The Beast

[Harry Dalek]

You can call me John. I am younger than most of the forum users I think and most younger internet users tend to use real names less to an extent. Yeah it helped me too... there was a notable lack of available software for that purpose with fully customizable line and frame rates. Still want to make a viewer some day but motivation and time are fleeting right now.

Well thank you John its really good Software You might out live us all ! carry on the NBTV dream ; ) ,i have yet to use every thing it can do adjust ,every thing works great i was infact just using it which i will show the results after i reply here .
Yes a multi system viewer would be great but being able to test on standard and off standard video is pretty useful to me ,any one making a monitor would be foolish not to make use of this .

Both of those ports dying though. USB is the only standard I see sticking around for a while.

That's True useful if you have a older pc laptop laying around so i understand the USB idea .

Maybe I'll take it on myself one day.

i was looking into it a little this morning getting my head around what's about and which way to go if possible .

That might work. I do wonder what the practical bandwidth is. 384KHz is way above anything humans can hear. While the internal DAC may very well do it, I wonder how well the analog side of the circuit actually handles those frequencies in practice. Only way to find out would be for someone to be brave enough to risk buying one and scoping it.

[/quote][/quote]

So these USB DAC sound card thingies could work outputting a wider bandwidth so looking at one with the highest sampling rate what we are looking for ...Bit of research is a good way to start as in what about costs possible to buy just the chip is it worth the trouble ,Just noticed Steve has posted up on this also my understanding of this side of things is very little so i would have to ask Questions .
I have seen a fellow make a USBDAC sound card if that the right name don't think the thing was that much better than an internal sound device ,self detecting when plugged in so i don't think he had to use write software for it i will post it up under here when i track it down after the posts today , any case interesting .

[Harry Dalek]

Yes, USB. It's wonderful, except when you're trying to use it for a minority application as per here. There are 'break-out cables' available from the likes of FTDI, which I have used to turn a USB port into a 5 or 3.3V UART. They work well but only go up to 3Mb/s, 300k bytes/sec. In a lot of cases that's plenty enough. The speed limitation is set by the chips used, FT231 and FT232...made by FTDI themselves. I don't think they've released a faster version, but I haven't checked recently.

Alternatively there's micro controllers with a built-in USB engine, e.g. PIC18F14K50 (20-pin DIL or SMD) which are quite capable of handling Full-Speed USB (12Mb/s) or 1.2M bytes/sec. You can obviously program them to produce whatever I/O you wish. 1.2Mbytes/s should be quite capable of handling a 500kHz bandwidth 8-bit D-A or A-D.

However, if you're even just glanced at the programming of a USB port, you hope that someone, somewhere has done it already. Undoubtedly they have, but they seem to keep the code to themselves - in a way I don't blame them, a lot of effort must have gone into it. I felt myself literally wilt as I started wading through the USB spec. It's not for the faint-hearted! As you may have gathered - I gave up!

Steve A.

Its does sound very hard complex a thing to make if you have had trouble !
I was wondering if it could be a off the self gadget you could buy ,a USB DAC sound device with a wider band width .My understanding here is very very low seeing there are these things built that do out perform your PC's sound card , so would it outputed a video within its higher bandwidth is that all we would need i am more than likely missing something .
Just wondering if you have looked what's about that could do the job at least do better than what our PC sound device does ?

[Steve Anderson]

I've just toured around the Microchip (PICs) website, and there does seem a lot of tools available now to get you going with USB. When I'm in the right frame of mind (probably inebriated) I'll have a go...

Steve A.

[Harry Dalek]

Looking into improving yesterdays look at syncing with a pot at various line rates i did build a multi position switch with 5 positions with resistance's from yesterdays tests to give me that syncing range with a switch position this time ,that works bottom photo ,so put that aside and looked into displaying lower frame rates higher line rates in sync .
Below are the tests i did today Slow NBTV with a frame rate of 1.5hz and 240 lines syncs works well (did have to replace my frame oscillator cap from 1uf to 4.7uf for the test )tomorrow i will look into what it can and can't do on this side of things .
Kitten Koala Testcard ; ) below

[Harry Dalek]

I've just toured around the Microchip (PICs) website, and there does seem a lot of tools available now to get you going with USB. When I'm in the right frame of mind (probably inebriated) I'll have a go...

Steve A.

OH thank you Steve , i enjoy testing these monitors it s a little frustrating not having a little more bandwidth to display stuff it would be good to see if displaying any thing above 64 line is worth it with a PC at NBTV frame rates at least ,as in the post above my other idea is just dropping the frame rate least on a video still testcard it shows what could be .

[Steve Anderson]

OK, let's see what's needed - to an approximation...
64 lines with an aspect ratio of 1.5:1 = 96 pixels say. So a full frame = 6,144 pixels. At 12.5 frames/sec = 76,800 pixels/sec. The bandwidth required will be half that figure, 38,400Hz, i.e. 38.4kHz. That's pushing it even on some of the best sound-cards, those with 96kHz sampling. Plus you still have the LF AC-coupling headache to deal with, below 20Hz or so. (Note I'm ignoring sync-time here, it is an approximation).

