STE V1.5 32MHz BOOSTER - STATUS

[exxos]

I've updated the V1.5 booster info which is now the 36MHz booster...


https://www.exxosforum.co.uk/forum/viewt ... =24&t=1018

[PhilC]

Well done. I can imagine you are relieved to now have the booster stable enough to sell.

Still, makes those three 40mhz ones even more valuable doesn't it... :lol:

[exxos]

Well done. I can imagine you are relieved to now have the booster stable enough to sell.

Still, makes those three 40mhz ones even more valuable doesn't it... :lol:

Indeed! I built up a second one, 36MHz... first run 73mins, second run 63mins.. this in "fast mode".. bizzare...

[PhilC]

Could it be something else not liking playing at 36mhz? rom chips getting too hot or something daft? or even the dreaded sockets?

The CPU pins in my STF are the hottest thing on the CPU when using the old STE cpu but cool to touch when using the one you supplied. Something similar perhaps?

[exxos]

Could it be something else not liking playing at 36mhz? rom chips getting too hot or something daft? or even the dreaded sockets?

The CPU pins in my STF are the hottest thing on the CPU when using the old STE cpu but cool to touch when using the one you supplied. Something similar perhaps?

I think its just timings... I mean take 55ns ROM.. 20ns goes on GLUE decoding.. at 32mhz you have 2x32mhz cycle before DTACK (64ns) so total ROM access speed is 75ns and the CPU is expecting data valid after 64ns...So the ROM is already pushed to the limits at 32MHz.. nevermind higher speeds...

BTW you guys who got the 40mhz one, which motherboard type was it, external blitter or integrated blitter ?

[PhilC]

It's this one

20181109_223658.jpg

And with the older Rom board before the red one if that helps.

[exxos]

It's this one

Thanks.. I have a feeling those external blitter types work better for some reason.. I need to do more tests, but if I keep messing about with these things they will never make it to my store :roll:

[PhilC]

You don't have one to test it in then?

I would volunteer but my experiences with sockets lately tells me to leave it alone :D

[exxos]

You don't have one to test it in then?

I do yes.. but then it will turn into another week of messing about with the thing... I need to catch up on other stuff as well yet...

[exxos]

It has been bugging me how it can run for 1-2 hours when "cold" then 30mins, then 10mins then 5mins...

So...

IMG_3501.JPG

I had this running last night (or should I say this morning ) for 146mins 45 loops... Which is getting on for 3 hours run time! I pretty much guarantee it would fail under a hour when first turned on, and run time after that is generally under 10 minutes.

I turned it on about an hour and a half ago this morning, is currently on loop 25, which is about an hour and a half run time..

My thought last night was to check the voltage on the ROM board to see if there was a voltage drop their considering is a fair distance away from the PSU.. My thought was as well longed inductance of the supply rails... But the voltage seemed fine, but then realised I don't think there is any actual bulk capacitance anywhere around the ROM circuit.

Now I do have a 1500uF capacitor bodged on a RAM capacitor.. But obviously this still isn't enough. The whole issue of it running initially when the machine is cold then failing quicker and quicker as the machine warms up, is pretty much tell-tale signs of bad electrolytic's (I really should have thought about this sooner! :roll: )

The only capacitor there is C204, but that is part of the RS232 supply generator circuit.. I cannot see any other electrolytic's anywhere around the area of the motherboard.. So it is a bit odd to say the least... But in any case there should be a bulk capacitor somewhere on the motherboard but it does not look like there is :roll:

The irony in all this, is the ROM datasheet does actually say use something like 100uF bulk capacitance, but I use something like 10uF ceramic, should have been fine, and has actually been fine up until this point. Though I actually wrongly assumed there was some bulk capacitance on the motherboard itself close by, but there actually isn't! :roll:

I can only really assume that as things heat up, they may well pull a few more mA per chip across the board, and it crosses tipping point after a while with the stability of the of the power rails.. giving the impression of a bad capacitor.. It is only a guess though..

So I will leave this board on over the weekend and see what happens.. When I get back tomorrow night , assuming it has run the whole time without crashing, I will see if I can hack on the ROM board a 100uF ceramic and then do the testing again..