Re: Radio Interference Problem

[Fred Townsend]

Wasn't it Winston that said we are two countries separated by a common language? hFE is NOT = to hfe. Gain bandwidth product (Ft) (specified in MHz) is only mildly related to hFE (specified as a ratio). By JEDEC definition upper case notation is used for DC parameters and lower case are ac parameters. In short that is not the end of the horse that you use to check his teeth. I have never heard of a RFI problem that was related to DC parameters. I am talking about ac parameters which are largely unspecified for this type device. This is where you get into problems when substituting different types or even the new for the old same type.  Newer devices almost always have a higher Ft because the manufacturer has upgraded their processes. The manufacturers can upgrade their product without changing the specification because the maximums were never specified in the first place. Declan Moriarty wrote: Recently, Somebody Somewhere wrote these words I'm having a lot trouble understanding what you are saying. Let me state my belief. The aluminum case TO3s that you replaced the existing transistors with probably have a gain bandwidth product (Ft) 10 to 1000 times greater than the old transistors. This means that instead of driving the transducer with a sinusoid waveform it is now a nice crisp square wave. This could cause ringing and RFI. Furthermore, the hfe is now greater while the parasitic capacitance is now reduced by a factor of up to 100. This now permits parasitic oscillations that are heard as RFI. Poor grounding, inductive spikes, and poor earthing are the red herrings. Don't get me wrong. These are issues to be dealt with but not the source of RFI which I see as the subject of this posting. Fred Townsend I checked the gains. The Manufacturer's transistors died (my fault), and I now have the replacements in place. No you have the substitutions in place. BUX 80 (The manufacturer's job) has a minimum hFE of 3 BU225s (The Replacements) have a minimum gain of 10. You are up side down. It's not the minimums but the maximums one worries about for problems relating to RFI. Yes one worries about minimums to make the basic circuit work but RFI, in this type circuit, is almost always caused by too much, ac, gain. Curiously, the BU225s are the older, as I am using old stock (93/94 Vintage) and they are steel, whereas the manufacturer's ones are aluminium. As BU225s have a much higher VCE, (2250V vs 800V) I do not expect great gain from them. The entire thing, of course, as an ultrasonic cleaner is enclosed in a steel casing which has this convoluted earth. I know better than to think it is a perfect seal of all electrical noise - far from. One can affect the transmission path by attacking the antenna but one normally kills the transmitter to suppress the transmission. You might be able to minimize the spikes by putting an RC damping network (snubber) across the primary of the transformer (between the collectors of the transistors). This is what is typically done with triacs and SCRs to lower the level of flyback transients. I also found a lot of power supply schematics that put the snubber across the secondary instead. If there isn't one there, I will certainly try that and see what effect it has. I'm not too worried about radiated RFI, as there is a stell casing which (theoretically anyhow) should act as some sort of Faraday Shield. I had pointed out Ever use a cell phone inside a car? The car is a steel case. Moving the radio off the same wall socket lessened interferance considerably. Trying to tell people that when I moved the u/s clearer to a different mains point from where the radio was connected, the interference was lessened considerably. That would tell me that the interferance was passing via the power lines, which led me to suspect diodes. Suspect the diodes as being the source of the problem or merely in the transmission path? Fred Townsend BTW, is there something funny with your box, Dave? A button stuck? I got 5 copies of your reply.