• Please review our updated Terms and Rules here

Kicking around a Tava PC

SomeGuy

Veteran Member
Joined
Jan 2, 2013
Messages
4,444
Location
Marietta, GA
Recently grabbed a rustbucket Tava computer off of eBay. I had been looking for one of these for a while. After a LOT of cleaning, I have it fairly presentable. It is not perfect, but it has been very interesting to derp with.

The Tava PC was among the first clones to use somewhat generic form parts, and it did not try to add features beyond that of IBM. The Tava brand seemed fairly short lived, but it did make it in to the compatibility lists of a number of programs.

Tava Computer Front.jpg
Behold the Tava PC - Yawn at its magnificence! :p

Like some of the other shorter lived early clones, there is little information out there. Hopefully this provides a bit of useful information.

The name of the game here is CHEAP, CHEAP, CHEAP. Reportedly, Tava massively undercut IBMs prices. And looking at it, I can see how.

Let's start with the motherboard:
Tava Mb Front.jpg
Tava Motherboard Back.jpg

The motherboard is an IBM 5150 clone. It only has 5 sockets, and uses the same spacing as the 5150. The board is the same size as a 5150 and has the same mounting holes. Aside from the missing cassette port and a different power plug, this board could directly replace an IBM 5150 motherboard.

The sticker on the back calls this a "Model 96000, Config 1.5". The motherboard in this one differers a bit from that shown in a PC World review, as well as another one that had been on eBay a way while back.

This board features on-board serial and parallel ports, uses an NEC 8088, and accommodates 256k of motherboard RAM. It *looks* like the earlier motherboards may have only held 128k.

I had wanted to dump the BIOS, but disappointingly, the BIOS was at some point "upgraded" to a generic 1986 DTK BIOS. While this suggests that the motherboard hardware is IBM PC compatible, it also suggests that the original BIOS had problems that warranted an upgrade.

The construction of this motherboard is a bit shoddy.

There are some re-soldered chips, one chip with a messy "piggyback", and various patch wires. A string of patch wire on the bottom between the ISA slots suggests the traces for power were not thick enough. The board is held in with plastic screws, as traces run too close to the mounting holes and could short. A couple of chips next to the ISA slots are right up against each other. The parallel port uses a DIP socket and plug. And one of the ISA sockets isn't quite soldered in right!
Tava Piggy Backed Chip.jpg

Amazingly, this mess actually boots.

Reportedly, the original boards were manufactured by Faraday Electronics, who sold boards only to OEMs like Tava. I don't know if this one was a Faraday board. The only markings on it are Tava, copyright 1984, "SS", and "Part No. 1001 Rev B"

According to several small snippets from Google Books, Tava did at some point start obtaining similar motherboards from manufacturers other than Faraday. Tava continued to use copies of the Faraday BIOS in these boards in violation of their contract, which landed them in some hot water. So perhaps this is one of the later non-Faraday boards?

Tava Case.jpg
The case itself is not really generic. The slots and mounting will fit an IBM 5150 motherboard, but the the power supply form and connector is proprietary. The drive bays are designed for two full heigh 5.25" devices.

BTW, the case and cover attach with a pair of latches in the front of the case. To remove the top part of the case, you must have the system on a good flat surface, unscrew only the two lower screws on the back, and then push the top off with quite a bit of force. Similarly, you must push it back on with quite a bit of force.

Tava Floppy Power Supply.jpg
The power supply itself looks kind of cheap. I had an issue with the +5 volts line to the floppy drives not getting power. Looked like a bad solder joint, but a touch up fixed that.

Tava Floppy Drives.jpg
Although the case is designed for full height drives, Tava PCs came with half height drives. The two drives are awkwardly held together with a set of shims. If the screws aren't tight the top one can slide around. Reportedly, the earlier ones used Qume drives, however this one uses Matsushita JA-551-2N drives. They don't look impressive, but they work and test out OK.

