zxMarce

Zak, Well, who would have guessed? The Braasr Brasvision actually is a Dynacom. It puzzled me that on a corner of the the solder side there was a big blob of red paint, where the modulator "was" (I stripped it of its parts on my machine and will add the components for A/V there). The red paint covered the word DYNACOM!!! Also, the big electrolytic filter cap C27 is actually not connected to anything! Since the RF Channel Select switch is not populated, the negative pin of the cap is not connected to GND, and therefore it just sits there doing... nothing. I think Brasar/Dynacom wanted to save a penny by not populating the switch but they did not realize the cap was left "floating in the breeze" (saved one penny but not two, hehehe). Will implement your video out circuit and I'll also try to make it emit stereo sound, as the TIA puts audio in pins 12 and 13. Regards, zxMarce.

Zak, Read somewhere that the NTSC TIA's Chroma Out (pin 9) goes to positive with some kind of pullup, then it has a small coupling cap and after the cap another pullup. The article confirms not getting decent chroma by connecting directly to pin 9; instead chroma should be derived from after the small coupling cap. I guess R24 is after said cap. That is why I would end up drawing a schematic of my mod but including some of the board components, so people looking for these "solder points" may correlate their circuits to mine even if the boards are different. But not this week, we're getting the worse of a heat wave and blackouts are common. Regards, zxMarce.

Zak, Paying closer attention to your board pic and mine, it seems to me that both the Dynacom and the Brasvision clones "share" the board. Noticed that you also removed SW1, the TV Mode (B/W - Color) selector. I assume you just jumped it fixed in the Color position (why would the original console have this switch? I guess those were "other/older times"). The PDF is OK, but I'd also include the TIA type and a couple of other bits of info so anyone with a maybe different board can understand the basics and adapt it to other consoles. It is what I was looking for in my original post. A more detailed pic of the board with the connection points for the mod wires clearly marked is a big help. By the way, did you include the isolation capacitor in the video output? Does not show in the PDF. Will try to do my own conversion and post the result. Thanks for your help and pointers! zxMarce.

Zak, Thanks a lot, good to know, and nice schematic. Will give it a twist and a test later. What worries me is the lack of DC separation between the +5V and the video output; this can result in some DC voltage offset in the output. The circuit seems a modified emitter-follower (a basic circuit normally used as buffer), the emitter voltage "follows" the voltage in the base (minus appx 0.6V). If there is DC offset in the base, it will appear in the emitter and will get to the TV/Monitor, possibly affecting the picture. I'd add an electrolytic cap (would try 22uF, 47uF or even 100uF at 16V) between the transistor's emitter and the 68 ohm resistor to block any DC offset. Cap's positive side to the transistor (same position as shown in Audio Out) since the load (Video Out) goes through 75ohm to ground in the TV/Monitor. Regards, zxMarce.

Hi there. I've been gifted with a Brasar Brasvision console, an N-PAL clone of a 2600. It was a give away or throw away situation for the previous owner. Wherever I see it turns out, pretty much counterintuitively, that this is not a Brazilian machine, but an Argentine one. The site Compuclasico lists the console as manufactured by Argentine company Lakshmi S.A (now defunct), and the sticker in the bottom confirms this. Googling for BRASAR and or BRASVISION is useless... unless you mean to look for women bras, or Brazilian surveillance systems. I would like to make it spit A/V, but the board is not like any other in any tutorial I've seen, and to make matters worse, most YT tutes just say things like "Lift resistor Rxx and..." or "Remove transistor Tyy...", but give no detail as to what role the parts play or any hint that might help me map the target components to my board. I do not know if the component designators match the original 2600 schematic, so I'm at a loss here. The chips in the board were covered in some black marker or paint, but ordinary rubbing alcohol (not even isopropyl) removed that. The board has: R6532-RIOT, R6507-CPU, CO10444-TIA, and a standard TTL part, a 74LS74 - Double D-type Flip-Flop near the modulator. I think I saw "my" TIA type listed as NTSC here in AtariAge, but not sure. The console has a single 3.582056MHz crystal (N-PAL chroma burst for Argentina). Does anyone know of a "normal" 2600 schematic, but with the A/V mod already applied? If I could get one, I think I could map the mods made there to the ones I need to do in the Brasvision. Just for reference, a pic of the console's guts (yes, the DE-9 connectors are in the back of the board and... riveted): Thanks, zxMarce.

