DAC Problem


Reto HB9TRT
 
Edited

We have here 6 Ody2 built. 2 works perfect, and 4 has an issue in TX. We did much trouble shooting but have no idea what causes this issue. 

The output power of this 4 odys is not adjustable by thetis. They allways have full power. We took some measurements directly at the DAC and the problem must be there. 

On odys working there is a voltage of 100 to 900mVpp. The 4 with the problem shows 1.6Vpp.

On pin 18 of the dac the working odys shows 290mV and on the 4 others 2.5V. This is the only significant difference we can test. Pin 18 is full scale current adjust.

All odys has a 2k resistor from pin 18 to gnd. The resistors are OK. So we guess this have to do with software. Because it's nearly impossible that the 4, which were built from different oms has the same hardware issue. 

The working twos are the first which were built. Also my one. I tried first with different software releases. David's and also Davide's. The 4 newer ones were programmed with the actual firmware and they have never had an older one programmed. Is it possible that a register, memory was set by an older version and was not reseted by writing a newer firmware? A memory cell which will not correctly written if a new firmware will be programed? I don't know if the whole fpga will be 'formatted' if new software is programmed, or the mcu.
Because we have no idea where to search, the software is now in the focus. 

Again, we think it cannot be hardware, because it's nearly impossible that all makes the same error on 4 different devices. 

Now we hope enyone has an idea how to find this bug. 

Vy 73 
Reto


David Fainitski
 

Reto, this is hardware.
Check the OpAmp datasheet, you will see many kinds of package pinout is existing. Are you shure you have used correct one ?


Sergei Moissejev
 

This sounds indeed a hardware issue - have one working in Estonia and on in the UK and both are fine.

Serge / M0TLN/ES2SRG

On Sun, 14 Feb 2021 at 23:56, Reto HB9TRT <hb9trt@...> wrote:

[Edited Message Follows]

We have here 6 Ody2 built. 2 works perfect, and 4 has an issue in TX. We did much trouble shooting but have no idea what causes this issue. 

The output power of this 4 odys is not adjustable by thetis. They allways have full power. We took some measurements directly at the DAC and the problem must be there. 

On odys working there is a voltage of 100 to 900mVpp. The 4 with the problem shows 1.6Vpp.

On pin 18 of the dac the working odys shows 290mV and on the 4 others 2.5V. This is the only significant difference we can test. Pin 18 is full scale current adjust.

All odys has a 2k resistor from pin 18 to gnd. The resistors are OK. So we guess this have to do with software. Because it's nearly impossible that the 4, which were built from different oms has the same hardware issue. 

The working twos are the first which were built. Also my one. I tried first with different software releases. David's and also Davide's. The 4 newer ones were programmed with the actual firmware and they have never had an older one programmed. Is it possible that a register, memory was set by an older version and was not reseted by writing a newer firmware? A memory cell which will not correctly written if a new firmware will be programed? I don't know if the whole fpga will be 'formatted' if new software is programmed, or the mcu.
Because we have no idea where to search, the software is now in the focus. 

Again, we think it cannot be hardware, because it's nearly impossible that all makes the same error on 4 different devices. 

Now we hope enyone has an idea how to find this bug. 

Vy 73 
Reto


Reto HB9TRT
 
Edited

Hi David

yes we have exactly the same Chip on all Odys.
same Type, but 3 different batches, if I interprete the Nummer on the third row correctly. They have Production Date 2017,2019 and 2020. From the 2020 we have 4 with the same Production week.
3 of them make the Problem, one works fine.
the 2017 is my one and works fine. The 2019 is also my one and isn‘t finished built. The other non working I don‘t know the date behause it is not here but I guess 2020 behause the OM ordered it at the Same time like the others (this Year).

Another question: i made some Modulation Tests today with the MH31 Mic. Modulation ist basically Fine, but there‘s a relaively high Basic noise while pressing the ptt without speaking. There is on noisy Environment. A mic Problem? Or thetis settings? 

vy73
Reto


Reto HB9TRT
 

The dac is a ad9744 aruz


Reto HB9TRT
 
Edited

Hi serge 

We also have 2 working. And all have the same dac. That's special. It's also not a problem of the final stage of the ody. The issue can be measured after the transformer where IOUTA and IOUTB goes in. 

The only difference of the two working is, that they are finished 6 or 7 months before and we load older firmware in. I know, the today firmware is not new, but I test also 1.11 anp. 

I know that sounds crazy, but we must find a difference. It could not be possible thst the 4 newest all have the same hardware issue, built by different builders. I try to help the oms to get their odys working, and at the moment I have no idea how to search. 

Vy 73 
Reto


David Fainitski
 

Reto, this is hardware.
Check the OpAmp datasheet, you will see many kinds of package pinout is existing. Are you shure you have used correct one ?
OpAmp - is an Operational Amplifier shortly.


Reto HB9TRT
 

Hi David 

You mean the mcp6001T?

Vy 73 
Reto 


Reto HB9TRT
 
Edited

Hi David 

You're right. The om's ordered a wrong type of MCP6001T... Because the 6001T was not in stock at Digikey, they ordered the 6001UT which has a completely different pinout. Because I have the correct one also in the BOM, I never thought of that. Sorry about this, but I learned much from the ad9744... I read the whole datasheet :)

Now the ordered the correct ones at Farnell or mouser where they are in stock. 

Vy 73 
Reto


Reto HB9TRT
 

Hi David, 

I try to solfe the MCP6001(U)T situation and saw, that in the HFPU the U variant will be used 4 times. Vout on Pin4. But there's a fault in the schematics of HFPU. U4 and U4 are marked as MCP6001 where the Vout is on Pin 1. This is not correct. There also must be the U type where Vout is on Pin 4. 

I  have no Schematics for the Ody2, so I took the one from Angelina 100D. At least the part arround the AD9744 seems to be the same as on ody. The Vout goes to the REF IO Pin of the DAC. and here it's the MCP6001(T) with Vout on Pin1. 

The Last builders use the MCP6001U(T) instead of the 6001(T) and so the Ody dont works correctly (Vout on Pin4 -> Clould not work). On the HFPU we all don't analyzed it accurate and took the MCP6001U(T) for all 4 positions which is not the same as on the schematics, but which is luckily correct. SO I wonder, that my PA shows Power forward and reverse with the wrong OPAMP installed... it was not the wrong one, it was correct, only on the Scheamtics there's the wrong type, schematics is correct, but the designbator and Pin Count for the 6001U(T) is wrong. 

At least this is what we analyzed last night. Could you check this?

A fine little difference ;)

vy 73
Reto


Reto HB9TRT
 

Here a picture for explaining. 

vy 73
Reto


David Fainitski
 

Hello, Reto
U - version is used in HFPU-100 everywhere, for ADC lines too, see the pic below
Simple version is used in Odyssey-2
BOMs are correct both.
The schematic of HFPU-100 does not indicate U version for ADC lines buffers, but a schematic in general should not indicate correct pinout for IC.


Reto HB9TRT
 

Hi David, 

Thanks for the Explenation. I don't se the U on the board. I thought the Schematics is correct. But now I think all mysteries about the OPAMP are solved. The 4 non working Odys now also works fine. 

vy 73
Reto