Sadly this unit did not include its original video card. I believe it had been upgraded at some point, and then whatever video card removed (possibly by the seller).

It is not clear if TAVA designed and provided the video cards themselves. Video card compatibility would have been a major factor in determining overall IBM compatibility. It would have been good to test.

This system does work with a standard 8-bit ISA compatible VGA card.

It also came with a WDXT-GEN hard disk controller, but no hard drive. That controller is of a bit later vintage, and would not have shipped with it. Given how this system seemed to have been stored, any hard drive probably would have been dead.

Tava Floppy Card.jpg
It did come with what is probably its original floppy controller. It is a fairly ordinary low-density controller, however it has TWO internal headers. Like the IBM PC/XT floppy controller, it can control up to four drives, but unlike IBM's the third and fourth may be internal.

With everything put together, the next trick is booting it.
Tava Computer Assembled.jpg

Surprisingly, the power supply did not explode.

I thought it was a fluke that my Columbia Data Product 1600 did not like XT/AT switchable keyboards, however this motherboard also does not like them. So, I had to use my one earlier non-switchable XT compatible keyboard. I tried dropping in a Phoenix BIOS that works on another machine, but it still did not like the keyboards. Whatever the issue is must be in the hardware.

Once powered up, the TAVA happily boots IBM PC-DOS. Throwing a few test applications at it, it seems over all compatible, and acts like any other PC clone.

Speed wise, it tests the equivalent of a 4.77mhz IBM PC or XT.

So, it is a mess, but it is a historically interesting mess.

Tava also produced a small 80186 laptop-like computer called the "Tava Flyer".

However, TAVA did not last much longer. A 1985 PC World reported: "Battered by IBM's price cuts and the limitations of a one-product line, Tava was swallowed by Replitech, one of the Tava PC's primary distributors. Tava management sold all manufacturing and marketing rights to the New York firm, which promptly began producing a PC compatible of its own-under the Tava label. Still with us? Coinciding with Tava's metamorphosis, CompuShack, the California- based computer wholesaler that originally spawned the Tava PC, ceased operations."
 
Nice to see this thing documented. I saw the one posed in the Clones on eBay thread, and it's PSU intrigued me. Not enough to buy it and have it shipped to the UK though. Never seen anything like that on a clone in the UK. For some reason I thought TAVA was a Russian clone. Any idea of how much cheaper it was than the IBM?
 
I really enjoyed this write-up! The 1980s spawned some cut-rate ramshackle computing contraptions, and this one takes the bill. It's hard to believe this computer was the product of a California firm that ran full-page ads in PC Magazine, and not some far-east no-name affair.

Some of these build and design choices leave my mouth agape with the words "what on earth were they thinking?" Like the unused solder pads for expansion slots in between expansion slots, for instance: They don't line up with the case and they're so close together that cards probably wouldn't fit if you tried to use them, so why are they there in the first place?

Is the front bezel plastic (à la XT), or stamped metal?
 
Any idea of how much cheaper it was than the IBM?
That completely depends on exactly when and what configuration. From what I have read, the difference was big bucks until IBM slashed their prices.

This July 1984 PC Magazine article https://books.google.com/books?id=Bwng8NJ5fesC&pg=PA239 compares bare bones clone motherboards, and lists a bare bones Tava PC with Faraday motherboard, case, and power supply for $1295.

If anyone has ever wondered where the unpopulated (no chips) ERSO XT clone boards came from, just look one entry up at the "Super PC". Bare board $95! (A photo a few pages up shows it as a common ERSO style board)

Different Tava Computer.jpgPC World Farady Board Tava.jpgPC World Tava PC.jpg
For completeness, here are couple of photos from several PC world articles and a photo from a eBay listing that had been on eBay a few years back (but I missed the chance to bid on).

These show what appears to be the earlier motherboard, and possibly a more "stock" configuration.

