Page 14 of 21

Re: RJ's Attempt. Maybe he's going to finish this time!

Posted: 06 Mar 2020 15:51
by rubber_jonnie
exxos wrote: 06 Mar 2020 14:54
rubber_jonnie wrote: 06 Mar 2020 14:44 I presume it's a case of ordering your own patch PCB, or do you and Icky have some we can buy?
@Icky Got a few made which he's going to send to me at some point soon... I'll probably just put them in the post for people in the UK who want them, is not worth bothering adding them to my store just for the sake of a couple of boards..

Icky paid for the boards to be done, so me selling them is slightly unfair to Icky. But also unfair on others if I force them to buy boards to fix something I ballsed up in the first place.

For people overseas, they can get them from OSH cheap with free postage anyway.. There was also no point people in the UK ordering from OSH as it will take over a month just to get the boards..
@exxos Cool, no rush for now though, got plenty else to fix on the board first.

Re: RJ's Attempt. Maybe he's going to finish this time!

Posted: 07 Mar 2020 12:05
by rubber_jonnie
OK folks, here we go!

:girldance: :girldance: :girldance: :girldance: :girldance:

Here we are, properly running using its own onboard ROM:

IMG_20200307_115349.jpg

For some reason the C input of the LS11 just wasn't making a connection to the pad, lord knows why, so it was bodge wire time.
I guess if I'd installed it earlier in the build, then I wouldn't have to rework now.

A 3 input AND doesn't work too well with one input missing it seems.

IMG_20200307_115655.jpg

Seems to pass basic diag cart checks like RAM, system timings etc.

Re: RJ's Attempt. Maybe he's going to finish this time!

Posted: 07 Mar 2020 13:09
by JezC
:goodpost: :cheers:
Success at last.. and all the sweeter for the delay.

Well done. Hope you get to enjoy your preferred beverage from the H4 mug when it arrives!

Any plans to post the Gembench figures for your H4?

Re: RJ's Attempt. Maybe he's going to finish this time!

Posted: 07 Mar 2020 13:10
by exxos
:bravo: :girldance: :girldance: :cheer: :cheer:

Re: RJ's Attempt. Maybe he's going to finish this time!

Posted: 07 Mar 2020 14:06
by rubber_jonnie
JezC wrote: 07 Mar 2020 13:09 :goodpost: :cheers:
Success at last.. and all the sweeter for the delay.

Well done. Hope you get to enjoy your preferred beverage from the H4 mug when it arrives!

Any plans to post the Gembench figures for your H4?
Looking forward to receiving my H4 themed beverage conveyance :)

I will run some Gembench tests as soon as I get the ROM sorted. I need to swapit out for a 1.02/1.04 ROM, as I think I dropped in a 2.06/1.04 by mistake. Going to look at that now and see what's what.

Tested so far:

Floppy drive
Video out (Unsurprisingly)
Serial port using diag cart
IDE interface
Reset circuit
Cartridge port
RAM
GLUE, MMU, CPU & MFP timings
Keyboard

But will keep at it. Incidentally I tried my PAK68/2 and it wouldn't work, CPU fitted or not. I need to do the DMA patch really and test ACSI too.

Only weird thing thus far has been this.

Boot up, no floppy fitted and use the Desk menu to pop up the TOS version, no problem, pops up with the Atari rainbow logo.

Connect the FDD and power up, boots fine, floppy all works fine, try to show the TOS version from the Desk menu, and I get bombs and a reboot. I need to work out how many bombs, but it's pretty quick, so will likely need to video it.

EDIT: Swapping out the ROM for a TOS 1.02/1.04 combo has stopped the TOS version problem from occurring, seems fine now. Better go order some 27c4096's as I seem to have run out.

Re: RJ's Attempt. Maybe he's going to finish this time!

Posted: 07 Mar 2020 14:26
by Icky
rubber_jonnie wrote: 07 Mar 2020 14:06 Looking forward to receiving my H4 themed beverage conveyance :)
It’s in the post got dropped off at the post office yesterday. :)

Re: RJ's Attempt. Maybe he's going to finish this time!

Posted: 07 Mar 2020 15:03
by rubber_jonnie
Icky wrote: 07 Mar 2020 14:26
rubber_jonnie wrote: 07 Mar 2020 14:06 Looking forward to receiving my H4 themed beverage conveyance :)
It’s in the post got dropped off at the post office yesterday. :)
Thank you Sir!

Re: RJ's Attempt. Maybe he's going to finish this time!

Posted: 07 Mar 2020 15:59
by exxos
rubber_jonnie wrote: 07 Mar 2020 12:05 For some reason the C input of the LS11 just wasn't making a connection to the pad, lord knows why, so it was bodge wire time.
I've had no end of odd soldering issues like that.. I half think to wipe the whole board with IPA before soldering.. But it doesn't seem to help. After talks and research, it seems the "hot air leveling" finish on the PCBs is a issue. Basically because the solder isn't perfectly flat. I mean the idea of tin plating PCBs is to make soldering easier, but it doesn't hold up with fine pitch SMT . so I am told, the gold plating on PCBs is the proper thing to have for SMT work, basically because its plated the finish is level. It only take a tiny bump on the PCB plating and several pins can run off. I think that could be one reason the ste boosters failed so much .

If there was another run of the boards, it may be better to use gold plated.. Only it would add about 30% onto the PCB price :( though if it increases soldering reliability , it may be with it. I do plan to replace the caps and resistors all with SMT.. In fact Icky has started doing just that. Idea being we may be able to get JLC to assemble the generic SMT stuff.. But with 10 H4s still in stock , and I doubt you guys want to do more H4 builds, I'm skeptical if will go ahead. Of course there's a lot of addons under development so its not like its the end of the road so to speak :)

Re: RJ's Attempt. Maybe he's going to finish this time!

Posted: 07 Mar 2020 17:22
by czietz
When hand-soldering 0.5 or 0.8 mm pitch I always go for ENIG PCB finish, precisely for that reason. It massively eases the buildup process. (I find it's not needed with fairly big stuff like 1.27 mm SOIC.) In contrast, our Lightning/Thunder/Storm cards are built semi-professionally with stencils, solder paste and a reflow oven and there Matthias tells me HASL is not a big issue. Probably because the unevenness melts in the oven and SMD ICs auto-center during reflow.

Re: RJ's Attempt. Maybe he's going to finish this time!

Posted: 07 Mar 2020 23:40
by rubber_jonnie
exxos wrote: 07 Mar 2020 15:59
rubber_jonnie wrote: 07 Mar 2020 12:05 For some reason the C input of the LS11 just wasn't making a connection to the pad, lord knows why, so it was bodge wire time.
But with 10 H4s still in stock , and I doubt you guys want to do more H4 builds, I'm skeptical if will go ahead. Of course there's a lot of addons under development so its not like its the end of the road so to speak :)
Well, potentially I might be in for another board, probably end of March or early April, my main problem is that I'm waiting for my tax code to be unf**ked by HMRC, so need to watch what I spend this next month. I've had 5 different codes since our company was acquired last September, so not a happy bunny. I'll get code number 6 on 6th April, fingers crossed that'll be it.

My other problem is lack of donor boards at the moment. I can probably come up with most chips, but I'm short of GLUE and PLCC CPUs, but I would like to try and get an H4 built without all the cock ups from this one. At least it works now, which is the important thing.