Hi,
If this info is already around, sorry, I did a search and everyone seems to think the PVRs are limited to 160GB..

After much messing about, I managed to get my Techwood (Think it is a Vestel T816 box) to recognise a 320GB drive (rather than the now full 80GB drive it came with!).

The drive is a WD 320GB, I got as far as setting channels, formatting the drive, recording a snippet and playing it back. I will try a little more testing tonight and post more details on the drive I used And also if I manage to clone the info off the existing drive.....

In order for it to work, I had to ensure that there were NO partitions on the drive (windows disk management tools and a USB/IDE converter are your friends!) and I put the drive into CS mode.

Cheers
K
Last edited by karleccleson : 14-01-2008 at 09:50

What does the drive info screen say though? Although the drive works, it might only be using 160Gb. The T810 certainly only used upto 160Gb (even if larger drives were fitted). Although the T816 might be different, and the T825 certainly is different since it can be bought with a 250Gb drive.

That was the first thing I checked and it proudly boasted 319GB

As I say, I haven't tried to use the drive yet other than a quick test (only managed to get it formatted at midnight last night!).

I hope to do alot more testing tonight.

Evesham sold a T816 with 250GB and the Grundig GUSTB250IV T816 with Version 4.7 software I bought from Makro was also 250GB.

Colin

Ok,
I tried recording, dual recording, pause while recording and playback during recording. All worked fine

The machine claims to have 116Hours of recording time (2GB/h).

Using one of the numerous "Ext2 on windows" drivers out there I successfully copied the contents from the original drive to a windows PC. Unfortunately, copying to the larger drive caused my PC to reset (when writing directory information to the second partition I think). This trashed the second partition, so I think this is a no go at the moment....

I think the source code came with the driver, so I may see if it is possible to add the patch that has been metioned on other threads to the driver to allow the second partition to be written to.

Failing that, I still have 4x the storage

(I'll post the HDD details later, forgot to copying them down!)

It's always cool to do a mod like this but I wonder what the speed of the library menu etc will be if you actually use your new 4x-the-storage capacity? I only ask because I have a Wharfedale T816 250GB model which frankly has a silly amount of storage on it, and if I use more than about half, I find actually loading the library and navigating up and down it excruciatingly slow. It took me ages the other day to go through deleting programmes I had allowed to accumulate there - I may have missed something, but I couldn't mark several and delete them at once, I had to delete each one individually. I actually think a 160GB or even 80GB model would have been fine for my needs, as I would then have got into the habit of archiving things I wanted to keep on my DVD recorder instead of just being lazy and allowing my PVR to fill up!

I have found that 80Gb is definately not enough. My previous PVR had 128Gb (after an upgrade) and that seemed to suffice. But I have noticed that the recording sometimes fails to stop when a programme ends, so I end up with a 3hr (5Hr in some cases!) record rather than a 1hr. It doesn't take too many of these to fill up an 80Gb drive.

Also, I have a couple of 320Gb drives that I have removed from an (rubbish!) SC101 NAS box looking for a home.

Hi Karl.

I have been intregued with your efforts with the Digifusion Models, and appreciate that your level of computing knowledge is somewhat beyond mine!

I originaly used Win2fs to transfer files (in both directions!), and found that it was other functions within the windows OS that caused the impromptue reboot. I eventually just ran the driver and nothing else, and made sure I did exactly what I had planned without klicking my mouse too much elsewhere. This stopped the rebooting.

I noticed that the other windows drivers struggle because of the large second partitions block size. The patches you mention, effectively created a dummy file area on disk to solve the problem (for what reason, I didn't look into). I noted that the Win2fs driver copes in a similar way, since when the drives are mounted into widows, a new 'Cabot' directory appears on the second partition to mirror that of the first, thus effectively creating a 'dummy' directory. This doesn't appear to have had any detremental effect on the function of the STB, and appears to serve no other useful function.

