Terrible Fire Accelerators - Page 187

hooleydog · 13 October 2018, 01:19

Quote:

Originally Posted by plasmab

Big round ones = 10uF
Regulator ones as per manual for regulator.
Everything else 100nF

Thanks so much, and don't stay up too late (Not good your your health)...

Oh, can the board run a bit faster than 20MHz....

hooleydog · 13 October 2018, 04:48

Hey Guy's I'm gonna do a soldering test using my solder pot. The 2 CLPD's are held in place using a small dab of epoxy (Like in a real SMD machine) Then I'll tape off the rest of the board off using cap tape.. Add flux to the chips and then slide the board over the solder in the pot (so the solder covers the chip) and see if it works (No shorts I hope). The temp of the pot will be no hotter than 270 degrees (so the epoxy won't melt, or the board getting burned). If it don't work then not a problem, as I have more than one board and spare CLPD's too..

But if it works.. I'll do the same to the ram chips (Quick and easy)...

Oh.. One of my jobs was to send the boards (Big and small) through a wave solder machine. All the SMD components are held in place with adhesive (so they don't fall off going through the wave solder).. Flux is sprayed under the board (Covering all the SMD and through hole components).. Then the board moves over the solder wave.. All the components get soldered, and the boards come out hot (Have to wear gloves).. But the components on the board can take the heat, and I've hardly seen a short on the boards (The boards then go to Quality control) ...

project23 · 13 October 2018, 05:23

Quote:

Originally Posted by hooleydog

Hey Guy's I'm gonna do a soldering test using my solder pot. The 2 CLPD's are held in place using a small dab of epoxy (Like in a real SMD machine) Then I'll tape off the rest of the board off using cap tape.. Add flux to the chips and then slide the board over the solder in the pot (so the solder covers the chip) and see if it works (No shorts I hope). The temp of the pot will be no hotter than 270 degrees (so the epoxy won't melt, or the board getting burned). If it don't work then not a problem, as I have more than one board and spare CLPD's too..

But if it works.. I'll do the same to the ram chips (Quick and easy)...

Oh.. One of my jobs was to send the boards (Big and small) through a wave solder machine. All the SMD components are held in place with adhesive (so they don't fall off going through the wave solder).. Flux is sprayed under the board (Covering all the SMD and through hole components).. Then the board moves over the solder wave.. All the components get soldered, and the boards come out hot (Have to wear gloves).. But the components on the board can take the heat, and I've hardly seen a short on the boards (The boards then go to Quality control) ...

You're doing all of this from home? If so I have crazy amounts of respect for what you're doing, and you should consider documenting. That's the kind of bat$hit insane thing I like to do. I've worked with epoxy before, I guess you'd just use a temperature controlled hotplate with enough solder molten in it, then place the card in on both side one after the other? Or am i mistaken?

Utterly fascinating. Would make soldering those tiny CPLD's an absolute doddle!

John

pgovotsos · 13 October 2018, 06:58

Quote:

Originally Posted by project23

You're doing all of this from home? If so I have crazy amounts of respect for what you're doing, and you should consider documenting. That's the kind of bat$hit insane thing I like to do. I've worked with epoxy before, I guess you'd just use a temperature controlled hotplate with enough solder molten in it, then place the card in on both side one after the other? Or am i mistaken?

Utterly fascinating. Would make soldering those tiny CPLD's an absolute doddle!

John

I do this using a spare furnace I used to ladle cast bullets. It won't take boards bigger than about 8" x 8" but with accurate thermostat and a quick dip of both sides it is easy to use. It works fine with the pitch of the Amiga chips and even a bit finer. Good epoxy and flux are your friends

With the tremor in my hands these days this works better for me. Just don't put so much epoxy under the components that it lifts them off the pads or covers the pads when you push down and it squeezes out.

plasmab · 13 October 2018, 10:54

I found this didnt work with RAM chips. the pins need physically pushed by the iron to make contact. Its also alot more work and as a few people found.. you end up with shorts where there the solder paste dribbles under the chip and shorts between vias.

Anyways good luck and let me know how it goes.

