Very recently, I was thinking of expanding my satellite system with a third and fourth receiver, only to be shocked to discover that my favourite tuner (the pricey, but otherwise rock solid TBS 6925) has been discontinued. The replacement, the TBS 6983, features twice as many “professional” mode capable tuners for the same price, but does not feature a loop-out port. This means you cannot daisy chain tuners together to receive different transponders on the same polarity from the same LNB output, or get some spectrum analysis happening. This is a pretty desirable feature for those who know the limitations, as often, multiple simultaneous feeds are on the same polarity and this would save a lot of LNB outputs, cabling, or the installation of a multiswitch ahead of the tuner.
It seems like most tuners have since removed the loop-out port, and they are indeed, rare to find. One of the alternatives to the TBS 6925, recommended by some on satellite tuning forums due to Crazyscan support, is the Prof Revolution S2 8000. After discount codes, my purchase came out to AU$110 per card, which wasn’t that cheap, but I did buy two as I had wanted to use both in a new HP Microserver N54L machine.
The unit comes in a cardboard glossy fold-open box, which advertises many features on the front. One of the “features” is the PTG Engine 1.7, which is just their fancy marketing term for their drivers. They claim on their site that their drivers are extremely well tested, BSOD-free, stable and compatible with a vast majority of software, but this is couched in the most broken English and marketing fluff. They make vague and irrelevant claims that their drivers somehow improve the signal reliability.
In reality, as most tuners are based upon modules, making a tuner card and its drivers should not be highly complicated. As long as the driver implements the code to control the “bridge” interface between the PCI-E bus and the network interface module (NIM), to control the NIM itself (the front end and the demodulator) and to shuffle this data “synchronously” in parallel through the bridge into the RAM, and then pass it off to the BDA stack, it should operate just fine.
Making a driver BDA compliant should mean it is capable of working with the vast majority of tuning software. Specialized blind-scan support would just mean the driver also has another interface for direct hardware access to the serial bus to control the demodulator and front end directly.
The card itself does have a loop-out port as seen in the pictures, supports low-profile bracket usage and is based on the STV090x demodulator but without support for 16APSK, 32APSK or ACM/VCM modes.
Curiously, there is a logo claiming that they are #1 in the world, but by what measure, I don’t know.
The box is sealed with a QC Pass label, fairly similar to the TBS cards. The rear tends to repeat most of the text already found on the front of the box, but with several gems, including “BDA (Broadcast Driver Architecture) driver supports many 3rd party applications party applications“. Okay, lets party!
It also has a claim the card is designed in Germany and made in China in Shenzhen. I’m not too sure about this, but they do claim having tested with Hotbird as part of their marketing spiel.
All over the other sides of the box are logos to the level of almost overload. The unit has many “self made” logos which are irrelevant as the unit doesn’t come with its own tuning or blind-scan software except for a data tuning program, and thus support does also depend on your supplied software. The card itself has a unique MAC address for data services as well. Guess what? 00 22 AB is the OOI for Shenzhen Turbosight Technology Ltd., better known as TBS! Isn’t that a surprise?
Inside the package, you get the card inside an anti-static shielding bag. The card comes pre-configured with the full-size rear bracket, and has a connector for additional power supply, of the same style as a PCI-E 6-pin VGA power connector. This seems to be a bit of an annoyance, as the card certainly doesn’t require that much power (10W maximum), and is more a precaution against weak power supplies.
The board is based around a Conexant PCIe A/V Decoder CX23885 which has the job of shuffling data from the metal-can NIM via a parallel TS interface through the PCIe bus, and receiving commands to pass back to the demodulator and front end of the NIM over I2C. The card has a lot of surface mount solid electrolytic capacitors, although they seem barely soldered onto the card, and I did have one “knock off” as I was trying to squeeze it into the microserver’s case. Luckily, I was able to use my hot air gun to resolder it back, as the joint failed and the trace didn’t lift. The power circuit seems to be protected by a self-resetting polyfuse FU1 near the middle-bottom of the card.
The card is dated Week 38 of 2012, making this pretty old stock. It does have a version number of v2.1, in the same position you’d expect it based on TBS cards.
The rear of the card doesn’t have any components, but the card itself did have a small scratch on the blue solder resist. The soldering quality seemed a little inconsistent on the top side of the through-hole connector.
