Photos of DTK PTI-265W containing Winbond W83759AF.
Used in my 486 in MS-DOS and Win95A.

Bruno128 wrote on 2023-10-17, 10:00:

Photos of DTK PTI-265W containing Winbond W83759AF.

Nice looking board. It's not in TH99, but it looks like an updated revision of https://th99.dosreloaded.de/c/C-D/21633.htm

I haven't been able to get a UM8672AF board yet. It would be interesting to see if it is faster than a UM8672F

douglar wrote on 2023-09-12, 00:18:

I have a Boca Research 4475 based on an Atronics2015pl https://arvutimuuseum.ee/th99/c/A-B/20774.htm https://fccid.io/EUD5U9BRI […]
Show full quote

pshipkov wrote on 2023-09-11, 21:44:

Anyone ?

I have a Boca Research 4475 based on an Atronics2015pl
https://arvutimuuseum.ee/th99/c/A-B/20774.htm
https://fccid.io/EUD5U9BRI4475

I've been really busy at work for the last few months and my project sort of stalled.

I ran these tests a while back without any dos driver when I was checking to see if it worked:

Processor : AMD Am5x86(tm) 133 MHz
BIOS vendor : MRIcroid Research, Inc. 23 (01/26/94)
OS version : MS-DOS 7.10

Copy code to clipboard

132GB Msata SSD on a PATA adapter
2                  Tested in FAST mode
3                  Random access time     : 0.26 ms
4                  Linear read speed      : 1088 KB/s
5                  Min/Max read speed     : 1085 KB/s / 1103 KB/s
6                  Linear write speed     : 1024 KB/s
7                  Min/Max write speed    : 937 KB/s / 1207 KB/s
8                  Buffered read speed    : 5867 KB/s

Copy code to clipboard

1512MB SD on a Sinitechi
2                  Tested in FAST mode
3                  Random access time     : 0.80 ms
4                  Linear read speed      : 1479 KB/s
5                  Min/Max read speed     : 1479 KB/s / 1479 KB/s
6                  Linear write speed     : 1777 KB/s
7                  Min/Max write speed    : 1776 KB/s / 1779 KB/s
8                  Buffered read speed    : 1481 KB/s

Let me know if you want to see some other test and I'll run it

What is your test program to get those values?

Speedsys 4.78

I boot from floppy, choose a device driver through a config.sys menu, and then run a batch file that launches:

1speedsys HD0 T1 WT SP R PCX:%1.pcx REP:%1.txt 

Params:
HD0- Test Hard Drive 0
T1 - Quick HDD test
WT - Write test
SP - Skip PnP device detection.
R - Generate report files
And "%1" is a command line parameter that is the name of the output file I want to save.

Then I feed the output files into a parsing script that sticks them in a database.

I've recently learned that I should do a cold boot/power cycle after loading some device drivers because ctrl-alt-del doesn't always reset the IDE<-->controller speeds.

Working through the pack of cards, I'm up to the ADI/2 version of the Acculogic sIDE-4/VL

FCCID: IIV0041900A00

Testing Notes

• The Appian Turbo ADI2 driver didn't approve of this card: https://www.vogonsdrivers.com/getfile.php?fileid=2063
• The Adaptec ADI2 Flexi driver was OK with loading: http://vogonsdrivers.com/getfile.php?fileid=2046
• The Adaptec ADI2 Flexi driver only uses 2KB (!) of RAM
• The ATA speed change from the driver seems to persist through a warm boot
• This card would not run benchmarks with the Transcend Consumer CF, so I substituted a PQI Consumer CF that it did like
• This card would not run benchmarks with sata devices connected through a Jmicron Sata bridge, so I substituted a period correct WD Caviar 21000
• There were some unexpected latency fluctuations that make me want to retest this again at some point
• There was a line of undocumented jumpers near the ADI/2 chip that likely allow for speed control overrides but I didn’t play with them

Edit -- I reviewed and corrected Sinitechi / MR BIOS numbers near the bottom. Warm boot doesn't reset the ATA speed on this controller.