I never understood why people are so scared of just soldering with iron though. I honestly find soldering a CPLD on easier and less of a PITA than putting through hole resistors on board.

plasmab · 13 October 2018, 14:42

New tf534 alpha:

https://github.com/terriblefire/terr...0_13_alpha.zip

I tested this at 25Mhz. I can confirm the AAIB issue. In an effort to get a baseline the Ram is back in the ZII area. this is temp.

project23 · 13 October 2018, 14:55

Quote:

Originally Posted by plasmab

I never understood why people are so scared of just soldering with iron though. I honestly find soldering a CPLD on easier and less of a PITA than putting through hole resistors on board.

That's coz you've done a million of them!

It's true though, yesterday was only my 4th and it took a fraction of the time of the 1st CPLD I did.

hooleydog · 13 October 2018, 15:01

Myself.. I've hand soldered lots of SMD and through hole components. I even recapped my old A1200 plus my A2000 motherboard (Removing the old caps and sucking the solder from the holes), and changed my 500+ plus A2000 from NTSC to PAL (I don't like NTSC ).. But I've never soldered a CLPD (Where's my microscope at)....

Oh, to be safe.. I even replaced the old caps in my A2000 power supply too, and I added a a/c switch to the front of the 2000, so It's easier to turn it on/off instead of doing it at the back of the computer (Which I hated)...

plasmab · 13 October 2018, 15:27

Ok. I guess i really need to find out why AIBB is the only application with corrupt graphics.. Is the author on EAB? I wonder if i could get some sourcecode snipets to reproduce the issue.

Other applications do show this too. In the menus. Its just a bit random. Does anyone know how intuition renders the menus? I likely have some kind of cycle collision but i need to narrow down whats actually happening.

CRP · 14 October 2018, 01:04

Hi, just wanted to let you know that my TF530 is officially completed, up and running. Everything works as advertised. I have just burnt a dual kickstart eprom and am running both wb 1.3 and 3.1. Everything is speedy, ide interface works perfectly.

It has been a bumpy road, but thanks to Mr. Leary for designing this thing and making it available to everybody, and thanks for all the guys who have helped out in this forum. I literally started from zero, have learned a lot and had lots of fun!

Thanks!

plasmab · 14 October 2018, 01:27

Quote:

Originally Posted by CRP

Hi, just wanted to let you know that my TF530 is officially completed, up and running. Everything works as advertised. I have just burnt a dual kickstart eprom and am running both wb 1.3 and 3.1. Everything is speedy, ide interface works perfectly.

It has been a bumpy road, but thanks to Mr. Leary for designing this thing and making it available to everybody, and thanks for all the guys who have helped out in this forum. I literally started from zero, have learned a lot and had lots of fun!

Thanks!

Congrats.

In other news I hacked my SDG1025 up to a SDG1050 today.. so i can now go above 25Mhz on my signal generator.

alenppc · 14 October 2018, 01:47

Quote:

Originally Posted by plasmab

New tf534 alpha:

https://github.com/terriblefire/terr...0_13_alpha.zip

I tested this at 25Mhz. I can confirm the AAIB issue. In an effort to get a baseline the Ram is back in the ZII area. this is temp.

Flashed. I dropped the clock down to 36 Mhz now. IDE is working again. It seems that CPU Fastrom is also working again (gives the crazy IDE speeds once more).

I have an issue with the mouse pointer jumping a bit randomly around the screen (doesn't move smoothly and jumps back and forth).

As for the graphics corruption issue, I wish I knew where the author of AIBB is, since this is one of the most precise benchmark tools on the Amiga. It really would be great to hear from them.

All I can tell you is that the first public release of the TF530 firmware had the same issue. You'll also notice that the CPU clock is not being detected correctly by AIBB, issue which also occured in the original public release of the TF530 firmware.

Both issues were fixed on the subsequent update... so whatever you did then is what fixed it... sorry I am so vague...

Still talking about AIBB, now it shows a crazy low memory latency of 3.5 (TF328 with 0 wait state ram was 4). 14.19 mb/sec transfer rate. I think this is faster than 530...

plasmab · 14 October 2018, 01:57

Quote:

Originally Posted by alenppc

Flashed. I dropped the clock down to 36 Mhz now. IDE is working again. It seems that CPU Fastrom is also working again (gives the crazy IDE speeds once more).