Bundled with the card is a low profile bracket, which is easily installed by removing two screws to undo the full size bracket, inserting the small bracket in its place, and replacing the screws. A cabled IR receiver is also provided, and a 6-pin to three molex power adapter cable. The wires in the adapter cable seem very thin, and the use of three connectors is a bit overzealous in my opinion.
The unit comes with a quick installation guide, an 8cm pressed driver CD and a remote control, but without any included batteries.
The Driver “Situation” – What’s Up Prof?
While the hardware looks rather fine, and the website continues to reassure its market about the quality of the drivers, what I found in my installation could not be described as plug and play, nor could it even be described as stable.
The target machine runs Windows 7 x64 edition, which would be amongst the most popular operating systems today. The first step was to start with the CD – inserting it greets us with this surprise – a volume label of TBS-. This is another indication of the TBS and Prof “link”. Unfortunately, opening the CD and the wizard has no drivers for Windows 7 x64 on it, instead it redirects us back to the website to search for some.
The page for drivers for the 8000 is a big mess and currently looks like this:
I suspect more experienced users will not be too deterred by this, but the large number of links with old and new drivers available seems to suggest someone has problems making a reliable driver everyone is happy with. Furthermore, it seems the Linux support revolves around severely outdated kernels, and the patches themselves for video4linux may not actually be applicable anymore. Don’t expect any help, seems to be the message.
Unfortunately, the driver itself does not come with a driver certificate, and even disabling signature check will not allow this driver to load at all, instead continually giving a Code 52 error.
Being a little smarter, I would assume that I might just try with Version 188.8.131.52 Windows 8 driver, since it is closely related to Windows 7. After downloading and decompressing the drivers, it seems that they were written all the way back in June 2013, and development of the driver has not happened since then.
Sadly, this driver was also refused by Windows 7. This time, it complains that it may be corrupted or missing. This could be due to specific authoring of the driver which relies on Windows 8 only interfaces.
This has been rather frustrating. My next attempt was to install Version 184.108.40.206, but it still failed to initialize correctly when I had two tuner cards in the system, with an exclamation point and the same error message.
Success was finally found with driver installation on installing an earlier Version 220.127.116.11 driver. The driver is actually signed by Shenzhen Turbosight Technology Co. Ltd. and is dated 27th June 2011, which makes it very old.
Now that the drivers are installed without any warning symbols, you would expect everything to work just fine, right? Sadly, this is not the case. With two tuners installed, I could never get DiSEqC commands to work, as they would come out of the wrong card or not at all. Once I had eliminated DiSEqC switches from the system, I was capable of doing Crazyscan scans on just one card at a time, otherwise I would get an RFScan error. Despite turning up valid spectrum graphs, and SNR figures, when actually asked to stream the transport stream, the card returns no data whatsoever.
This did not just affect Crazyscan – even ProgDVB was unable to utilize even one card in a two-card system. The card would scan and show a lock but since there was no data on the channel, it would conclude with zero channels.
This driver is slightly more recent, from 19th April 2012 (no prize for guessing why that date is significant for me). I hope that running this driver would give me the best experience possible with the one card running.
The PTG Engine version number hasn’t changed, nor has the fact it is not digitally signed. There is an interesting conflict in seeing the driver signed by TBS, but the driver itself is copyright Prof Tuners Group.
In truth, this points to a much bigger issue with the Prof Tuners Group (PTG) itself.
It’s not unknown that PTG and TBS have been trading shots over the past few years. There has been talk, since removed from forums, that PTG was started by a disgruntled TBS employee which stole their intellectual property including their code and signing certificates and went to start his own company.
I have no idea how true this is, as the card itself and box looks very similar to TBS packaging, and the PCB silkscreen marks and the included CD point to TBS involvement in the production of the product. Despite this, as talk of PTG cards increased, TBS may have felt some threat from the new company, and sought to increase the reputation of their products above that of PTG.
To that end, they had support forums and encouraged people to actively review their products. PTG saw this as a direct attack, and they hence announced they had nothing to hide, implying TBS were moderating their forums to remove negative sentiment about their products.