Tava Brochure 1.jpgTava Brochure 2.jpg
Also, a couple of graphics from a Tava brochure.

Is the front bezel plastic (à la XT), or stamped metal?
Yes, the front bezel is plastic.

(Oh, and the black 5.25" faceplate blockouts are just some random "modern" ones I stuffed in there)

The number of expansion slots on a 5150 were sometimes seen as a bit limited. It seems like smaller, thinner cards without a backplate might just fit in such slots. Many of these early clone boards were designed for "embedded"/industrial use and might not need the standard configurations. Haven't checked to see if any of those slots would line up with an XT case.
 
Last edited:
Well, I managed to acquire a Faraday motherboard. This is labeled as a FE 6400 series board. There was nothing to indicate what kind of system it was removed from, but looking closer at the earlier eBay Faraday/Tava picture (see post #4), the board looks almost identical - including the lack of Tava specific branding.

Faraday Board - Front.jpg Faraday Board - Back.jpg

As you can see, this is a much more professional looking motherboard. There are a few patch wires on the back, but it seems much more solidly built. Looking at the chip date codes, it looks like it was built in 1983, although the BIOS has a 1984 copyright date.

My Tava motherboard appears to have been built in early 1984. So this Faraday board is probably a bit earlier.

This Faraday board has 256k of ram, in a rather odd layout, and TWO built in serial ports that use DIP socket connectors.

A little bit of oddness with the power connectors that I didn't notice earlier. The connector on the Faraday motherboard has two extra pins. There is a pin for an extra ground and a pin for an extra +5v. Both connect directly with with other ground and +5v. My Tava motherboard has solder pads for these but they are not used, nor does my power supply plug provide these.

Still, after verifying the other pins matched, I was able to connect the Tava power supply to this board without anything blowing up. It just leaves the top two pins unused. Not really needed unless drawing a lot of power.

Looking really closely at the full sized image of the earlier eBay Faraday/Tava, it looks their power supply plug does the exact same thing.

For reference, the power pinouts are:

(pin 8 is closest to keyboard connector at the back of the machine - only numbed on the Tava board)
8: Ground (not provided by Tava PS, no pin on Tava MB)
7: +5 (not provided by Tava PS, no pin on Tava MB)
6: Ground
5: -5 (not provided by Tava PS)
4: -12
3: +12
2: Ground
1: +5

Wait, what? No -5 volts from the power supply?

Looking closer at my Tava motherboard, it has a regulator on it to provide -5 volts from the -12v.

Also, again looking closer at the earlier eBay Faraday/Tava, the ONE difference is the presence of a regulator.

So, this Faraday board does not have a regulator installed, suggesting it was not previously used with a Tava power supply. If I want -5v using this power supply, I would need to add a regulator. Meh. I don't think anything here needs it.

Now, the next trick is getting it to boot. As it turns out, there is some issue with this Faraday BIOS.

With this BIOS installed, it just sits there like a lump on a log. No video, no nothing. The dumped contents look fine, but it seems to be expecting something.

Also tried feeding this Faraday BIOS to the PCE emulator, and PCE also sat there like a lump on a log.

The catalog on Bitsavers mentions a serial monitor program. Perhaps it is waiting for some response on the serial port? I probably don't have the switches set right.

ftp://bitsavers.informatik.uni-stuttgart.de/components/faraday/1986_Faraday_OEM_Catalog.pdf

However, plopped in the generic DTK BIOS and the board springs to life and boots MS-DOS. Yay!

Hmm, I'll have to do more testing and fiddling later.

Here are dumps of the ROMS:
View attachment Tava DTK ERSO BIOS 2.38 1986 (U87).zip

View attachment Faraday 6400 R1_7 BIOS 1984 (U110).zip
 
The catalog on Bitsavers mentions a serial monitor program. Perhaps it is waiting for some response on the serial port? I probably don't have the switches set right.