I have an issue with the mouse pointer jumping a bit randomly around the screen (doesn't move smoothly and jumps back and forth).

As for the graphics corruption issue, I wish I knew where the author of AIBB is, since this is one of the most precise benchmark tools on the Amiga. It really would be great to hear from them.

All I can tell you is that the first public release of the TF530 firmware had the same issue. You'll also notice that the CPU clock is not being detected correctly by AIBB, issue which also occured in the original public release of the TF530 firmware.

Both issues were fixed on the subsequent update... so whatever you did then is what fixed it... sorry I am so vague...

Still talking about AIBB, now it shows a crazy low memory latency of 3.5 (TF328 with 0 wait state ram was 4). 14.19 mb/sec transfer rate. I think this is faster than 530...

excellent feedback.

Can you tell me if audio is ok? I have some issues but i suspect its my capture card. Simon the sorcerer was a bit odd with audio when i tried it.

I can put an alpha together with the same bus timing as the last TF530 release. I am pretty sure i've tried that already but i'll push it out for you to try tomorrow.

alenppc · 14 October 2018, 02:08

No audio issues whatsoever

project23 · 14 October 2018, 02:37

Quote:

Originally Posted by CRP

Hi, just wanted to let you know that my TF530 is officially completed, up and running. Everything works as advertised. I have just burnt a dual kickstart eprom and am running both wb 1.3 and 3.1. Everything is speedy, ide interface works perfectly.

It has been a bumpy road, but thanks to Mr. Leary for designing this thing and making it available to everybody, and thanks for all the guys who have helped out in this forum. I literally started from zero, have learned a lot and had lots of fun!

Thanks!

Well done dude, I hope to be close behind you. Had it programmed up, but socket problems, so had to start again (broke the board).

super_adam · 14 October 2018, 08:50

Quote:

Originally Posted by plasmab

Congrats.

In other news I hacked my SDG1025 up to a SDG1050 today.. so i can now go above 25Mhz on my signal generator.

I wish you you could hack it up to at least 1060

plasmab · 14 October 2018, 14:16

Quote:

Originally Posted by alenppc

No audio issues whatsoever

Can you test something for me...

Disable Data Caches and re-run AIBB. Issues go away for me.

Which has me worried because CIIN is asserted. Do the data caches ignore CIIN?

plasmab · 14 October 2018, 15:09

Ah. I was hitting this with AIBB.

http://eab.abime.net/showthread.php?t=56255

New alpha.

https://github.com/terriblefire/terr...0_14_alpha.zip

plasmab · 14 October 2018, 15:21

Beta time... ram back in ziii space.

https://github.com/terriblefire/terr...10_14_beta.zip

time to break it guys.

EDIT: While you guys do that i'll see if i can get cache burst going

plasmab · 14 October 2018, 15:58

Ok i think i understand the slowram issue now. Even though I dont have any.

If you have an 030/MMU? You will need a software patch to cache inhibit the $00Dxxxx area except where you have ram (and want it cached). The accelerator cannot do anything about this.

13 October 2018, 04:48	#3722
hooleydog Registered User Join Date: Aug 2017 Location: Southfield Posts: 53	Hey Guy's I'm gonna do a soldering test using my solder pot. The 2 CLPD's are held in place using a small dab of epoxy (Like in a real SMD machine) Then I'll tape off the rest of the board off using cap tape.. Add flux to the chips and then slide the board over the solder in the pot (so the solder covers the chip) and see if it works (No shorts I hope). The temp of the pot will be no hotter than 270 degrees (so the epoxy won't melt, or the board getting burned). If it don't work then not a problem, as I have more than one board and spare CLPD's too.. But if it works.. I'll do the same to the ram chips (Quick and easy)... Oh.. One of my jobs was to send the boards (Big and small) through a wave solder machine. All the SMD components are held in place with adhesive (so they don't fall off going through the wave solder).. Flux is sprayed under the board (Covering all the SMD and through hole components).. Then the board moves over the solder wave.. All the components get soldered, and the boards come out hot (Have to wear gloves).. But the components on the board can take the heat, and I've hardly seen a short on the boards (The boards then go to Quality control) ... Last edited by hooleydog; 13 October 2018 at 05:18.

