T Harrison

  • Content count

  • Joined

  • Last visited

Community Reputation

0 Neutral

About T Harrison

Contact Methods

  • ICQ
  1. Have you considered using a sub-audible tone such as 20Hz rather than a super-audible tone (20kHz)? This might act as a suitable stimulus signal, without being in the conscious hearing range of the animals? You will likely need to experiment with the frequency and level somewhat, as you don't want to saturate the loudspeaker's internal transformers and generate unnecessary heating of the transformer. Also, you need to be careful that a constant low frequency signal which is getting near to DC is at a sufficiently low level not to heat the voice coil too much. Conversely, you may find the heating effect prevents driver problems in very cold temperatures during the winter months, even if it results in earlier driver failure than if it had no stimulus. Just some food for thought...
  2. We have since had more instances of it mis-behaving. We have discounted a single SHARC chip, as it has happened on multiple inputs and outputs across different chips. We have checked with the local electricity company who have admitted to problems. They have since fixed many local faults, yet we have still had it mis-behaving since. The electricity company is putting a fault logger on our incoming power supply soon, so we'll see if that shows any more power related issues. Once the incoming power is shown to be fault-free, we'll be able to isolate whether its the unit or the supply. Watch this space... Tim Harrison
  3. I'd just like to add that I regularly get Access violations running IQwic offline under XP with no components connected. I have experienced this from IQwin 5.11 onwards. I wondered if it might be graphics resolution related, as I run two monitors at 1280 x 960. However using the same dataframe on-line with a WinXP laptop using lower resolution (1024 x 768 I think), the program is very stable. It often seems to occur when closing a component control panel (8810C), but I haven't found any other repeatable identifiable actions that cause crashing. The dataframe in question was started on Win98 and ported over from IQwin to IQwic, and has been through hundreds of revisions. Have also had similar, albeit less frequent problems with various dataframes running under WinNT4. Haven't tried Win2K. Hope this helps in the ongoing quest for bug-free perfectly behaved software!
  4. Thanks Brad. All very useful and practical background which the contractor appreciates. The only area which remains unanswered is no. 4... Tim
  5. In the past few months we have experienced a couple of odd occurrances of losing control of an 8810C (V3.00). The unit is left permanently on, to ensure the scheduler can recall presets at times when the main sound system is not on. In the two instances, which occured approx 2 months apart, the system appeared to work normally, but when using IQwic with it (the PC running IQwic is not running permanently), the metering and mute status of some inputs and outputs was incorrect. In our case, Main output B and some of the aux outputs was wrong. We went through the system, component by component, thinking it must be elsewhere in the system. We finally concluded it was the 8810C, so I power-cycled it, and all was well again. Our only theory at present, is that it may be suffering from any mains voltage irregularities (under/over-voltages, brownouts etc). We are therefore looking at installing a line conditioner or UPS to see if that prevents this re-occurring. Curiously, we had a client at work who had an IQ-USM 810 "lose" its three presets. The client was convinced that it was power related, as they have very high voltages (up to 260V) present in that part of Scotland (measured UK voltage is 240V in most places, not 230V as most non-UK residents often think. Not truly unified with Europe yet, I'm afraid ). Working for a distributor, I see many products with switch mode power supplies suffer problems of various types. This is not necessarily a criticism of them, rather an observation, that perhaps more research needs to go into their specification, and tests on long-term real world performance. I think manufacturer's reputations can often suffer as a result of buying in an OEM module.
  6. On a similar, but slightly different topic, I have experienced difficulties with what happens with control polling when changing presets. After changing presets, the 8810C recalls the value of the parameter stored in the preset, and disregards the control port value, until it next updated. In my case, I have a remote potentiometer connected to analogue control port 1 acting as a volume control for a remote zone. What can happen is that volume pot is set by someone in the remote zone to minimum (off). Then a preset is recalled in the main zone of the system. It cross-fades to the new preset over 10 seconds and recalls a level of 0dB, which then disturbs the remote zone. This can confuse the users in the remote zone who don't understand why the volume has changed. Also, as soon as they move the volume fader slightly, the volume jumps back to the fader's position. I can see that for some applications, you would want the preset's stored value to override the control port, however it would be helpful if there was the option (per individual control input), to ignore recalled preset's values. Having used many different vendor's DSPs audio systems, they all vary on the flexibility and methods available of remote control implementation. The difficulty for a system's integrator is that it's impossible to simulate how the system will work conclusively, until you've installed it and programmed it, by which time its too late if it can't be made to behave as you require. We looked at using scenes as an alternative, but found it to be impractical. Aside from the huge list of parameters that would need programming, changing scenes is instant; there is no ability to cross-fade between scenes - again another feature which could be helpful. Hope this is useful user feedback. More feature requests for the melting pot...
  7. Some questions for the IQ team regarding the detail of load monitoring: 1) Need more explanation of HOW the load monitoring is performed, in order to explain to a consultant and contractor. 2) How is it possible to calculate the minimum signal required to ensure that the system can accurately load monitor at all times, as the system needs some stimulus in order to monitor. 3) Need advice on how to set the high and low impedance limits. The system concerned would be a three-way active system, so will have different components on different channels. 4) It appears not to be possible to measure the impedance curve of the load, although this feature used to be available in the now discontinued SLM-8 module (curve viewer under the View menu), and is available on competitive product - the EV P series amps for instance. Is this planned to be available for the future? We have scoured the documentation for greater explanation, but without success. Any further information on the subject would be greatly appreciated. Tim Harrison Fuzion plc.
  8. What's new in firmware version 3.004 since version 3.003? It would be very handy to have some description somewhere on the site as to any improvements between software versions. I guess this is one for Bruce?
  9. The question you've got to ask yourself is how much EQ are you using?! From my experience, you've got to have seriously large amounts of EQ going on to "run out of gas". I just tried a quick experiment, and had 8 bands of EQ on every output, and 2 on every input, without reaching "out of gas". Sounds like you need to look elsewhere in your system for your problems. If the system needs that much EQ, then surely speaker and microphone locations, quality of speakers, mics or acoustic environments would be the obvious starting points to look at. Hope this helps.