Today I tested the second Acculogic sIDE-4/VL, which as a CMD640 chip instead of the ADI/2. FCCID IIV0043400A00

First I tried the v2.09 driver. The driver was sweet because it gave all sorts of useful info but I could not get it to provide any benefit and it used 23K memory.

Then I tried https://www.vogonsdrivers.com/getfile.php?fileid=2073. The dos driver was v2.13 in that bundle. All the same sweet info but it only used 9KB.

At first I could not get it to provide any benefit either. Eventually I tried the config tool out of frustration. It informed me that I can't use the PIO override unless I also set the bus timings, and then I could get both the O004 and O005 overrides to work.

Notes:

• The ATA speed increase would persist through a soft reboot, but not the Multi word IO, which makes sense.
• XUB didn't like this controller. The drive name detection strings were screwy and the bench marks would lock up doing the seek tests.
• I lost some of my early benchmarks copying from the Gotek because they were mixed up with some questionable results and I deleted the lot of them, I'll have to redo them tomorrow
• I tested with CMD640x2.sys that was v2.13 with a 9/28/2023 date
• The driver reported that the Jmicron Bridge could only do ATA0, and MR Bios and DTC BIOS had trouble with it showing really low speeds
• PIO 5 was getting in the 3K range, but PIO 4 was much better, getting things in the 7000-10000 range, so I tested with "A33 o004" parameters
• The more I test the more I realize that there's probably a lot of tweaking I could do to get things to work or work better, but things were much easier 30 years ago when I had 1 controller, 1 driver, 1 disk and lots of time to work on it and no choices because I just spent all my money of the stuff I bought. Now I'm inundated with too many options and not so much time.

Reviewing the the numbers I've posted so far, it seems like things break down into 4 basic groups:

• Under 1000 KB/s This usually means that something is going wrong. I can often fix it by reseating the card. I had a group of cards that were blocked by the keyboard controller in the VLB slot I had been using, and it's taken a number of insert/remove cycles to make the other port reliable. However, on the CMD based card, there is one scenario that just reports < 1000KB/s regardless. Seems like something is having trouble there.
• 1500-1700 KB/s This seems like the default range for 33Mhz VLB with a 133Mhz AMD chip.
• 3500-5000 KB/s This seems like the throughput is doubled, minus some overhead that didn't double
• 8000-10000 KB/s and the throughput is doubled, once again minus some overhead that didn't double

It would be nice if these bands correspond to PIO 2, 3 & 4 but somehow I think multi sector transfers and 32bit transfers are complicating the situation and it is just coincidence that things have been lining up in 3 clear categories.

Scope or LA would let you verify disk side timings.

rasz_pl wrote on 2023-10-22, 07:22:

Scope or LA would let you verify disk side timings.

Would be nice if I had those and knew how to use them, but I don’t. I studied math in scool, not electrical engineering, so I’m going to take a theoretical approach. I’ll revisit one of the drivers that provides good debug info to see if I can get the device to tell me what is going on.

I seem to be stuck in mode 2 with the lion.sys driver without the optional parameters. Can anyone disassemble the lion3.sys driver from this archive: http://vogonsdrivers.com/getfile.php?fileid=1996 ?

Also, I've found evidence that these drivers existed, but I can't seem to find them:

• ts8310.zip: TS-8310 VESA/PCI IDE DRIVER TANS CO.,LTD. I can see an archived copy here but it is corrupt: https://web.archive.org/web/20140207072735/ht … trol/ts8310.zip
• 6560A.zip: I got the driver for HT6560B from here: Re: Holtek 6560 vlb ide controller but it doesn't like the HT6560A, and I can see a file listing here: https://storagedrivers.com/drivers/66/66313.htm

Hi, does anyone have drivers for the VLB controller on SIS 83C611 + WinBond W83757F.
Thank you in advance.
Regards

piokum77 wrote on 2023-10-25, 00:37:

Hi, does anyone have drivers for the VLB controller on SIS 83C611 + WinBond W83757F.
Thank you in advance.
Regards

Sis 83C611: http://vogonsdrivers.com/getfile.php?fileid=2036

Thanks.

piokum77 wrote on 2023-10-25, 06:04:

Thanks.

