ISA slot motherboards from AliExpress" so this discussion is not a diversion.

I too would like to see a review. But not so much I'd buy one to do it.
Preferably a review by someone that uses an O'Scope when they review.

I'd like to see a review also tbh 😀

I also found some other motherboard and there reads it would be a new one. I will probably order some motherboard and see what I get. What I would like to find are new AT computer cases, but it seems almost impossible.

http://www.aliexpress.com/item/Special-new-58 … archweb201560_9

markot wrote:

I also found some other motherboard and there reads it would be a new one. I will probably order some motherboard and see what I get. What I would like to find are new AT computer cases, but it seems almost impossible.

http://www.aliexpress.com/item/Special-new-58 … archweb201560_9

Given the ad title says "with CPU of memory spark cutting computer gongs" I wouldn't count getting what you think you're getting.
.

PCBONEZ wrote:

markot wrote:

I also found some other motherboard and there reads it would be a new one. I will probably order some motherboard and see what I get. What I would like to find are new AT computer cases, but it seems almost impossible.

http://www.aliexpress.com/item/Special-new-58 … archweb201560_9

Given the ad title says "with CPU of memory spark cutting computer gongs" I wouldn't count getting what you think you're getting.
.

This is probably true. I sent some questions about this motherboard to the seller and asked for specifications and manuals. The answer I got was "no,thanks".

Hi,
I had spent a lot of time and money hunting ISA+audio stuff in the past. I have maintaned modern ISA-enabled MB list on my personal site:
http://rayer.g6.cz/hardware/mbisa.htm
I would like to clarify one thing - it doesn't matter if the ISA bridgde is manufactured by Winbond, ITE, Fintek or someone else... What matters is how it is connected to host system. If it is PCI2ISA bridge and the southbridge is ICH6 or newer you have no chance run ISA DMA because intel removed some important legacy lines from the ICH so blame intel and not the bridge manufacturer. If you have ICH5 and older it still may not be enough to support ISA DMA because it depends if all necessary lines was routed properly (and if it is properly configured by BIOS). I have one MB with i845+ICH4 with Winbond PCI2ISA bridge and it works fine with my SB AWE64: http://rayer.g6.cz/hardware/azxab100.htm
On some minority MBs is used a different host connection - LPC2ISA bridge. I received a positive report from one guy that have such MB (intel G41 with Fintek F85226FG bridge) and after some tweaking of bridge config he got ISA DMA working. I'd like to hear about another successfull story on similar or newer MB.

^^ I've tried various PnP soundcards, including AWE32 and AWE64, and they all worked fine on a BCM BC875PLG (i875-based). The only non-PnP card I tried was a SBPro 2 CT1600, and that locked-up during booting.

RayeR wrote:

On some minority MBs is used a different host connection - LPC2ISA bridge. I received a positive report from one guy that have such MB (intel G41 with Fintek F85226FG bridge) and after some tweaking of bridge config he got ISA DMA working. I'd like to hear about another successfull story on similar or newer MB.

That's really encouraging. I wish that someone skilled would create an LPC to ISA bridge/backplane that would connect to TPM headers found in so many motherboards. I would buy that in an instant.

FuzzyLogic wrote:

That's really encouraging. I wish that someone skilled would create an LPC to ISA bridge/backplane that would connect to TPM headers found in so many motherboards. I would buy that in an instant.

Yes, I belive it should work. I also have TPM (LPC) header on my current Gigabyte MB. I have a plan to start with something more simple - make a proto board with 2nd superIO chip to enable floppy support (that I'm missing) or additional COM port (easy to test). It will also need some chipset configuration - as BIOS wouldn't do it for me I will need to write some DOS initialization program. I already searched for a similar project but nothing found. I wonder how 8-bit community is producing a bunch of HW tweaks for they loved machines like modern CF/SD storage for C64 and so on but nothing for PC HW 🙁

RayeR wrote:

Hi, I had spent a lot of time and money hunting ISA+audio stuff in the past. I have maintaned modern ISA-enabled MB list on my p […]
Show full quote

Hi,
I had spent a lot of time and money hunting ISA+audio stuff in the past. I have maintaned modern ISA-enabled MB list on my personal site:
http://rayer.g6.cz/hardware/mbisa.htm
I would like to clarify one thing - it doesn't matter if the ISA bridgde is manufactured by Winbond, ITE, Fintek or someone else... What matters is how it is connected to host system. If it is PCI2ISA bridge and the southbridge is ICH6 or newer you have no chance run ISA DMA because intel removed some important legacy lines from the ICH so blame intel and not the bridge manufacturer. If you have ICH5 and older it still may not be enough to support ISA DMA because it depends if all necessary lines was routed properly (and if it is properly configured by BIOS). I have one MB with i845+ICH4 with Winbond PCI2ISA bridge and it works fine with my SB AWE64: http://rayer.g6.cz/hardware/azxab100.htm
On some minority MBs is used a different host connection - LPC2ISA bridge. I received a positive report from one guy that have such MB (intel G41 with Fintek F85226FG bridge) and after some tweaking of bridge config he got ISA DMA working. I'd like to hear about another successfull story on similar or newer MB.

