LMX2595: Specific frequency and low frequency problems. But otherwise working.

Hi Ian,

Did you cross check your register configuration is correct as compare with the configuration generate by TICS Pro?

For example:

I have older designs that use the same hardware and software. To me this is a really strange problem. Something has changed in this revision of PCB and I can't find it in either the schematic, layout, or components. I have tried most everything i can think of. it is a hard failure, there is a specific frequency (and lower) where the LMX just stops working. With software that operated older systems correctly. Here is the configuration file. I have tried to read and write to the registers but this is new to us, so I'm not sure we are getting is right but it seems like a lot of the registers are correct so I think we have it.

R112   0x7000ff  
R111   0x6f00ff  
R110   0x6e010d  
R109   0x6ddc00  
R108   0x6c00f9  
R107   0x6bcc01  
R106   0x6a0000
R105   0x690031  
R104   0x680000
R103   0x670000
R102   0x663fc0  
R101   0x650019  
R100   0x640000
R99    0x630000
R98    0x620300  
R97    0x610ccc  
R96    0x600000
R95    0x5f0000
R94    0x5e0000
R93    0x5d0000
R92    0x5c0000
R91    0x5b0000
R90    0x5a0000
R89    0x590000
R88    0x580000
R87    0x570000
R86    0x560000  
R85    0x55fb00  
R84    0x540001
R83    0x530000
R82    0x521f00  
R81    0x510000
R80    0x507777  
R79    0x4f0037  
R78    0x4e0027  
R77    0x4d0000
R76    0x4c000e  
R75    0x4b0a80  
R74    0x4a0000
R73    0x49003f
R72    0x480001
R71    0x4700c1  
R70    0x46e378  
R69    0x450000
R68    0x4403fc  
R67    0x430000
R66    0x4201fe  
R65    0x410000
R64    0x401bcc  
R63    0x3f0000  
R62    0x3e0333  
R61    0x3d00fc 
R60    0x3c0000
R59    0x3b0001
R58    0x3ac001 
R57    0x390020
R56    0x380000
R55    0x370000
R54    0x360000
R53    0x350000
R52    0x340820
R51    0x330080
R50    0x320000
R49    0x314180
R48    0x300300
R47    0x2f0300
R46    0x2e07fc
R45    0x2dc0f2 
R44    0x2c3220  
R43    0x2b0000
R42    0x2a0000
R41    0x290000
R40    0x280000
R39    0x270001  
R38    0x260000
R37    0x250104  
R36    0x24010e  
R35    0x230004
R34    0x220000
R33    0x211e21
R32    0x200393
R31    0x1f43ec
R30    0x1e318c
R29    0x1d318c
R28    0x1c0488
R27    0x1b0002
R26    0x1a0db0
R25    0x190624  
R24    0x18071a
R23    0x17007c
R22    0x160001
R21    0x150401
R20    0x14d848  
R19    0x1327b7
R18    0x120064
R17    0x1100e0  
R16    0x100080
R15    0x0f064f
R14    0x0e1e70
R13    0x0d4000 
R12    0x0c5001
R11    0x0b0018
R10    0x0a10d8
R9     0x090604
R8     0x082000
R7     0x0740b2
R6     0x06c802
R5     0x0500c8
R4     0x040a43
R3     0x030642
R2     0x020500
R1     0x010808
R0     0x002418  

Hi Ian,

I have no problem to lock our eval board with your configuration at 150MHz, 180MHz or 2.4GHz.

Could you share your schematic?

1. We have been able to change registers separate  from running our full software. We have found that we can program 300 MHz and the IC will operate correctly. But if we change the divide to 64 (according to table 8) it fails. We can move the VCO lower to get to 150 MHz, but when the SEG3 is turned on by the divide by 64 it always fails. So it "seems" like it's not a specific frequency but a specific divide.

 2. We rework these PCBs, the 10uF caps on pins 27 29 33...) do not have a full wrap around termination (F380J106MMAAH2). If these are not connected I can't inspect it. Would there be a specific pin associated with the divider (especially SEG3) that if not connected to the capacitor would make the divider fail? What pins are associated with the divider (power, bias, gnd..)?

Hi Ian,

Did you mean, the problem happens only when the capacitors at "bypass" pins were not connected properly?

No. I only meant to say that I can't see the connectivity of these parts, and was wondering what would happen "IF" there were disconnected during rework.

I keep getting stuck on the issue of not being able to get to the divide by 64. How can the whole IC operate correctly until I program it for a reasonable divide (that works on other designs using this IC). This software works on other PCBs with the 2595 on it. We use this part on other PCBs that have the same synthesizer circuit section (usign the same serial connections and chip selects) , but different analog after the synth. signal is made.

Hi Ian,

Those capacitors are for the internal LDOs, if they are not connected properly, LDO output is either not stable or has more noise. 

Can you use a wider span to view if there is indeed no signal output or frequency is not correct? Is the VCO locked? You can check the LD pin or read back register R110[10:9]. 

You can also try, program R32[6] = 1 and then = 0. This will reset the channel divider. 

Some of the systems that are not working do have a problem where the frequency is way off. One makes 327 MHz instead of 300 MHz. Generally when we read back the register the has LD, it shows a locked value. Also, isn't the LD signal multiplexed on the MUX OUT pin. That would mean we would have to switch the MUXOUT pin from readback to LD every time we wanted to change from register readback to LD. I don't currently have a way to visually check RD, I'd have to connect a VOM or something.

Can you tell me which specific power and bias pins, anything that supplies power,  to the divider section?

I have gotten the lock detect selected and it shows the PLL being locked for both good and bad frequencies. I am checking it with an oscilloscope so that I can see if it glitches. I don't see any dip. The voltage is lower than I would think. The part is well supplied with 3.3 volts, it only goes to 2.9 volts. I do have a pull up connected (3.0 K Ohms) so that when we make the MUX OUT serial out, it is not floating. We have had trouble with serial not being managed. The MUXOUT might not have enough drive for this resistor.

Hi Ian,

No need to change MUXout to LD output, keep using it as read back output and then read R110[10:9] for lock status. 

At failure, please check

- if there is no output or output frequency is not correct

- if the VCO is locked, read R110

- if R32[6] can resolve the problem

To debug:

- try VCO = 9600MHz, CHDIV = 64 (150MHz output) and CHDIV = 48 (200MHz output)

- try VCO = 7680MHz, CHDIV = 64 (120MHz output) and CHDIV = 48 (160MHz output)

- try VCO = 11200MHz, CHDIV = 64 (175MHz output) and CHDIV = 32 (350MHz output)