So, let's have a look at the FTDI 'cables', as previously. They can go up to 3Mbits/sec, a byte is 8 bits, but when used as a UART (e.g. RS232) you have to add one start bit and one stop bit per byte, a total of 10 bits per byte. That becomes 300kBytes/sec, far more than the 76,800 pixels (8-bit samples/bytes) per second needed. So they're suitable.

The next choice you need to make is the baud rate for the UART side of things. In theory a PC can use any baud rate that you wish (within limits), even non-standard ones. However it's plain sense to stick with a standard baud rate. The next one up from 76,800 bytes/sec, or 768,000 baud is 921,600 baud or 92,160 bytes/sec., so we have 20% 'spare' for sync-time. 921,600 baud is generally as fast as most PC terminal software will run, though maybe not Hyperterminal which is useless for this anyway. I use RealTerm as a Terminal program (free download), others like TerraTerm (also free) may well be suitable too, but I've never used it.

Using a terminal program is not as easy as using Audacity or other audio record/playback software, it takes a bit of getting used to, but it's not that difficult. The SSTV wav files on this forum (somewhere) were created with RealTerm, but obviously at a much slower rate...

Steve A.

P.S. Someone check my math...I think it's OK though!

Below is how simple it is to interface one of these 'cables' to a micro...the 'cable' is powered from the USB source, in this case your PC. It can also supply 5V at up to 250mA to your project if needed...though not used in the case below...the transistors shown are not 'naked', they have inbuilt base resistors and base-emitter resistors...so easy to use...

PC NBTV Interface 2C.gif (7.25 KiB) Viewed 6758 times

KRC241-246.PDF

(65.49 KiB) Downloaded 210 times

[Steve Anderson]

...and with the above arrangement all the LF problems go away, it's DC-coupled...0.01Hz, no problem! Not a single capacitor in the signal path...except those constraining the HF response...

Steve A.

[Harry Dalek]

The USB part of it does look simple wiring wise but i suppose as your post mentions in detail its complex to feed it correctly for our needs .
So far i Know how to wire up a USB this a start .
I think your idea would be to feed a wav file from the pc to something like the SSTV project idea we were looking at a while back speculating it would output the video at correct bandwidth ...i am not a computer digital person so got be slow with me here

[smeezekitty]

Keep an eye on the "pixel clock" field. If its less than your soundcard sample rate, you should be ok. If you're sampling at 48KHz, 3.5FPS should be possible at 240 lines

[Harry Dalek]

Keep an eye on the "pixel clock" field. If its less than your soundcard sample rate, you should be ok. If you're sampling at 48KHz, 3.5FPS should be possible at 240 lines

Hi john
My sound card says it can do HD audio 96k which i set it at ,i can never tell if that is True or not i can't think of a way to test it apart from outputting higher NBTV rates , I am sure 3.5 FS is possible i went with the maximum with 240 line 1.5 frame for syncing and image quality ,also i have noticed even number framing number such as 2 frames will not lock perhaps 2.5 or as you say 3.5 will ,i will experiment tomorrow ...we are in Lock down here in Melbourne so i have a bit of time on my hands not being able to go to work so i will work on this

[Steve Anderson]

It's the old case of 'buyer beware'. Just because it says it samples at 96kHz doesn't mean it can record up to 48kHz. All A-Ds should/must have an anti-aliasing filter on the input. How good that filter is is anyone's guess, it's rarely if ever specified. An analogue version isn't going to have the 'brick-wall' response ideally required, well, at a sane cost at least. So mostly this is done in software, an FIR filter being the most common and easiest version to implement. Even that isn't perfect, but somewhat better.

So for most consumer 'sound cards', internal or external to the PC, assume 20kHz and no more. There is no reason when used for sound to go beyond that, it means they can sample at 96kHz, sure, but the filter specifications can be relaxed costing less in hardware. It means that you end up paying for the lack of a decent filter in needless file sizes. Remember an audio CD, which is good enough for me, is only sampled at 44.1kHz. But it only needs sampling once for CD mastering, so they can afford to throw a lot of cash at that filter.

The only way to prove or otherwise the response of a 'sound card' is with an audio oscillator[1] and an oscilloscope as a minimum...you might be horrified with what you find!

Steve A.

[1]..that goes up to 100kHz as a minimum...

[Harry Dalek]

It's the old case of 'buyer beware'. Just because it says it samples at 96kHz doesn't mean it can record up to 48kHz. All A-Ds should/must have an anti-aliasing filter on the input. How good that filter is is anyone's guess, it's rarely if ever specified. An analogue version isn't going to have the 'brick-wall' response ideally required, well, at a sane cost at least. So mostly this is done in software, an FIR filter being the most common and easiest version to implement. Even that isn't perfect, but somewhat better.