Wow after the heated debate of if someone copied another company's MB but still no one has bought and tested the "IMIG41-2ISA" which could be the most compatible from the list that might work with the SB ISA Sound Card? Curious if anyone ended up buying that motherboard and testing it...
🙄

Stumbled upon an interesting adapter...

http://www.phantom.sannata.ru/forum/index.php … =372&a=do_print

http://rd.foto.radikal.ru/0709/4a/452d6ffb815e.jpg

95DosBox wrote:

Wow after the heated debate of if someone copied another company's MB but still no one has bought and tested the "IMIG41-2ISA" w […]
Show full quote

RayeR wrote:

Hi, I had spent a lot of time and money hunting ISA+audio stuff in the past. I have maintaned modern ISA-enabled MB list on my p […]
Show full quote

Hi,
I had spent a lot of time and money hunting ISA+audio stuff in the past. I have maintaned modern ISA-enabled MB list on my personal site:
http://rayer.g6.cz/hardware/mbisa.htm
I would like to clarify one thing - it doesn't matter if the ISA bridgde is manufactured by Winbond, ITE, Fintek or someone else... What matters is how it is connected to host system. If it is PCI2ISA bridge and the southbridge is ICH6 or newer you have no chance run ISA DMA because intel removed some important legacy lines from the ICH so blame intel and not the bridge manufacturer. If you have ICH5 and older it still may not be enough to support ISA DMA because it depends if all necessary lines was routed properly (and if it is properly configured by BIOS). I have one MB with i845+ICH4 with Winbond PCI2ISA bridge and it works fine with my SB AWE64: http://rayer.g6.cz/hardware/azxab100.htm
On some minority MBs is used a different host connection - LPC2ISA bridge. I received a positive report from one guy that have such MB (intel G41 with Fintek F85226FG bridge) and after some tweaking of bridge config he got ISA DMA working. I'd like to hear about another successfull story on similar or newer MB.

Wow after the heated debate of if someone copied another company's MB but still no one has bought and tested the "IMIG41-2ISA" which could be the most compatible from the list that might work with the SB ISA Sound Card? Curious if anyone ended up buying that motherboard and testing it...
🙄

Stumbled upon an interesting adapter...

http://www.phantom.sannata.ru/forum/index.php … =372&a=do_print

http://rd.foto.radikal.ru/0709/4a/452d6ffb815e.jpg

Sadly IEI IMBA-G412ISA you mentioned doesn't have functional ISA just like Intel said. I had one, though I'm currently running it for a different purpose.
For DOS purposes, at best you can make good use of your sound card's FM synth for some old games and for music trackers, in a way better than PCI sound cards. (AFAICT when Intel removed PC-PCI related stuffs they also broke everything PCI sound cards relied on to get functional DOS audio, except Aureal which seemed to be a different approach)

Again, I'm yet to hear of any practical ideas on making the ISA slots on boards of ICH6 or later actually useful...

EDIT: Should additionally mention something that's been bugging me about the IMBA-G412ISA board I'm running.
1. Its BIOS contains both IRQ and DMA assignment settings (Available for PCI/PnP or Reserved for Legacy ISA), despite the board doesn't appear to have any DMA capabilities (hence the DMA switches are mostly useless).
2. The BIOS changelog contained a line about leaving IRQ11 Reserved for ISA PnP, and an earlier AMI BIOS revision mentioned about something called Bear Card.

4. IRQ11 reserve for ISA PnP card.

############################################################################# # Action : SA28MR12 (base on SA28MR11) # Date : 0 […]
Show full quote

#############################################################################
# Action : SA28MR12 (base on SA28MR11)
# Date : 06/26/2012
# CheckSum : 2E45h
# Sign_On_Message: SA28 V1.2 (06/26/2012)
#############################################################################

1.Fix ISA Bear Card test fail.

