Love this series! - I remember seeing these DAI computers in magazine ads back in the day. It looked like a very capable machine!
@dannybloe
Жыл бұрын
Loved this episode. Learned a lot! Thanks
@Breakfast_of_Champions
8 ай бұрын
The wild, untamed and astonishing jungles of 1970s power supplies' land😀
@kausthubhk735
Жыл бұрын
Thanks for your great videos on restoration. I love your videos.I am waiting for the 5th part.👍🙏
@mwaddams
Жыл бұрын
44:19 Interesting fan, telling us to take it easy :)
@mwaddams
Жыл бұрын
Sorry to reply to my own comment, but Bernardo, did you know of this? Did you see it? I want one bad :)
@ColinChristie1
Жыл бұрын
Hi Bernardo. So cool to see the PS circuit diagram featuring the SG3524. I headed Silicon General's manufacturing in the 80s including our wafer fab in So. Calif (Garden Grove) where this regulating PWM was built and the assembly operations in the Philippines where the die was packaged. Some trivia you and your viewers might enjoy. The SG1524/2524/3524 series was one of our highest volume products in the '80s. (The 1524 is the mil-spec version with operating temperatures of -55C to +125C, the 2524 is for -25C to +85C, and the 3524 is the standard version for 0C to 70C). This product was original built on our Bipolar "40-volt Process" on 3" wafers on proximity aligner lithography (negative resist) before we upgraded the fab to 4" wafers and projection alignment lithography (Perkin Elmer 140). Today's IC's are built primarily on 12-inch wafers! In the 90s, Silicon General rebranded to Linfinity Microelectronics. In 1999 Microsemi aquired the company, and then in 2018 Microchip Technology acquired Microsemi. The wafer fab in Garden Grove continued to operate until it was closed around 2016.
@thebyteattic
Жыл бұрын
Fascinating! Thank you!
@tajtrlik1111
Жыл бұрын
Hello Bernardo, this is my first comment on this channel, im from Slovakia and I don't usually write comments on English speaking channels because I don't know enough English to write a meaningful comment, I had this comment translated by Google Translator. I am writing because this seems important to me. If the transformer you now have in that DAI computer gives a peak value of 24V between the center and the end of the winding (both windings, because they are the same), (so the effective value of the voltage between the center and the end of the winding should be something around 17V), then fault is not in the transformer according to me, because the filter capacitor will charge to a peak value minus the loss on the rectifier and the ripple will cause the rectified voltage to drop by only units of volts, so the rectified voltage in this case should be around 20V and that should be enough for the correct operation of the voltage regulators. Namely, the formula you showed (in time 42:58), which determines the size of the rectified voltage, applies to the average value of the rippled DC voltage without the filter capacitor connected, but in this case the filter capacitor is there, so the resulting voltage will approximately correspond to the peak one (minus the loss on the rectifier and the ripple). If it is much smaller and very wavy, then maybe the 4700uF capacitor behind the rectifier is bad. Try using a multimeter to measure the AC voltage between the center and the end of the winding (first one, then the other end) of the transformer, it should be around 16-17V effective (that is, around 24V peak, but multimeters measure the effective value), only that I would have misunderstood and the transformer it gives the 17V effective between the ends of the windings, and between the center and the end there would then be only 8.5V effective (that is, 12V peak), then the transformer would really be wrong. But I definitely wouldn't put a transformer there that gives 36V peak between the center and the ends of the windings, because the filter capacitors would be charged to this value and that could be too much for the voltage regulators and they could be damaged. The transformer you need should have a secondary winding of 2x17V effective (that is, the 17V effective must be between the ends and the center of the winding), then it will have a peak voltage of 24V, to which the capacitor after the rectifier is charged minus the loss on the rectifier and some ripple on top of that, so the rectified and smoothed voltage will be around 20V, as it should be. As I said before, I wrote this comment because I think it's important and no one has pointed this out yet, otherwise I really like your videos, I've seen all of them on your channel and gave a like to each one, I just decided to write a comment only now because of my poor English (Google Translator used and correced on my own).
@thebyteattic
Жыл бұрын
Thanks for the elaborate comment. The new transformer, with parameters calculated in this video, does solve the ripple on the linearly regulated 12V, as expected.
@tajtrlik1111
Жыл бұрын
@@thebyteattic OK, thanks for reply
@spacedock873
Жыл бұрын
Very good basic explanation of how a switched mode PSU works. It was helped by the fact that this is a very basic PSU! I have to say that you rather embarrassed yourself by criticising the designers of the PSU, saying that they didn't know what they were doing, and then introduced two problems with your two modifications! 😳 I personally would have run the PSU with a dummy load after the rebuild to test it in isolation before risking connecting it to the rest of the main board. Having said all that, this is a very interesting series and I am looking forward to the next video 👍
@thebyteattic
Жыл бұрын
I'm used to embarrassing myself ;-). But the PSU's design is still pretty embarrassing too ;-)
@spacedock873
Жыл бұрын
@@thebyteattic 😄👍
@johnsonlam
Жыл бұрын
I assume they want to keep the circuit simplicity, also keep cost down, when transformer output current is not large enough to blow things up, nowadays I think "not to regret" better than saving the few dollars. Also thanks for sharing the trouble shooting experience since it's mistake everyone may made, really good real life experience!
@lennartbenschop656
Жыл бұрын
At this point I would visit the Home Computer museum, the DAI computer they have there and check the secondary voltage of its transformer. Instead of guessing it.
@thebyteattic
Жыл бұрын
This is a pretty safe guess in a simple PSU like this ;-). I think the DAIs at the museum are defective anyway, not sure Bart would risk turning them on. But then again, it isn't needed.
@ovalteen4404
11 ай бұрын
According to the datasheet, 33K compensation is a far saner value than 220. So I'd weigh differences between the reverse-engineered schematic and your supply in favor of what's in your supply. If the timing components match the schematic, the supply should be switching at around 30KHz.
Пікірлер: 22