OH i Agree i can set all the window sound device to this and the programs i use to this setting but as far as knowing if the sound device can do it i have my doubts ,i am not so sure i could produce say a frequency sweep wide enough to output to a scope and test ,every thing i have seen for this is up to 20khz only .

If you can think of an easy way to know for sure would be a useful test .

So for most consumer 'sound cards', internal or external to the PC, assume 20kHz and no more. There is no reason when used for sound to go beyond that, it means they can sample at 96kHz, sure, but the filter specifications can be relaxed costing less in hardware. It means that you end up paying for the lack of a decent filter in needless file sizes. Remember an audio CD, which is good enough for me, is only sampled at 44.1kHz. But it only needs sampling once for CD mastering, so they can afford to throw a lot of cash at that filter.

So even a external sound device is just going to stuck at the low 20 KHZ ..i did hope there might be a external sound device that had the bandwidth i just get confused with the sample rates s and such with these things all i know more is better for both .

The only way to prove or otherwise the response of a 'sound card' is with an audio oscillator[1] and an oscilloscope as a minimum...you might be horrified with what you find!

Steve A.

[1]..that goes up to 100kHz as a minimum...

[/quote]

You sort of answered in a way what was on my mind ! but you think it would out put up to 100Khz or mistaken ?

Feeding an oscillator to the pc sound card and seeing what coming out on my scope would be a good Test ,only problem i could see would be on the mics side the circuit is set the same and the output side i see not as easy to do in windows 10 as it is in windows 7 have to look into it .

I was think at first some thing like this below with a better chip might work


youtu.be/otrGbcWVfp0

http://users.abo.fi/jskata/JEDAC/

Hes unit based on the PCM2707C from Texas Instruments which takes care of the USB communication and outputs I2S audio data
the PCM1794A, a 132dB SNR 24-bit 192kHz DAC which receives I2S protocol
the OPA4134, a high performance audio operational amplifier

I think it would be cheaper just to buy a built sound usbdac unit be very hard to solder one of these things together

[Harry Dalek]

Hi had a look today John on the framing at 3.5Hz very flashy and results were not very good down to 100 line i would have to adjust my ramp oscillator cap is way off for this rate by the looks of it


Framing at low 2.5 and below 1 second still worked tried the lowest setting in freeNBTV and highest line rate but the problem is having to readjust circuit trimmers for size control and such as image becomes very large for the external controls which go out of range .

So low wise keeping my monitor as is 1.5 to 2.5 HZ framing seems better on results low end with out have to adjust every thing out of range

[Steve Anderson]

The 'Audio Soundcard' above looks OK, but without downloading the datasheets (yet) I can't find any reference to sample rates or other specifications...unless that's in the soundtrack with the video (I haven't played it yet).

It's also just the output side, i.e. playback only, no record function, unless I've missed that as well...entirely possible!! Of course the hardest part is the record function...

Steve A.

In your added text 192kHz sampling is mentioned which in theory can record up to 98kHz, assuming it can record. Again that depends on the input anti-aliasing filter (if there is one)....and again you end up with bloated file sizes.

Having had a quick look at the first datasheet for the PCM270xC on the very first page it says, "Sampling Rates: 32 kHz, 44.1 kHz, and 48 kHz" No mention of 96 or 192kHz anywhere that I can see...

Now this is becoming confusing...why then is there a second D-A that can go at 192kHz? (the PCM1794A). Maybe I'll come back to this once the coffee kicks in...

OK, the caffeine has done its job...but I'm still confused! No problem, the fact is the PCM1794A can go at 192kHz and the device is playback only. Now where did this all start?...

FYI The PCM1794A is around US$16++ each at Digikey. Not only can it handle 192kHz but is 24-bit.

[Steve Anderson]

Going back a few steps, I mentioned an 'up to' 100kHz oscillator so you can confirm or otherwise a sound card indeed has that input filter required.

Ideally as you sweep up the input frequency toward half the sample frequency the playback output should drop very rapidly - really very rapidly!

At a sample frequency of 48kHz as the input frequency gets near 24kHz the output should vanish. If you go beyond 24kHz there should be zero, absolutely nothing. With a poor or non-existent filter with an input frequency of 25kHz - you'll get 1kHz out! At who knows what level - hardly useful! With 26kHz input - 2kHz out, and so on...

So using a sampling rate of 192kHz the response should be up to (in theory again) 96kHz, hence the mention of a 100kHz oscillator...

Steve A.

You may question where are those sorts of frequencies going to come from? In music, percussion instruments, e.g. drums have sharp rising edges, particularly snares, and of course cymbals - full of high frequencies, even the innocent tamborine. In much the same manner the sharp rising and falling edges of logic contain frequencies that can alias, so they too need filtering. Think of sync pulses for example...

Another way to ease the burden of the anti-aliasing filter of by over-sampling by a factor and reducing the sample rate rate in software.