Please take the following with a grain of salt; I did not check the TI's schematic; I'm talking video in general, as a TV repairman in the '80s would. Actually, there's more to video than just 50/60 and RGB - Composite - YPbPr... In the race for Colour TV, several standards were developed (we're talking post WWII here). We'll focus on NTSC (USA, Japan, etc) and PAL (most of Europe, France being an exception with SECAM). To encode chrominance (chroma, or just colour) on the luminance (luma, luma+sync, brightness, or just Y) signal, a subcarrier is added and the chroma signal is sent as a phase-modulation of this subcarrier "on top". The luma (Y) signal alone comprises "ye olde" brightness and horizontal/vertical sync signals originally sent to B/W only TVs. Incidentally, this is why most of the time something is not right, you just get B/W: Luma+Sync (Y) is the foundation of TV broadcast. Colour comes... "afterwards". NTSC (strictly speaking NTSC-N) uses a subcarrier of 3.579545 MHz (3.58 for short) while European PAL (PAL-B) uses 4.43361875 MHz (4.43 for short). This is a most important bit to know. The various TMS99x8 chips encode the video differently, and the composite (NTSC, TMS9918) should be sent modulating a 3.59MHz, while the luma and chroma-difference signals (PAL, TMS9929) should modulate 4.43MHz. Any mixup will result in B/W display because the TV cannot lock its internal chroma oscillator to the input signal. And some noise may be seen in the B/W picture as the chroma subcarrier would not be filtered (dot-crawl is an example). Also, PAL will phase-invert the RED component every scan line to fix -and make redundant- NTSC's hue shift (and its knob). Again, this is for the dated CRT TVs available back in the day - With modern equipment, it is actually up to the TV model to say if the signal is "acceptable" or not. That is also why "some" modern TVs work and "some others" do not. To add problems to trouble, the NTSC chip TMS9918 uses a single pin to spit out just composite, while the PAL devices TMS9928/9 use three pins, one for luma (Y) and two for chroma-difference (Pb, Pr) signals. Regarding TI-99/4A modulators, the NTSC ones would just mix audio and composite, and amplitude-modulate the TV channel RF signal. An easy task. While PAL modulators had to work harder by creating composite out of Y-Pb-Pr, and mounting it on the TV channel RF signal. That's where the LM1889N entered the stage to do the heavy-lifting. All in all, I guess that if you just swap chips you will not have a proper picture. At a minimum, I'd say the chroma crystal may need to be changed. Also if you're going NTSC->PAL you need to route the difference signals Pb and Pr to the output. Hope that sheds some light, zxMarce.

For the ones after it, this is the same KiCAD 5 schematic (corrected an omission of two caps' references and values), but including the italian-like video out modification. Hope it serves someone. Regards, zxMarce. Modulator N-PAL - Video Out.sch

Here are some pics I took today of my modded N-PAL modulator, PHA-2031. The last pic is the image as I see it on a 32" Philips TV via A/V Input. Sometimes better, sometimes worse. The modification I made -I realized later- is one from the TI99IUC. Only changes I made was to replace the 68 Ohm resistor with a 75 Ohm one by connecting two 150 Ohm resistors in parallel (did not have any 68 Ohm at hand), and I used a SS9018 transistor (from a disassembled Timex/Sinclair 1000 RF modulator) instead of a BC547. I also connected my transistor between LM1889N's pins 13 and 14, given that pins 14 and 16 are both positive supplies. I suspect something in my chroma subcircuit is dodgy because whenever I touch some component leads (mostly anything on LM1889N's pin 15, audio subcarrier input) I can make the circuit output color (noisy of course, but color), while lately I mostly get noise-free B/W. Maybe my 3.582056MHz crystal' s gone; I played a bit with its tuning trimmer without result. For the eagle-eyed viewing my pictures, I also changed the non-shielded cable to shielded in the video mod; this I did after taking the photos. Hope this helps someone; as I said, I'll do a second schematic with the modifications. The modulator front: The back: The side with the A/V RCAs: The guts: Detail of the TV/Computer switching relay: The horrible mess - I added the multi way connector to be able to separate the board from the input cable (and forgot to include my own outputs - oh well): Detail of the multi way connector: Detail of the video modification: Same, but from the other side: The... er... "picture" I get from my TV: Regards, zxMarce.

Well... Got most of the original circuit for the PHA-2031 N-PAL modulator on a KiCAD 5 schematic. Attached. The circuit does not include the LM1889's RF parts (for the TV channels), as I only wanted to get composite out of the device. Basically, these RF parts are a L-C tank for the desired TV channel and that's pretty much it. Maybe an extra decoupling cap. I'll make a second schematic with my modifications. Regards, zxMarce Modulator N-PAL Schematic (KiCAD 5).7z

Hi there, I just signed up for some TI goodness. A brief history, off the top of my head - Actual facts may slightly differ. The first home computer I saw was a TI-99/4A with its PEB at a computer dealer in the early '80s. It was being used by a young person, and before my eyes a Parsec-style game slowly started to appear - the person was coding the game right there, live! At the time, Texas Instruments did have some HQs here in Argentina (guess it still does); they not only had representation but they actually did manufacture transistors here (sadly, this is no more so). It appears that "our" TIs had a national modulator (PHA-2031) designed and manufactured locally for our own PAL-N systems (50Hz vertical, 15.625kHz horizontal, 3.582056MHz chroma burst, 4.5MHz audio subcarrier). Not the same as the European (PAL-B, as we know them here) ones. Googled hi and lo, but could not get a schematic of it, so I got my hands dirty. I have a very rough and somewhat incomplete schematic of the beast (pencil!). I did not get to the RF part, basically because I just wanted to get to the A/V signal, and the RF part is just an L-C tank for the LM1889N it uses plus some support passives. I did get some A/V out of the modulator, albeit video is slightly noisy (maybe a cap or two need replacing, be it in the modulator and/or the TI power supply itself). I have the modulator board scan, a Pinta project with board's track and pad layers (flipped, so I could just pencil-draw over it the components by copying their positions from the board), and the handmade -partial- schematic of the modulator. I did not make a second schematic with my modifications yet; I will do it and upload the whole bunch later to this thread. My hopes are that someone sees what I did and yells me something along the lines of "Hey, if you get video this way [.....] it will be clean" or even "change caps Cxx, Cyy [...] and video will work". Regards, zxMarce.