I dunno what this Bear Card really is, but an assumption would probably be one of the ISA test cards here... Still, I don't know what it really means about fixing failed tests, is it about DMA, or something else that shouldn't (e.g. I/O, IRQ)?
3. The board has an early UEFI-capable BYOSOFT BIOS available (released until 2011, the BR14 version. It works, but feels kinda buggy, given the amount of bugs mentioned in the changelog). In the BYOSOFT BIOS changelogs, contains statements about ISA support, and also another reference about Bear Card.

4. SA28 project support ISA Slot for ISA device.

----------------------------------------------------------------------------- Version : SA28BR14 Base On (SA28BR13) Date : 20 […]
Show full quote

-----------------------------------------------------------------------------
Version : SA28BR14 Base On (SA28BR13)
Date : 2011/07/06
CheckSum : 6183H
Sign-on Message : NO
Action :
1：Set IRQ7/IRQ5 default as "Reserved" to patch bear card test fail.
2: Set user press "DEL" or "F2" can both enter BIOS setup.

Those official statements seem to talk about ISA support that has been tested working, but in reality the ISA capabilities are the same as other boards of ICH6 or later. This is what confuses me. Additionally, chipset register hacks regarding ISA DMA obviously won't work.

>LSS10999
Yes, as I looked at IEI IMBA-G412ISA it uses PCI2ISA bridge ITE IT8888 (very popular on industrial MBs) so no way for DMA. Probably lot of industrial ISA cards are some I/O controllers or ADC modules that uses only IRQ no DMA so their users maybe happy. I found this reference for Bear Cards: http://www.beartech.com.tw/BearFunctionTestCardEng.htm - just a common diagnostic POST code display card - no need for IRQ/DMA, it just use I/O ports. The mentioned fix may include a proper configuration of the PCI2PCI/PCI/ISA bridges to pass writes to IO port 80h to the ISA slot (some magic of PCI substractive/aditive decoding that I don't understand but it's necessary to be properly configured to pass an I/O access through entire long path CPU-NB-SB-bridge-slot...).

I wonder if someone else tested a MB with LPC2ISA bridge...

RayeR wrote:

I wonder if someone else tested a MB with LPC2ISA bridge...

I believe someone had tested an Intel D845WR board with an LPC2ISA bridge and had zero success getting sound to work on it:

Re: Intel D845WR s478 no sound with ISA audio

---

I purchased a couple of the HT845-ISA/1 boards from AliExpress in the last couple years. They go by a lot of different names. Currently using one of them with a P4 2.8ghz/512k/400. These boards supposedly support 533 and can be configured 400 or 533 with a jumper, but I couldn't get it stable with a 3.06ghz@533. You can stick a 533 in there and set the board to 400, you'll lose 25% of your clock speed, but the BIOS will recognize this and tell you what you're actually running at compared to what CPU is in there.

These boards do work with SB16 and up. The DMA is incompatible on earlier cards and suffers from stuttering. YMF719 also stutters. I think the only two sound card lines I did not have issues with were SB16 and up, and AZT2316 based Aztech ISA cards. I believe this board is an almost identical copy of an Aimmer board, the AMI-I845GV-ISA. Variants with 3 ISA slots exist, and those come with a HT845-ISA/3 sticker instead. These boards are all keyed for 1.5v AGP only, so no older AGP cards allowed.

This board does not POST when the ATI Radeon 4670 AGP is installed. Additionally, video output fails on some faster AGP cards when using any high resolution (Geforce 3 Ti 500 and FX 5700 both gave me issues here.)

Also, if you get the 1 ISA slot variant, beware! The board is *ever so slightly* larger than Micro ATX spec, by a couple millimeters. This caused the board not to fit in my Micro ATX case because of the overlap from the PSU.

SaxxonPike wrote:

I believe someone had tested an Intel D845WR board with an LPC2ISA bridge and had zero success getting sound to work on it:
Re: Intel D845WR s478 no sound with ISA audio

He wrote about Winbond W83626F PCI to ISA bridge not LPC2ISA bridge. His MB should have ICH4 but maybe intel didn't wired all necessary signals to the bridge. I belive that most of 845/ICH4 MBs works...

RayeR wrote:

>LSS10999
Yes, as I looked at IEI IMBA-G412ISA it uses PCI2ISA bridge ITE IT8888 (very popular on industrial MBs) so no way for DMA. Probably lot of industrial ISA cards are some I/O controllers or ADC modules that uses only IRQ no DMA so their users maybe happy. I found this reference for Bear Cards: http://www.beartech.com.tw/BearFunctionTestCardEng.htm - just a common diagnostic POST code display card - no need for IRQ/DMA, it just use I/O ports. The mentioned fix may include a proper configuration of the PCI2PCI/PCI/ISA bridges to pass writes to IO port 80h to the ISA slot (some magic of PCI substractive/aditive decoding that I don't understand but it's necessary to be properly configured to pass an I/O access through entire long path CPU-NB-SB-bridge-slot...).