Connect a terminal. I'd like to find one of those A-tease motherboards. Interested to know how compatible these actually are. It looks like they gave up on the serial console idea. I'm guessing due to software commonly bypassing the BIOS routines for video output.
 
Hmm, tried connecting a laptop as a serial terminal and got nothing. Plugged the DTK BIOS back in and ran a COM port diagnostic and it seems like there is something wonky with the UARTs. I could send test data out on COM 2, but not receive, and nothing on COM 1 which also intermittently returned status failures. Not sure how much more I can do since the UARTs are soldered on the the board, and there are no schematics.

I noticed my Tava motherboard has the SW1 and SW2 labels reversed from the normal IBM DIP switch blocks. The switches on the Faraday board don't really have meaningful labels. I think I finally have them all set right, but hard to tell without documentation.

The DIP sockets these motherboards use for plugs are the opposite direction on the Faraday board than on the Tava board. Even more confusingly, looking closely at the PC Magazine TAVA, it looks like that one has each socket oriented in opposing directions. WTF?

Instead, tried plugging the Faraday BIOS in to my Tava motherboard instead. From what I have read I would have expected it to work, but it also sat there like a lump on a log, and no indication anything was happening with the serial ports.
 
I acquired yet another Faraday board. It is almost the same as the first Faraday board. The BIOS chip is labeled "Copyright 1983 Faraday Electronics 1.5", which suggests this one is slightly earlier.

Faraday Board 2.jpg

Unfortunately, when I tried to dump the BIOS ROM I discovered the chip was dead! My eprom programmer kept complaining that the +5v pin was not connecting. Sure enough, it measures as no continuity at all. Unsurprisingly, trying it in a ROM socket did nothing.

It's like this BIOS just wants to stay hidden to history.

The good news: I popped in a Phoenix BIOS and the board came to life! Even better, I ran various diagnostics, and both serial ports check out fine as well as everything else. So now I have a reference to repair the first Faraday board.

I also tried the Faraday 1.7 BIOS from the other Faraday board, but like all the others, it just sat there and did nothing. I have not yet delved in to the BIOS code to see what it thinks it is doing, but I suspect that ROM is just corrupt. So strike out on the Faraday 1.7, Faraday 1.5, AND the Tava ROMs.

There is one oddity about this board - the option ROM socket is populated. There is a hand written label that reads:

"PARK
SEEK306
SEEK1280
RECAL
FORMAT"

The entire content is:
Code:
00000000  B4 0C B0 00 B1 41 B5 31-B6 00 B2 80 CD 13 90 CC    
00000010  B4 0C B0 00 B1 41 B5 30-B6 00 B2 80 CD 13 90 CC    
00000020  B4 0C B0 00 B1 01 B5 82-B6 00 B2 80 CD 13 90 CC   
00000030  B4 11 B0 00 B1 01 B5 32-B6 00 B2 80 CD 13 90 CC   
00000040  B4 07 B0 06 B1 01 B5 00-B6 00 B2 80 CD 13 CC 9E    
00000050  FF FF FF FF FF FF FF FF-FF FF FF FF FF FF FF FF   
00000060  FF FF FF FF FF FF FF FF-FF FF FF FF FF FF FF FF    
00000070  FF FF FF FF FF FF FF FF-FF FF FF FF FF FF FF FF    
00000080  FF FF FF FF FF FF FF FF-FF FF FF FF FF FF FF FF    
00000090  FF FF FF FF FF FF FF FF-FF FF FF FF FF FF FF FF    
000000A0  FF FF FF FF FF FF FF FF-FF FF FF FF FF FF FF FF    
... more of the same FF fill ... 
00000FF0  FF FF FF FF FF FF FF FF-FF FF FF FF FF FF FF FF

It is not a device expansion ROM. It only contains a tiny amount of code which seems to be just a few BIOS calls. It is set to map in to memory at F800:0000.

Not sure why anyone would even put this in ROM unless it was executable directly from the BIOS.