13 October 2018, 15:01	#3728
hooleydog Registered User Join Date: Aug 2017 Location: Southfield Posts: 53	Myself.. I've hand soldered lots of SMD and through hole components. I even recapped my old A1200 plus my A2000 motherboard (Removing the old caps and sucking the solder from the holes), and changed my 500+ plus A2000 from NTSC to PAL (I don't like NTSC ).. But I've never soldered a CLPD (Where's my microscope at).... Oh, to be safe.. I even replaced the old caps in my A2000 power supply too, and I added a a/c switch to the front of the 2000, so It's easier to turn it on/off instead of doing it at the back of the computer (Which I hated)... Last edited by hooleydog; 13 October 2018 at 15:19.

13 October 2018, 15:27	#3729
plasmab Banned Join Date: Sep 2016 Location: UK Posts: 2,917	Ok. I guess i really need to find out why AIBB is the only application with corrupt graphics.. Is the author on EAB? I wonder if i could get some sourcecode snipets to reproduce the issue. Other applications do show this too. In the menus. Its just a bit random. Does anyone know how intuition renders the menus? I likely have some kind of cycle collision but i need to narrow down whats actually happening. Last edited by plasmab; 13 October 2018 at 18:12.

14 October 2018, 15:21	#3739
plasmab Banned Join Date: Sep 2016 Location: UK Posts: 2,917	Beta time... ram back in ziii space. https://github.com/terriblefire/terr...10_14_beta.zip time to break it guys. EDIT: While you guys do that i'll see if i can get cache burst going Last edited by plasmab; 14 October 2018 at 15:26.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Listening to Amiga music with too much treble makes it sound terrible!	Foebane	Amiga scene	4	08 October 2016 06:30
Terrible usb drive performance?	lovinggames	support.FS-UAE	5	11 March 2015 09:06
Broken MASPlayer - Terrible hissing.	SS454	support.Hardware	8	17 November 2013 01:41
Accelerators	pmc	MarketPlace	9	04 November 2009 20:23
Problem with a terrible, simply effect!	Camionsauro	Coders. General	13	19 April 2009 08:53

13 October 2018, 10:54	#3725
plasmab Banned Join Date: Sep 2016 Location: UK Posts: 2,917	I found this didnt work with RAM chips. the pins need physically pushed by the iron to make contact. Its also alot more work and as a few people found.. you end up with shorts where there the solder paste dribbles under the chip and shorts between vias. Anyways good luck and let me know how it goes. I never understood why people are so scared of just soldering with iron though. I honestly find soldering a CPLD on easier and less of a PITA than putting through hole resistors on board.

13 October 2018, 14:42	#3726
plasmab Banned Join Date: Sep 2016 Location: UK Posts: 2,917	New tf534 alpha: https://github.com/terriblefire/terr...0_13_alpha.zip I tested this at 25Mhz. I can confirm the AAIB issue. In an effort to get a baseline the Ram is back in the ZII area. this is temp.

14 October 2018, 01:04	#3730
CRP Registered User Join Date: May 2018 Location: Milano Posts: 59	Hi, just wanted to let you know that my TF530 is officially completed, up and running. Everything works as advertised. I have just burnt a dual kickstart eprom and am running both wb 1.3 and 3.1. Everything is speedy, ide interface works perfectly. It has been a bumpy road, but thanks to Mr. Leary for designing this thing and making it available to everybody, and thanks for all the guys who have helped out in this forum. I literally started from zero, have learned a lot and had lots of fun! Thanks!

14 October 2018, 02:08	#3734
alenppc Registered User Join Date: Apr 2012 Location: Canada Age: 44 Posts: 910	No audio issues whatsoever

14 October 2018, 15:09	#3738
plasmab Banned Join Date: Sep 2016 Location: UK Posts: 2,917	Ah. I was hitting this with AIBB. http://eab.abime.net/showthread.php?t=56255 New alpha. https://github.com/terriblefire/terr...0_14_alpha.zip

14 October 2018, 15:58	#3740
plasmab Banned Join Date: Sep 2016 Location: UK Posts: 2,917	Ok i think i understand the slowram issue now. Even though I dont have any. If you have an 030/MMU? You will need a software patch to cache inhibit the $00Dxxxx area except where you have ram (and want it cached). The accelerator cannot do anything about this.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)