SX52 OSCHS setting for 50mhz Resonator

Comments

• LoopyByteloose Posts: 12,537

  2005-10-21 15:16 edited 2005-10-21 15:16

  Gunther mentioned this to me, I think the 10k is correct and the other is a typo. I would have to search the Forum to find the exact explanation. In either case, the debate was whether it was at all necessary to put the resistor in. In breadboarding it is often omitted as a convience, but generally it is included in the final design. It seems it would be easy to be misled into thinking that it was never necessary.
  
  As you can see the OSCHS3 is chosen with the 1M. This consumes more power and is likely to create more EMI.
  
  In general, using the lowest power setting that gives you a stable oscillator is the best solution - Thus the OSCHS2 with the 10K resistor.
  
  I have been wondering where on the proto board are you going to insert the resistor? I guess either the X1 or C1 socket can be used as they both seem to have an XI and X0 which the resistor must use to create a parallel connection.
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  "When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)
  
  ······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
• Mike Cook Posts: 829

  2005-10-21 15:21 edited 2005-10-21 15:21

  Kramer,
  
  Thanks for the reply!
  
  I'm using a 10K on the SX52 board and put it in the X1 socket. I was having a heat problem running at 50MHz, this seemed to take care of the issue. However I have another SX52 running at 20MHz and the 10k resistor and the SX52 does not get warm at all.
  
  Mike
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  "OEM NMEA GPS Module" Now available on ebay for only $19.99
  
  Product web site: http://www.allsurplus.net/Axiom/
  
  
• LoopyByteloose Posts: 12,537

  2005-10-22 11:22 edited 2005-10-22 11:22

  Well, if you go back to those specifications, I think you will find a bit of 'voodoo'.
  
  At 20mhz, you might try 1meg of resistance [noparse][[/noparse]I believe that is what it says].
  Also, you can try to program the lowest driving power to the oscillator.
  I suspect going too low will not do any damage, just cause the oscillator to stop.
  
  I seem to be spending more time reading than actually building these things because I am away from home 12-14 hours a day. When I do get to building, you will have helped me tooo. Thanks.
  
  By the way, do not worry about the 'internal resistor' as the SX-Key software only turns it on if you ask for the internal oscillator. If you do turn it on and then later want to change back -- be sure to re-program before you plug in an external oscillator, crystal, resonator, or resistor.
  
  There seems to be the possiblity of causing greater overheating. THIS NEEDS TO BE VERIFIED BY SOMEONE AT PARALLAX.
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  "When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)
  
  ······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
  
  Post Edited (Kramer) : 10/22/2005 11:32:37 AM GMT
• Guenther Daubach Posts: 1,321

  2005-10-22 19:40 edited 2005-10-22 19:40

  Hi Guys,
  
  let me place some comments and answers to the requests I got via PM in the public:
  
  1. Internal oscillator active with SX-Key connected
  
  As long as the SX-Key is just connected to the OSC1/2 pins without actively generating an internal clock, this should not cause trouble at all. When you let the SX-Key generate a clock signal using the IDE's "run-Run", or "Run-Clock" options, an unpredictable situation might occur as the external clock fed into the OSC1 pin and the internally generated clock might interfere. I don't think that this will cause any danger to the SX-Key as its clock output line is protected against overloads. On the other hand, the SX might "go nuts" with such a combination - mabye "feeling" clocked at a high rate, so simply don't do it. The SX-Key and the SX-Key IDE can't take away all responsibility from the users, so "think before you click".
  
  2. External oscillator active with SX-Key connected
  
  When a resonator or a crystal is connected to the OSC1/2 pins, or the output signal of an external clock generator is fed into the OSC1 pin, a similar situation arises when the SX-Key generates its own clock. Both signals are offending, and you usually will not be able to debug the device in this configuration. Most likely, the Debugger will display that the SX is running, and the debugger buttons are inactive. Again - simply don't do it - disconnect the external resonator, crystal or clock source before connecting the SX-Key for debugging. I never tested it but I assume that debugging won't be possible either if the internal clock generator is active.
  
  On the other hand, when you just want to use the SX-Key to re-program the SX with an external resonator/crystal connected, this usually works out fine, i.e. the SX-Key takes over control by pulling up the OSC1 pin to 12V (the programming voltage Vpp). This makes the SX enter the programming mode, and it will no longer drive the resonator/crystal. After programming has finished, and Vpp has been removed, the SX usually does a reset, and starts executing the program found in program memory even with the SX-Key is still connected (not generating a clock signal, of course), driving the resonator/crystal to generate the system clock. Usually, crystals and resonators can stand the 12V Vpp, so there is no need to disconnect/remove them from pin OSC1 before programming the SX. This may be different with an externally connected clock generator. I once tried this, receiving a "Vpp generation failed" error from the IDE. This was because the external clock generator caused an excessive load to the programming voltage but neither the SX-Key nor the clock generator were damaged.
  
  3. Parallel Resistors, Resonators/Crystals OSCHSx Settings
  
  It never happened to me that an SX would not start generating the clock with no parallel resistor connected. Nevertheless, If you are going to design SX-based applications to be operated within the specified ambient temperature range (-40° ... 85°C), the parallel resistor would make sure that the clock starts running under extreme conditions. The same is true for the setting of the oscillator gain, controlled by the OSCHSx setting. To my experience, a 50 MHz ceramic resonator always worked fine with OSCHS2 and OSCHS3 where using OSCHS1 sometimes caused an unreliable oscillator startup.
  
  4. Heat Problems
  
  When you look at the diagrams presented in the SX28 datasheet, it shows a supply current of almost 90 mA at 5.5 V suply voltage at a clock frequency of 50 MHz when a resonator is used. This means that the SX consumes (i.e. dissipates into heat) almost 0.5 Watts at these ratings. Although not explicitely specified in the data sheet, I assume that this diagram does not consider any additional power dissipation caused by the currents and the voltage drops sinked and/or sourced by output pins. Therefore, it is normal that an SX gets warm, especially when operated at higher clock speeds, like 50 MHz. Nevertheless, in my applications I never found situations that the SX go so hot that you could not touch it with your finger. I had situations when this was the case but only when either outputs had a short, or when the SX was internally damaged.
  
  Just to make it clear, I don't think that selecting a resistor parallel to the resonator, or the setting of the OSCHx option has a major impact on the dissipated heat.
  
  5. Some precautions to protect your SX-Key and the SXes
  
  Although the parts on the SX-Key board and the SXes are protected against electrostatic discharge (ESD) to a certain extend, it is a good idea to not go and test the limits. Therefore, don't connect/disconnect the SX-Key or the SX while the system is powered. Before touching components on your setup, discharge yourself by first touching a lead or surface that is connected to ground. If possible, avoid areas with floor carpets, or other materials that might cause charging your body when handling electronic components.
  
  ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  Greetings from Germany,
  
  Günther