I wonder if someone else tested a MB with LPC2ISA bridge...

Apparently the Bear Card I found on that site does contain ISA functionality tests, which is in an included program that can be run under DOS and contains many aspects such as I/O, IRQ and DMA. I believe they simply skipped the DMA part since they were never functional, just they fixed all possible IRQ issues which might be required for stuffs like serial port or data acquisition cards. As far as I have tested, all ISA sound cards can work with everything that doesn't require DMA (notably FM synth).

Apparently IEI is doing some false advertising regarding the ISA functionality. They never mentioned a bit about ISA DMA being not supported, not to mention they even have non-functional DMA and Reserved Memory switches like earlier boards in BIOS. Generally speaking, most industrial motherboard manufacturers will clearly state that the ISA slots are Slave, or directly state that the board has no DMA support, when it comes to boards with ICH6 or later, so customers who require DMA should go after older but available boards that do support ISA DMA. By the way, MSI did not specifically state about no ISA DMA support on MS-98A9 which also uses IT8888, but a reseller (Adek) stated.

LPC2ISA bridge is rare, the RUBY-9719VG2AR is the only board I know that contained one, which has been discussed. I did try googling about anything that uses the bridge, Fintek F85226FG, but no useful results (mainly industrial SBC boards meant to be used on backplanes or such). To me it looks like LPC2ISA is rather uncommon among ATX boards, but common among embedded and SBC boards.

LSS10999 wrote:

Generally speaking, most industrial motherboard manufacturers will clearly state that the ISA slots are Slave, or directly state that the board has no DMA support, when it comes to boards with ICH6 or later, so customers who require DMA should go after older but available boards that do support ISA DMA.

In my case the Commell P4LA http://www.abi.cz/?cls=stoitem&stiid=13280&st … i-e-pci-isa-atx didn't noticed about non-working ISA DMA too. They admitted it when I already bought it and blamed to the support. Due to high post prices and taxes I didn't return this MB and use it in father's PC (ISA unused there).

RayeR wrote:

LSS10999 wrote:

Generally speaking, most industrial motherboard manufacturers will clearly state that the ISA slots are Slave, or directly state that the board has no DMA support, when it comes to boards with ICH6 or later, so customers who require DMA should go after older but available boards that do support ISA DMA.

In my case the Commell P4LA http://www.abi.cz/?cls=stoitem&stiid=13280&st … i-e-pci-isa-atx didn't noticed about non-working ISA DMA too. They admitted it when I already bought it and blamed to the support. Due to high post prices and taxes I didn't return this MB and use it in father's PC (ISA unused there).

Yeah, it's advised to take some serious look at what the board uses for ISA slots before buying in case of board description didn't explicitly state such. While most manufacturers do state, some don't. In case of MS-98A9, MSI itself didn't state about no ISA DMA support, but a reseller, Adek, stated such. AFAIK ITE IT8888 is the most common among ATX motherboards (especially the BGA variant IT8888G) followed by W83628F+W83629D combo. Both are PCI-ISA bridge solutions.

Some useful information for ISA chipsets were compiled in this page. There were mentions about virtualizing ISA DMA in software for AMD G-series APU, used mainly in embedded solutions. AFAICT modern AMD chipsets (notably 700/800/900 series) are legacy-free, that similar to the situation of those recent PCI-free Intel PCH, PCI sound cards couldn't access the built in legacy components such as FM synth even under Windows/Linux, so that a software solution was utilized most likely to address such dilemma when designing embedded SBCs with usable PC/104 slots (which is based on ISA) around AMD APUs. This whitepaper mentioned such, and from the archived page from the AMD Q&A, it mentioned that AMD worked with AMI for that functionality (the Q&A page also links to a PDF document, but it's not archived and access requires signup and with NDA).

EDIT: Should comment that IMBA-G412ISA is also not quite flexible. It doesn't have any option to tweak CPU settings (multiplier, Vcore, etc.), and it couldn't handle 771-modded Xeons properly. I tried a few Xeons (E5450/E0, X5460/C0, X5470/E0) and while they do boot, the system becomes unstable with program crashes (sometimes the CPU could even miscalculate CRC) and sooner or later BSoD with a random error code.

EDIT 3: Forget about it... guess the board is naturally incompatible with Xeons at some point... even after I tried modding the Vcore up a little bit...