PTG company does not have a forum and we do not plan to organize it. We release our equipment on a basis of “It simply works, and that’s all!” principle. We have nothing to hide about our production and you are free to discuss it on any forum available for you. Meanwhile, we have no opportunity to permanently moderate the forum, leaving only positive messages and deleting the ones from unsatisfied clients which will always exist, disregarding the quality of the product.
Sadly, amongst the crossfire, it seems that PTG has not had any direct investment, with their News page devoid of any content since 2013, and stocks of their USB 7500 depleted but never replenished. PTG smells of an operation in wind-down mode, with no expertise to write and update drivers to address issues, while simultaneously trying to hide the fact that their products still have significant problems.
Of course, PTG may have been a victim of TBS, but regardless, the consumer is the one that has lost out in the battle, with lacklustre product support, and a lack of certainty of future compatibility.
Single Card Performance
In single tuner mode, the card is much more usable, as it actually sends back transport stream data when requested. As a Crazyscan user, I was able to verify Crazyscan support by doing a scan. DiSEqC was also verified to function correctly.
The amplitudes reported by the tuner tended to have some strong “ripples”, which are partly due to the LNB and impedance mismatches in lower quality RG6 cable. The actual signal amplitude itself is very believable, and shows a good signal-to-noise contrast, unlike the TBS 6925 which might show about 4-5dB between peak and trough for a 10dB SNR signal. This suggests the TBS 6925 might have a non-linear front-end amplifier, and the Prof 8000 doesn’t. The reported SNR values from both tuners are very similar, as expected, as they are based on the same series of demodulator.
In blind scan performance, signal locking was not as reliable as the TBS 6925 in my experience. Some transponders took multiple attempts to blind-scan lock, and other times, more difficult transponders would not lock at all. However, the constellation plot does differ to that of the TBS 6925, without any “filtering”. The TBS 6925 would often give plots of noise clustered around the middle, whereas the Prof 8000 gives one with noise even out to the extreme edges, forming a “square”. I suppose this could be due to a different configuration of the demodulator between drivers.
In some cases, the lack of a lock may be because of ACM/VCM being in use, and this was more clear where the tuner had achieved a phase lock, by looking at the resulting constellation diagram. The tuners’ signal to noise ratio will jump between -10dB, 0dB and the “actual” value.
In general, broadcast services using DVB-S and DVB-S2 were more reliably locked, and plots were easily generated. The problem was that the tuner does not always return a proper TS stream and will in about >20% of the initializations, return garbage.
This results in TSReader complaining of Sync losses, because the stream itself doesn’t seem to be anything but random garbage. This suggests the driver has problems in returning the data possibly because of synchronization issues. Changing tuners in Crazyscan to Default Device and or changing back to the Prof 8000 causes a re-initialization which sometimes cause the tuner to return sensible TS data. Other times, a full restart is required to bring the tuner back to life.
The discontinuation of the TBS 6925 was a shock to me, and a big disappointment. The Prof Revolution S2 8000 was claimed to be a tuner that “just works”. What I found could not be further from the truth, and the user experience was disappointing.
Despite the Prof 8000 being claimed as a good “alternative” for Crazyscan blind-scan enthusiasts, my experience from having finally bought a pair of cards seems to show the Prof 8000 being significantly behind that of the TBS 6925 in terms of its driver support, and blind-scan locking performance.
The constellation display and RF level is, however, more accurate than the TBS 6925 with no obvious “filtering” and amplitude compression from a (possibly) non linear input amplifier. However, it is important to remember it lacks support for 16APSK, 32APSK and ACM/VCM streams. The design of the card in requiring so many molex plugs of additional power is just over-the-top and unnecessary inconvenience.
Because of the state of PTG as a company, and the lack of support and driver updates, it is hard to commend this card for purchase. The company appears to be in a state of winding-down, and future support for the card is hardly guaranteed.
For single card usage, it shows signs of provable instability in use, which is undesirable especially for “automated” recording or IPTV streaming purposes where reliability is key. In interactive usage, if you are willing to tolerate some unreliability, then it might be worth considering, but you should eliminate all other options first.
For now, it might just be best to stick to TBS products, if you want some hope of being supported. Those not needing Loop-Out will find the dual professional tuner card has a similar “per tuner” cost with the advantage of professional APSK/ACM/VCM mode support, so why pay for inferior support?