On my controller, I got the best performance using the parameters with 0 IDE Wait, 8 sector max read like this: Device = C:\611B01.SYS 0 8
Not all drives can do the 0 IDE wait though.

Here are the benchmarks that I got:

Questionable latency reading on the Xceed industrial CF that I need to go back and double check.
Without a driver, I'd say that it performed very well for a card without jumper based speed settings, however with a driver, it was definitively in the ball park of PIO 3 cards.

If you were wondering how I go so many VLB IDE cards, I the answer is I sold off my Alaris Leopard, Voodoo 5500, and low serial number sound blaster 1.5 for this:

+ a few more not pictured here.

Sort of like trading in some rare jewels for an assortment of 19th century trowels, but I'm getting a lot more use out of these than I was the museum pieces.

Ahhh, memories of the under-the-table $1 box at computer fairs of yore 🤣

@douglar
After reading your message i am starting to think that i should have ship you 25 or so cards for the testing you are doing here. : )

Btw, i think it will be more useful if you have a common chart with all numbers in one place.

Do you have GoldStar 711 and 712 cards ?
I have only one 712, but wondered if 711 model will be any different.
I have a note about the in the other thread that they are probably the same, and actually a relabel, but still.

douglar wrote on 2023-10-25, 11:50:

If you were wondering how I go so many VLB IDE cards, I the answer is I sold off my Alaris Leopard, Voodoo 5500, and low serial number sound blaster 1.5 for this:

Good man. Can you do PDC20230 and PDC20630 next please? The PDC20630 was very fast for me when I had a VLB system. There are two drivers for it. One of them would show only a smiley face ASCII character as its TSR identifier in mem /c.

I figured this is the place to ask. I have a DTC 2277 card which is clearly a re-badge of the ATronics CPDC200BRC using the 2020L chip. Despite the onboard BIOS being dated 1995, its LBA support is pretty broken. The DTC sticker says "VIC41" (c) 1995. I don't know what ATronics revision it is as removing the DTC sticker will likely trash the ATI one underneath. Is there any BIOS updates for these cards? DTC clearly didn't provide them. There also doesn't appear to be any 32-bit protected mode drivers for any other OSes out there either, so this thing likely doesn't work on Windows NT at all.

Looks like a premium caching VLB IDE controller otherwise, although I don't know how it is performance wise. Also, I have the manual if anyone needs it.

NJRoadfan wrote on 2023-10-25, 23:25:

I figured this is the place to ask. I have a DTC 2277 card which is clearly a re-badge of the ATronics CPDC200BRC using the 2020L chip. Despite the onboard BIOS being dated 1995, its LBA support is pretty broken. The DTC sticker says "VIC41" (c) 1995. I don't know what ATronics revision it is as removing the DTC sticker will likely trash the ATI one underneath. Is there any BIOS updates for these cards? DTC clearly didn't provide them. There also doesn't appear to be any 32-bit protected mode drivers for any other OSes out there either, so this thing likely doesn't work on Windows NT at all.

Looks like a premium caching VLB IDE controller otherwise, although I don't know how it is performance wise. Also, I have the manual if anyone needs it.

I've got a DTC 2177a with 1994 BIOS on two chips, VIC41-F and VIC41-B and I'm in the same boat.

The issue us that your computer doesn't talk directly to the drive. If I understand things right, your computer talks to the Atronics 202oL chip, which in turn talks to the AMD N80C188 chip, which is the CPU that talks ATA to the drives. The AMD N80C188 runs on code from the BIOS, so if the BIOS doesn't do LBA, there's no driver over overlay that will get you where you need to go.

Any chance you can upload your BIOS?

Correct. Caching cards need their own BIOS in order to work as your onboard BIOS' Int 13h firmware wouldn't know what to do with it. The card might expose the standard IDE registers, but that would bypass any caching. Both DTC and ATronics provided zero information about the ROMs on these cards. I don't have a reader, so no dumps.