exxos's DFB1 trials

[exxos]

At work now, but I dont know if they're sufficiently visible in one of my old videos?

Cannot make out the values but I can see one of those 330K is missing. I think these are intended for calibration resistors. Problem being they only seem to alter the voltage 0.02V making them rather pointless anyway. :roll:

Capture.PNG

[Badwolf]

Cannot make out the values but I can see one of those 330K is missing. I think these are intended for calibration resistors. Problem being they only seem to alter the voltage 0.02V making them rather pointless anyway. :roll:

Yeah, can't see a damned thing in that vid.

Will have a proper nosey later on.

BW

EDIT: but I think, ultimately, if the user isn't pushing 5V to the board, they'll have to downgrade their speeds or upgrade their PSU.

[dml]

FWIW I have a Falcon here with an Atari (but recapped) PSU and with R18/R19 = 330k.

The voltage is reading 4.96v across the 5v rail.

[exxos]

EDIT: but I think, ultimately, if the user isn't pushing 5V to the board, they'll have to downgrade their speeds or upgrade their PSU.

Certainly looking that way. We have a recapped PSU. a lot of people out there probably don't. So there regulation is going to be even worse they could be bouncing between 4-5V continuously which is only going to aggravate the problems.

Reminds me a lot of the ST PSU's only output 4.80V a lot of the time. It gives the impression the machine is more efficient because of the lower voltage. The power supplies are generally terrible to start with anyway.

But I have found that some ST's even start to malfunction when you up the voltage to the proper 5V type of levels. :roll: I did not really look into it but likely all the ringing on the signals gets progressively worse with higher voltages. Which again leads you want to doing more mandatory fixes.

These machines really are evil on all aspects and levels :roll: I don't think the Falcon suffers in such a way. The Falcons I have worked on pales in comparison to the ST's anyway.

[exxos]

FWIW I have a Falcon here with an Atari (but recapped) PSU and with R18/R19 = 330k.

The voltage is reading 4.96v across the 5v rail.

Thanks.

So these things are basically going to be all over the place then :roll:

I will see if I can design a little mod PCB so the voltage can actually be altered higher or lower.

[exxos]

This is the draft circuit for a Falcon power supply voltage adjuster.

Capture.PNG

IMG_0292.JPG

330K does basically nothing so having 220K / 2 = 110K + 22K = 132K per branch. As 100K only alters about 0.1V, 132K is a reasonable value anyway. Aside from that, it is a voltage divider. So the preset at midway will not have any effect anyway.

Adjusting the preset one way or the other will rise or lower the 5V level. The 22K are there is basically " voltage ends stops". I just guesstimated that 22K will probably end up somewhere around 0.5V adjustment. Cannot be bothered to work it out properly. But these resistors will be through-hole parts and easily changeable anyway.

And they also look into changing the diode pack. As when I did this on the SR98 it improved voltage regulation and efficiency.

If anyone is interested then scroll down to DIODE UPDATE section.
https://exxosforum.co.uk/atari/last/psu/index.htm#SR98

[exxos]

Took out the one in the Falcon power supply. Ironically seems to be a reasonable one and likely better than the one I was using in the SR98 anyway :roll:

I will have a look and see if I can find anything any better as it has been awhile since I last looked.

IMG_0306.JPG

IMG_0305.JPG

STPS10H100CFP.pdf

Shindengen-S10SC4M.pdf

EDIT:

Looked through a few datasheets and a lot of them are missing internal capacitance off. There seems to be a huge variation in that respect on the ones I looked at. While there are some others which have lower voltage drops which may be better, the capacitance is higher. I am not wanting to get into all the particulars again with it all. But I think the diode pack which is in there is a pretty good one to start with. So will leave it at that I think.

[exxos]

And it goes a little something like this...

1.jpg

Capture.PNG

[Badwolf]

Right, no R18, like you noticed.

R20 and the one beneath look to be 5.1k, same as yours.

R19 looks like a 330k too.

So just the R18 different?

BW

IMG_6026.jpg

[exxos]

So just the R18 different?

Seems so. As R19 is a pull down, it would appear your power supply would have been slightly too high voltage and they cut R18 out to limit the voltage a bit.

But like I said earlier, only seems to alter 0.02V anyway. So it begs the question why they really bothered. It is basically down to the noise margin there anyway.

I see you had 5.014V On your multimeter test.. I guess if you slapped 100K for R18 it will probably drop it down to more my level of around 4.90V.

EDIT:

Maybe those calibration resistors were simply in the documentation when they were being tested ( assuming they actually was tested) during manufacture. The Muppets testing them probably had some documents stating if the voltage is too high then cut R18 etc. So they did exactly that, which did almost nothing and nobody ever questioned it :lol: