Terrible Fire Accelerators

[supaduper]

I know Amtech gel flux is prefered by lots of pro solders but unless you have a good ultrasonic cleaner it can take forever to clean, so I use this liquid flux and and swear by it , it cleans off great with isoprop and a quality kitchen towel and hardly any residue after.
And it last forever
https://www.ebay.co.uk/itm/RF800-Liq...53.m2749.l2649

[plasmab]

Originally Posted by supaduper

I know Amtech gel flux is prefered by lots of pro solders but unless you have a good ultrasonic cleaner it can take forever to clean, so I use this liquid flux and and swear by it , it cleans off great with isoprop and a quality kitchen towel and hardly any residue after.
And it last forever
https://www.ebay.co.uk/itm/RF800-Liq...53.m2749.l2649

I just leave the board in IPA overnight then wipe. The ultrasonic cleaner just does damage.

[solidcore]

I second the RF800 flux, it's easy to clean with Isopropyl and a good brush , goes a long way, very cheap, brilliant results! . Even used it to solder fine pitch 240 pin fpga no problem!. Thanks again to supaduper for the recommendation, still on my 1st bottle, bought a 2nd just in case!

[CRP]

If anyone's curious about my progress: I have been checking all solder points one by one. On the bus/cpu/oscillator side everything is ok, but I found pin 32 of the ram cpld and a lot of pins on the ram chips that are not making contact. It is impossible to see this, even a microscope (albeit a cheap usb 640x480 one...) does not reveal the bad solder points. But they definitely lack continuity.
Tomorrow I will put on some solder paste, get out the heat gun and see if I can fix these.

But a ram chip/ram cpld problem should not cause a black screen, correct?

[EzdineG]

Originally Posted by CRP

But a ram chip/ram cpld problem should not cause a black screen, correct?

Anything is possible, but it certainly appears to me like you’re not even getting past the initial tests involving the CPU.

Look over the entire board’s traces and make sure nothing between the 68000 socket and the rest of the board has been broken. Use the schematic and a voltmeter on the resistance/diode “beep” setting to test each pin to the first connectivity point on the board. I’ve found a few duds this way, as well as some boards damaged purely by (mis)handling.

If you’re using a socket relocator, make sure a normal CPU works where the TF530 doesn’t.

Lastly, make sure your 68000 socket hasn’t been damaged due to excessive testing. I’ve had to replace a few of these along the way, too.

[dalek]

I had to replace my 68000 socket too after too many insertions.

But also watch for the Kickstart ROM socket wearing out if you physically switch between diagrom and kickstart a lot - mine did and had me stumped me for ages!

Hi Everyone,
Couldn't find any info on the recommended voltage for each capacitor on the TF328 and TF530?

TF328
1uf - V=?
10uf - V=?
22uf - V=?
100nf - V=?

TF530
1uf - V=?
10uf - V=?

Any recommendations would be greatly appreciated.

[plasmab]

Originally Posted by RetroNut

Hi Everyone,
Couldn't find any info on the recommended voltage for each capacitor on the TF328 and TF530?

TF328
1uf - V=?
10uf - V=?
22uf - V=?
100nf - V=?

TF530
1uf - V=?
10uf - V=?

Any recommendations would be greatly appreciated.

Anything above 6V is fine. Go one voltage up if you feel need to for safety.

[chue]

I finally got it working! Here were my latest issues, and how they were solved:

1. "Unimplemented Instruction" in Diagrom.

I am using a 27c160 (1 Megabyte) EPROM and had burned Diagrom so that it would be duplicated in both the upper and lower 512k of the EPROM. I assumed that I could let the high address pin "float" and that the EPROM would select either the upper or lower image. This worked well enough to boot but was unstable.

After seeing that my original 1.3 Kickstart was mostly stable, I decided to use the EPROM and tie the high address pin to ground. Diagrom became stable.

I could finally run tests and found a RAM issue. A few pins were not soldered, so I fixed those. After that I reran the RAM tests and everything looked good!

It was time to install the CF card I had prepared, which led to my next problem:

2. With the CF card installed on the TF530, the Amiga would boot to a black screen.

After quite a bit of research I found out that each IDE cable has pin 1 denoted by a different color wire on the cable. Once I figured out which "edge" of the cable was pin 1, I realized I had installed the IDE cable on backwards... I changed the cable orientation, and it worked!

