Terrible Fire Accelerators

[plasmab]

Originally Posted by 8 Bit Dreams

Hello Stephen,

is it hard- or software related?

Firmware. I’ll maybe send you a beta

[8 Bit Dreams]

Originally Posted by plasmab

Firmware. I’ll maybe send you a beta

Thank You!
Will be happy to test and give full feedback

[plasmab]

Originally Posted by 8 Bit Dreams

Thank You!

Will be happy to test and give full feedback

Not sure it will fix all the issues but maybe a start

[chue]

Greetings,

I'm doing a TF530 Rev 3 build, and am having trouble bringing up (identifying) the CPLDs in ISE. If I short the BYPASS jumper on the board, ISE recognizes one CPLD. Without the bypass, ISE complains that there is some hardware issue.

My question is - which of the CPLDs is having trouble here, is it the RAM or is it the BUS CPLD?

MORE DETAIL:

I have most of the ICs soldered on - the exception is no RAM is soldered on yet.
All the caps are on.
The resistors (non-array) are on.
The resistor arrays are on, but instead of 1K networks I am using 4.7K

I am using a bench power supply and measure 5 volts at the 68K VCC pad. I also measure 3.3 volts at the JTAG jumper.

[plasmab]

Originally Posted by chue

Greetings,

I'm doing a TF530 Rev 3 build, and am having trouble bringing up (identifying) the CPLDs in ISE. If I short the BYPASS jumper on the board, ISE recognizes one CPLD. Without the bypass, ISE complains that there is some hardware issue.

My question is - which of the CPLDs is having trouble here, is it the RAM or is it the BUS CPLD?

MORE DETAIL:

I have most of the ICs soldered on - the exception is no RAM is soldered on yet.
All the caps are on.
The resistors (non-array) are on.
The resistor arrays are on, but instead of 1K networks I am using 4.7K

I am using a bench power supply and measure 5 volts at VCC at the 68K VCC pad. I also measure 3.3 volts at the JTAG jumper.

If you have this issue either the RAM cpld is dead or the soldering is dodgy.

[chue]

Thanks Stephen I will rework the RAM CPLD. I had a heck of a time getting it on, and probably killed it.

[alenppc]

Yeah, you most likely got a micro bridge on a cpld pin somewhere.

[8 Bit Dreams]

Originally Posted by plasmab

If you have this issue either the RAM cpld is dead or the soldering is dodgy.

as Stephen said:
1. Check pin1 for both cplds
2. Check both cpld's for shorts/bad soldered pads under the scope
3. if 1 & 2 are not the issue - it's possible that one of the cpld's is dead/not working/fake. where did You got them? Ebay or trusted seller?

[chue]

Originally Posted by 8 Bit Dreams

as Stephen said:
1. Check pin1 for both cplds
2. Check both cpld's for shorts/bad soldered pads under the scope
3. if 1 & 2 are not the issue - it's possible that one of the cpld's is dead/not working/fake. where did You got them? Ebay or trusted seller?

I apologize, I am not sure what you mean by the first. What do I check on pin 1?

I bought the CPLDs from digikey, so I assume they are real.

I just resoldered the RAM CPLD and still have the same issue with ISE not recognizing it. I checked for shorts and bad solder joints it, and seems ok.

I will resolder the BUS CPLD next and see if that fixes it. If not I will replace the RAM CPLD with a different one.

Thanks for all the suggestions everyone!

[8 Bit Dreams]

Originally Posted by chue

I apologize, I am not sure what you mean by the first. What do I check on pin 1?

I bought the CPLDs from digikey, so I assume they are real.

I just resoldered the RAM CPLD and still have the same issue with ISE not recognizing it. I checked for shorts and bad solder joints it, and seems ok.

I will resolder the BUS CPLD next and see if that fixes it. If not I will replace the RAM CPLD with a different one.

Thanks for all the suggestions everyone!

I mean is the dot marking pin1 on the right place?
How do You power Your TF? External or in the Amiga?

[chue]

Originally Posted by 8 Bit Dreams

I mean is the dot marking pin1 on the right place?
How do You power Your TF? External or in the Amiga?

Oh yes, the dots are facing "south" relative to the TF head logo. That is the dots point towards the IDE connector.

I am using an external bench power supply.

[plasmab]

Originally Posted by chue

Oh yes, the dots are facing "south" relative to the TF head logo. That is the dots point towards the IDE connector.

I am using an external bench power supply.

Don’t touch the bus cpld .. if you managed to program it it is not involved in the problem

[chue]

Originally Posted by plasmab

Don’t touch the bus cpld .. if you managed to program it it is not involved in the problem

Thanks for the warning!

I haven't done anything yet with it, including trying to program it. I only tried to get ISE to recognize it. I will try programming it today and if that works I will go back to reworking the RAM CPLD.

[EzdineG]

Originally Posted by chue

Oh yes, the dots are facing "south" relative to the TF head logo. That is the dots point towards the IDE connector.

I am using an external bench power supply.

Can you link the part number of the voltage regulator you’ve specced in your build?

[chue]

Originally Posted by EzdineG

Can you link the part number of the voltage regulator you’ve specced in your build?

I used this one:

https://www.digikey.com/product-deta...BCT-ND/1010516

[chue]

Originally Posted by 8 Bit Dreams

External or in the Amiga?

External, the TF530 is not connected to the Amiga.

[supaduper]

Just a heads to say that Digikey has re-stocked the CPLD`s ... types used on terrible fire boards..here :

https://www.digikey.co.uk/product-de...1388-ND/826991

https://www.digikey.co.uk/product-de...1385-ND/826988

[alenppc]

Thanks! They still priced them way higher than they used to be.

[chue]

Just an update - I haven't had much luck with my TF530 build. I have resoldered ALL components on the board, and ISE is not able to ID either CPLD. ISE also lets me test the JTAG chain and that test always fails:

// *** BATCH CMD : CheckIntegrity
Maximum TCK operating frequency for this device chain: 10000000.
Validating chain...
INFO:iMPACT:1206 - Instruction Capture = '1111111100000001'
INFO:iMPACT:1207 - Expected Capture = '10101010000XXX01'
INFO:iMPACT:2130 - Boundary-scan chain test failed . Please check tdi->tdo connection between the cable and device:'1' ( 'xc9572xl').
A problem may exist in the hardware configuration.
Check that the cable, scan chain, and power connections are intact,
that the specified scan chain configuration matches the actual hardware, and
that the power supply is adequate and delivering the correct voltage.

My conclusion at this point is that it definitely is some problem on the board. My guess is that it is some combination of dodgy soldering, dead CPLD(s), or bad board.

What is definitely ok:

My Xilinx software / programmer / cable are good, as I have a Xilinx dev board that I am able to connect to, and the setup can ID the CPLD on that board properly. The JTAG chain test works with this setup as well.

My power supply is good, as it worked with the above Xilinx dev board.

My next move:

I am going to try building a second board. The upshot of all of this is that my soldering has improved a lot, so the second board will hopefully be an easier build.

[plasmab]

I have seen many issues with TF530 builds but I’ve never seen a board fail to list the chips on the JTAg chain