Unfortunately, the Win2fs project hasn't been continued, which is a shame since it was very fast under windows. However, it is still possible to download the drive, and I think possibly the source as well. I have copies of both binary and source, if you are interested. I now also have the Ext2 Anywhere software from Paragon. It isn't free, but seems to do the job. Other Paragon software (partition manager for eg), which supports Ext2/Ext3, works OK as well.

Thanks for your work on the Digifusion. I was hoping that Syphon could have included downloading from off the FVRT90/95 models as well, as the files are TS rather than PS (which is of more help to me). Do you think you might support the FVRT95 (as models are now very cheap!) at some stage? Anyway, thank you.

Regards,

Tim.

Hi Tim,


Thanks for that, I'll give it a try Currently in the middle of adding a touch screen to my shiny new Asus eee (ok, it was my wifes shiny new Asus eee, but I have fallen for it!)


I think I have a copy buried somewhere, but need to get some tools to allow me to build it. If I cannot find the source, I'll PM you


Ok, I have some paragon stuff (magazine disks) kicking about too, so I'll see if this provides a quicker solution, I only need to do it once.


Your welcome, I am really enjoying the challenge!


Can you remind me, does the FVRT90/95 use the same STFS format as the 100's? Or does it use the new S/W base?
If it uses the same STFS, it shouldn't be too difficult, otherwise it would take some time!

Syphon Source is now available from my website for the brave

Hi Karl.

So sorry for the long lag time in response to your last post - seems a fair few of us have been having difficulty getting hold of Digital Spy since Friday! I understand a server somewhere was having difficulty. Anyway, It's back again.

I am not sure about the File System on the FVRT90/95s, but since the models I have just obtained very cheaply from a Maplin's clear-out are ex-Argos stock (and pre 2006!), I would say it wouldn't be a new version. I have come across a small program someone produced a year ago on the forum to extract the files, but that's all the program does - it doesn't allow any 'browsing' of the total disk content. The TS streams removed are also even more 'basic' than those on the Digihome/Vestel clones, and more difficult to get other software to directly read. At a brief glance at the Hex dump though, it is a bit difficult to see why. A little more investigation is necessary.

I am glad you have made the source for Syphon available to download. Although I was learning Visual Studio languages to get up and running with windows API, I would very much like to take a look at the driver's source, so thank you very much once again, and I will try to download the file from off your site, if I may.

Regards,

Tim.

Hi Tim,
I did a little digging, the FVRT90 is based on Access Devices firmware, so I don't see Syphon reading it in the near future....


Please do, it was written using Borland though, not VS.

Has anybody dared try anything bigger than a 320Gb drive? Just wondering . Also, what about different makes, other than Western Digital? I remember the T810 was quite picky in this regard.

I'm strongly considering swapping out my Samsung 80Gb drive for a new Western Digital 320Gb drive (WD3200AAKB), which are about £45 now. Ideally, a 5400rpm model would be better (IIRC, that's what the 160Gb T810 and T816 models were/are fitted with), but it seems these are now nigh on impossible to find (not going to get a secondhand one off ebay either as they tend to sell for more than brand new 7200rpm drives elsewhere).

It also seems to me that Vestel need to strongly consider switching to using SATA drives in future new hardware, as I notice these are cheaper than IDE drives now, and will most likely become even more so, whilst IDE drives will get rarer and more expensive as a result. Which reminds me, has anyone tried a SATA->IDE adaptor to use a SATA drive in their Vestel PVR?

I live in the Czech Republic. Sometime in October 2008 I've purchased a Vestel T816, sold under a local name of "GoGen 3050". Probably the same hardware still sells on the local market as Hyundai DVBT 530 PVR. Mine has Cabot firmware "v5.8 EU" and I've seen a piece of the Hyundai, purchased by the end of 2009, with the same firmware.