Here are some screenshots:

* The obligatory SysInfo screen
* The TF530 installed in the Amiga, with shielding cut
* Up close, showing jumpers - I am pulling 5 volts off the TF530 to power the CF card
* The INT2 jumper on the motherboard

Thanks again to Stephen for the TF530 and other members here for helping me with this!

[8 Bit Dreams]

+1

[lolof]

Originally Posted by plasmab

The ultrasonic cleaner just does damage.

Oh ! I soldered tons of Amiga stuffs and always cleaned all of them in ultra-sonic bath with propanol2.....

Never, never had issue with it.... I must be very lucky then.....

[plasmab]

Originally Posted by lolof

Oh ! I soldered tons of Amiga stuffs and always cleaned all of them in ultra-sonic bath with propanol2.....

Never, never had issue with it.... I must be very lucky then.....

It works but the bath does damage. Watch the Louis Rossmann video on the subject.

[CRP]

ok, so today I told myself: "let's fix the bad solder points", which I promptly did. Checked them all for continuity, and also verified that no adjacents pins are shorted. Then I reprogrammed the CPLDs, just to be sure they're fresh, and installed the board. Turned on the A500, and for a few second nothing happens. Then I smell something, and shortly after white smoke comes out of the empty coprocessor socket. I quickly turn off everything, and now on my tf530 GND and VCC are shorted across the whole board...

[plasmab]

Originally Posted by CRP

ok, so today I told myself: "let's fix the bad solder points", which I promptly did. Checked them all for continuity, and also verified that no adjacents pins are shorted. Then I reprogrammed the CPLDs, just to be sure they're fresh, and installed the board. Turned on the A500, and for a few second nothing happens. Then I smell something, and shortly after white smoke comes out of the empty coprocessor socket. I quickly turn off everything, and now on my tf530 GND and VCC are shorted across the whole board...

That’s mighty strange. Never had that. You must have a 5V short at the ram.

[CRP]

that's what I thought. I could not find it with microscope, so I desoldered all ram chips. Still short. desoldered all caps, still short. desoldered the coprocessor socket (a real pita! destroyed it in the process...) and short is gone...
so I am back almost at square one

[plasmab]

Originally Posted by CRP

that's what I thought. I could not find it with microscope, so I desoldered all ram chips. Still short. desoldered all caps, still short. desoldered the coprocessor socket (a real pita! destroyed it in the process...) and short is gone...
so I am back almost at square one

No i understand what happened now. One of your solder joints was a blob that made contact with the edge of the ground plane... keep the solder neat. Remember that the solder mask can chip and expose ground and if your solder is not neat it can cause this kind of issue.

[dalek]

can confirm, exactly this happened to me on my first board. I was/am such a noob but so much better now thanks to this project! Same sorts of shorts are possible on the ide and and 030 solder mask.

[CRP]

Yeah it really is bootcamp for a novice
Anyway, this time I want to take it a step at a time. Currently my board has the following still mounted:
- cpu socket
- two CPLDs
- all pin headers
- resistor networks
- smd resistors

my question is the following: if I install all caps, the voltage regulator, the oscillator, the buffers and the CPU, should I expect a black screen or will the amiga come to life, just with no additional ram? In other words, does the "bare no ram" accelerator show some sign of life only with diagrom or with ordinary kickstart as well?

Finally, I still am not really sure about the programming order of the CPLDs, which I perform with the Raspberry PI: Bus is first item in JTAG chain, RAM is second. But according to the wiki, programming should be done with parameter "-p 0" for ram and "-p 1" for the bus, which sounds like the opposite of the jtag order. Does anyone have a definitive answer to this?

Thanks

[8 Bit Dreams]

It will boot even without ram chips on it, memory will be reported as defective, though

[plasmab]

Originally Posted by CRP

Yeah it really is bootcamp for a novice

Finally, I still am not really sure about the programming order of the CPLDs, which I perform with the Raspberry PI: Bus is first item in JTAG chain, RAM is second. But according to the wiki, programming should be done with parameter "-p 0" for ram and "-p 1" for the bus, which sounds like the opposite of the jtag order. Does anyone have a definitive answer to this?

Thanks

Wiki is wrong then. Watch the video where I actually do it. I show my screen