One other thing, these Faraday boards have a header, labeled J14, for a reset switch. The Tava does not have that. Nice feature to have :p
 
Connect a terminal. I'd like to find one of those A-tease motherboards. Interested to know how compatible these actually are. It looks like they gave up on the serial console idea. I'm guessing due to software commonly bypassing the BIOS routines for video output.

I gave one away about a year ago, complete with schematics. Pretty close to a real 5170, but the onboard serial ports can be configured for RS423. Interesting in that it had an 8 LED DIP display that showed the POST codes. I used a DTK BIOS with it--someone had installed the original BIOS PROMs backwards.

This was part of a giveaway of a AT&T 6300 (with AT&T daisywheel printer) and a TRS-80 Model 16, both working. All the giftee had to do was show up on my doorstep and I filled his car up. I would have added more stuff, but he didn't have room.

I'm not a collector and I didn't want to be bothered with shipping this heavy stuff.
 
Another TAVA PC just popped up on eBay: https://www.ebay.com/itm/Tava-PC-IB...oard-and-Power-Supply/183839232195?nordt=true

Looks in fairly good condition, but it is pretty well gutted. Just the case, power supply, motherboard, and a floppy controller. No drives or video card. At least it seems a bit unlikely a bare-bones base system would have survived unused.

The tag on the back calls this a "TAVA-PC Config: 1.4 Part No: 76000", "Made in California, USA."

It appears to have a Faraday motherboard, and it looks like this board has a Faraday BIOS chip. There is a TAVA brand mark on the board. This motherboard has two serials ports and a parallel built in. The parallel port on the back appears to be Centronix.

Interestingly it has a "PRELIMLINARY TECHNICAL MANUAL FOR THE TAVA PC" with it. (That is really how they spelled it).

If anyone here gets it, a scan of that manual might be... interesting. As well as a BIOS dump. (While I am tempted, there are so many other things I need/want right now).
 
I did purchase this PC from eBay. i will work on getting the scan of the manual up. Unfortunately the unit arrived with A small amount of damage to the motherboard. There was no screws in the board and it was floating around. There was one IC missing. i will list the IC position later today. The IC missing is unusual location. The socket pins is all that remains
as the plastic shell and IC is gone. Possibly from floating around in shipping. Luck would have it that the IC has escaped out A rip in the shipping box. other than that the case is in great shape and board looks undamaged. Just the missing IC and socket shell. I hope someone can tell me what the IC should be once i get home I will list the IC location. Thanks
 
eBay TAVA

eBay TAVA

Another TAVA PC just popped up on eBay: https://www.ebay.com/itm/Tava-PC-IB...oard-and-Power-Supply/183839232195?nordt=true

Looks in fairly good condition, but it is pretty well gutted. Just the case, power supply, motherboard, and a floppy controller. No drives or video card. At least it seems a bit unlikely a bare-bones base system would have survived unused.

The tag on the back calls this a "TAVA-PC Config: 1.4 Part No: 76000", "Made in California, USA."

It appears to have a Faraday motherboard, and it looks like this board has a Faraday BIOS chip. There is a TAVA brand mark on the board. This motherboard has two serials ports and a parallel built in. The parallel port on the back appears to be Centronix.

Interestingly it has a "PRELIMLINARY TECHNICAL MANUAL FOR THE TAVA PC" with it. (That is really how they spelled it).

If anyone here gets it, a scan of that manual might be... interesting. As well as a BIOS dump. (While I am tempted, there are so many other things I need/want right now).

Got the TAVA from eBay. Will be scanning the manual soon as I get a break and reading the rom bios. Any ide what U65 should be it’s missing on mine. The pins of the socket is all that’s left the IC and socket plastic top are missing hope someone can identify what IC I need Thanks
 
Got the TAVA from eBay. Will be scanning the manual soon as I get a break and reading the rom bios. Any ide what U65 should be it’s missing on mine. The pins of the socket is all that’s left the IC and socket plastic top are missing hope someone can identify what IC I need Thanks
Oh, I can't believe I missed your post! Unfortunately users with less than 10 posts are delayed for moderation here, so they don't always show up on the new posts list. That, and I have been very busy the last couple months.