My Gogen box came with an 80GB IDE drive. By now the drive is hopelessly full of fairy tales and cartoons for our kid. The warranty is almost over, so I don't bother much anymore. I was wondering how much trouble it would be to replace the original drive with something bigger / less power-hungry. I have some 2.5" drives on my radar. While a 3.5" drive can draw some 10W, most 2.5" drives take only about 1.5W when active => switching to a 2.5" drive should save quite some juice and heat. In case you haven't noticed, IDE drives (parallel ATA) are getting very scarce. Seagate has phased out some former IDE drives as big as 750GB. Basically the only brand still making some IDE models is Western Digital, and even by that vendor, newly introduced models don't come in the IDE flavour anymore. The situation in 3.5" and 2.5" form factors is very similar. Besides, the SATA drives are cheaper (price per GB). So I was wondering, if there was some sorta bridge to allow me to connect a SerialATA disk to a legacy IDE controller. Somewhat to my amazement, such bridges do exist (so far I've only seen the other direction). After some searching on the local market, I've settled on this dongle. I borrowed two drives from my employer's stock and have had quite some fun with them

I borrowed a 1.5TB Barracuda ST31500341AS and some 500GB 2.5" Hitachi drive, both of them SATA models. I was uncertain about several things and had to cut some corners. I've first tested the SATA->IDE conversion dongle in a desktop PC (two very different PC's actually) to check how it behaves against a stock Intel IDE HBA. When using UDMA, the dongle worked amazingly well - it was hardly a bottleneck, the 1.5TB Barracuda ran at 100 MBps, close to its native SATA speed. When restricted to PIO though, the transfers were horrendously slow - around 2*MBps. In Linux, that is. Similar to some earlier experience with legacy IDE emulated by a SATA HBA. Armed with that knowledge, I unhooded the PVR. On the inside, it had a "sparse" 40pin cable, which would hint at PIO operation. The simple wiring of the IDE HBA in the Vestel box was another warning sign of possibly running PIO only. I actually had to drill an additional hole in the dongle's IDE connector, which has the classic "UDMA cable" lock instead of one of the pins (one hole missing in the female connector body), to be able to insert the dongle into the PVR's motherboard, which has a full set of male pins. Just to be sure, I powered the big drive by an external PSU during my experiments - to avoid overloading the internal PSU and whatnot.

Interestingly, the big SATA drives worked like a charm. The dongle has an activity LED, and clearly the PVR just spits a burst every second or two. Recording two channels and playing back an earlier recording was no problem either. To me this would hint that the IDE HBA actually runs in some UDMA mode. It would make perfect sense, the protocol is not that complicated and it saves quite some bus and CPU cycles to the host "computer" (PVR).

Knowing how long it takes to format hundreds of GB with EXT3 on a PC, I was amazed how quick the disk init went in the PVR. Later on, I attached both those SATA drives to a desktop PC and took a look with Linux on the partition structure. Fdisk showed two partitions, one of them small (just a few GB), and a bigger one. The first disappointment was, that the "big" partition is really just 320GB on both drives (500*and 1500*GB). The second disappointment was, that the Linux native in-kernel EXT2*driver refuses to mount the second partition. The error messages said "corrupt superblock", "inappropriate block group size" and "superblock magic number wrong", depending on what util I tried to unleash unto the rogue partition.
Especially given the relatively short init time, I would hazard a guess that this is not EXT2*anymore. Not just an obfuscated / crippled EXT2, it might be something completely different, maybe flat extents of video footage, indexed by some files on the small partition. I'm not skilled enough to reverse-engineer the format.

That's how far I got in my experiments. I haven't bothered to try and reallocate the second partition for a bigger size manually, using fdisk.

There are two unfortunate effects of this unknown format (file system) - it means that I won't be able to transfer the previously captured recordings to the new drive for the PVR, and less importantly, to a PC for archival.

Those were just lab experiments. For the upcoming "production upgrade", I've picked a Seagate Momentus 320GB 2.5" SATA drive. Essentially I trust Seagate more than any other brand. Western Digital are pushing the "Advanced Format" insanity, Hitachi drives over 128GB tend to have a quirk against older versions of the Linux kernel's ATA layer (LBA28/48 off-by-one thinko at the 128GB boundary, hinging around a different interpretation of the ATA standard). Fujitsu/Samsung/Toshiba? Not worth the bother if all we ever buy is Seagate.
I'll have to piece together a custom power cable, to power the SATA drive and the conversion dongle (floppy power connector) off the original IDE HDD-style Molex connector. No problem there. I also have some 3.5"/2.5" mounting brackets for the notebook drive. The worst part is, that I'll have to hack a male-female extension cord for parallel IDE, because the dongle is too tall, and would prevent closing the hood. I know that some DIY electronics stores have crimpable male and female connectors for flat cable, suitable for IDE. But that would cost me time and money, so I've just unsoldered a male connector from a scrapped motherboard and will probably just solder that onto a pigtail made from some "sparse" (40pin) PC IDE cable.