On my boards, U65 is a SN74125N. Hopefully nothing else got damaged. I'd still be interested in seeing a BIOS dump whenever you can get one. If you don't have time to scan the manual, could you at least post anything it has to say about the dip switches? Thanks!
 
Another Faraday board is on eBeh, but the seller wants an absurd amount of money: https://www.ebay.com/itm/174124329332

It sort of looks like an early board, as it is covered in patch wire. Looks like bios is "1.4" but the number is hand written. Here is an attached picture since eBay stuff disappears eventually.

Faraday Electronics motherboard Bios r1.4 - Small.jpg
 
This just popped up on eBay inside a lot of motherboards:


Advanced Logic Research 8088 PC motherboard.jpg

At first it looks like an IBM 5150 motherboard, but It is an Advanced Logic Research 8088 PC motherboard, that should fit an IBM 5150 case. It looks a lot like the later Tava motherboard, the main difference being that this one appears to have an IBM PC style power connector.

The advantage of this board over a 5150 would be that it has one built in serial and parallel port, as well as solder pads for extra cards that do not need a backplate.

Unfortunatly, it looks like this one has all of its socketed chips pulled.
 
Here's another Tava PC motherboard example. Appears to be the same MB from the chassis in post #1. Same part number "PART NO. 1001 REV B", stacked U82 and even has the yellow stickers on the CPU. I did not check the bodge wires matched but there are a few on this one.

tava_pc_front.png

This one has a Tava V0.8 BIOS. Internally the release date is "04/15/84". At offset 0 there's this copyright notice "(c) 1984 RMS Systems" followed by a call to display "TAVA PC VER 0.8 ". Zipped BIOS attached.

tava_pc_bios_V0_8.png

After a visual inspection only DRAM U9 needed cleaning up and its socket replaced. Thanks to the info in this thread putting together a power cable was straight forward. Equipped with a no-name monochrome card, DTK floppy card and a couple of 360K drives it boots various DOS 3.xx without issue. With a little more memory and loopback plugs the board passes all applicable CheckIt 2.1 tests.

A quick disassembly of the BIOS did not reveal any extra activity with the built-in serial port.
 

Attachments

  • Tava BIOS V0.8 1984 (U87).zip
    6 KB · Views: 5
Very nice. Thanks!

Just a quick test of the ROM in an emulator, it seems to crap itself executing non-existant code after displaying the copyright banner. When you dumped it, did you verify with the contents to make sure you weren't getting anything intermittent?

Although I recently fitted my TAVA board with a Phoenix XT BIOS, and the S1 and S2 were reversed, expecting new settings that did not even vaguely match the original way the DIP switches were set. Once I figured out what settings would make the Phoenix BIOS happy, the system ran very well.
 
The Tava ROM image runs fine when loaded it into an EPROM emulator at U87.

As a quick crosscheck I get following crc32 and sha1sum on the unzipped ROM:
crc32 - 5aeb1b17
sha1sum - 2278c5b278aa1e0e764e7cd89f4854c4ebb86f51

The ROM checksum 0x4F at 0x1FFF also seems correct.

Loading "DTK ERSO BIOS 2.38 1986 (U87)" from post #5 into the EPROM emulator results in a lot of screeching (not at beep code freq) and seemingly random beeps then nothing.

I found the configuration switches match those of the IBM PC. Here is my current setup in case you are interested:

S1
two floppy drives
no math co-processor
all MB banks populated
MDA graphics

S2
640K system memory

switches.png
 
If anyone had a Tava PC for sale, please let me know. This was the first PC compatible I ever purchased and started writing Qmodem on it. Thank you in advance!
 
Back
Top