Any further hints regarding the weird on-disk format would be welcome. I haven't tried the Windows EXT2*driver yet, but I don't give that much chance anyway.

Based on what I know about similar-priced competition, the Cabot firmware is pretty good - comfortable to use. The librarian module would use a bit more intelligence and features, but overall it's pretty good the way it is. Hell, my mom learned to use it instantly, based on just previous knowledge of a VHS VCR, and my brief 30second explanation that the "PLAY/PAUSE/STOP/FF/REW/REC" work just like on any old casette recorder, and that "LIB" meant a "library". My son recently started to use the "pause" button when he was barely 3 years old (doesn't yet grasp the difference between time-shifting a live broadcast vs. pausing the playback of a library recording). It's nice to hear that the code base is being devolped further, for new products. The hardware is also pretty good - both the IR remote and the DVB-T tuners are quite nicely sensitive. If only there was an easy way to buy the Vestel hardware straight from the manufacturer under a stable original brand, just like we were used to buying Sony or Panasonic. The way it is, if I want to buy a Vestel/Cabot twin-tuner PVR in my country, I have to wait for the single importer to get inspired to have a containerload of them assembled and ferried into our country under some obscure noname brand. There are long months of no availability. Of course I already have one, but I'm recommending them to my friends - so I know the supply chain appears like vapourware half of the time.

Maybe an opportunity for you to set up as an importer?

Whilst looking for something to do the opposite I came across many different versions of this on ebay, many different prices and sources, mainly china like this one, but basically the same design. (Hope that link is ok - converted it from a uk link to the czech one). It's more compact and self contained needing no extra power connectors. I can't vouch for its effectiveness, simply because it's an unknown quantity to me at present.

I have an update. After some thought, I picked a Seagate Momentus 5400.6 ST9320325AS as the right drive for my upgrade. And guess what: there is a gotcha. A show-stopper, really. The drive essentially works in the PVR, except for "waking up from suspend". If you turn on the PVR, wait a while for it to spin down the drive, and then press the "record" button, the drive obediently spins up, but the PVR firmware (user interface) hangs, saying just "starting to record - please wait". The only way out is via a cold power-cycle. Interestingly, it seems that read transactions (such as opening the library, which also spins up the drive) do not cause a hang.

It really seems as if, on spin-up, the PVR issues some comand and waits for a particular response, that however never comes.I don't know the details of how exactly the PVR puts the drive to sleep. The Linux-based HDPARM util offers two levels of explicit/immediate power saving commands: stand-by and sleep. In addition, you can also tell the drive to spin down based on its internal timer after N seconds. No idea which method the PVR uses. And, no idea exactly what the PVR does to wake up the drive from the spun-down state. Normal Linux just issues the normal access command, which takes some time to finish (due to the spinup required), which Linux doesn't notice / complain about. I wasn't able to reproduce the problem in vanilla Linux, using that same drive and that same IDE/SATA converter.

I checked again with a Hitachi Travelstar 500GB drive (2.5" 5400RPM notebook model with two platters), and interestingly, that drive works just fine. The original 3.5" drive is a Hitachi Cinemastar - maybe the brand matters for some reason

I've also tried to power the Seagate drive by an external PSU, to isolate a potential problem with the internal power supply being too weak... to no avail.

Another note: I've tried accessing the second (data) partition on the drive from Windows using Ext2fsd 0.46 from Windows, rather than native Ext2 from Linux - and guess what: the data partition is indeed accessible and does contains the files mentioned by others, yippee ))