I have updated the section on my site the describes some progress in home-brewing air-core based sim flight gauges. There is a new picture of two gauges that are beginning to show promise.www.mikesflightdeck.com/aircore_instruments_1.htm Mikewww.mikesflightdeck.com

Thanks for the update, Mike. This new generation is looking mighty good! Amazing how nice an old laser-printed faceplate can look. Now the big question is, "Can you make an air-core artificial horizon?" Probably need servos for that.Since TRC is no longer selling individual instruments, I forsee your project becoming of interest to more and more home cockpit builders.I look forward to the next installment of your air-core adventures.Peter http://bfu.avsim.net/sigpics/PeterR.gifBFU Forums Moderator[table border=2 cellpadding=0 cellspacing=1][tr][td][table border=0 cellpadding=8 cellspacing=0][tr][td bgcolor=#540000]http://bfu.avsim.net/sigpics/logo75t.gif[/td][td align="center" bgcolor=#FFFFF6]Bush Flying Unlimited"At home in the wild"Looking for adventure? Come join us! * [link:bfu.avsim.net|Web Site] * [link:www.cafepress.com/bfu,bfu2,bfu3,bfu4|BFU Store] * [link:bfu.avsim.net/join.htm]Join!][/td][/tr][/table][/td][/tr][/table

Peter,You are correct. Air-cores will not work for artificial horizons, at least, not in the present incarnation. They seem best for single pointer style gauges. Working with stepping motors and servos is on the list. I have a small selection of stepping motors and have played with them a bit already. I will likely start experimenting with servos by using those sold under the Tower Hobbies label. I have heard that they are really Futaba units made for Tower, not that it particularly matters to me. Since I will be ripping them apart, I'll start with the least expensive ones regardless of label.First though, I plan on wrappng up the air-core chapter and teaching myself PICmicro assembly language.Mikewww.mikesflightdeck.com

I had originally looked at hobby servos for instruments but due to their limited rotation range, they would only work for some. That's why I was very intrigued when I first saw your site a few months ago.Since you're now going to be using a PIC, are you also going to use the digital air-core driver chips? I saw that two or three companies make the chips in single, dual and triple drive models.Peter http://bfu.avsim.net/sigpics/PeterR.gifBFU Forums Moderator

Peter,I'm not planning on using an air-core driver chip. They're too hard to get. National doesn't make the LM1819 anymore. On Semi is on the last production runs of CS289, and while the CS 4121 and CS8191 are still in production, it's rare to find any in stock. I suppose they directly to the OEM users of air-core units and assemblies. Phillips makes the SA5775/7/8, but they are no easier to get. The analog prototype driver I put together uses cheap quad opamps, a few comparators and a D-flop. I put a few outboard transistors on the opamps driving the coils and wrapped feedback around them. Dreadful crossover distortion, but the coils don't mind.The general approach with the PICmicros is to use a simple piecewise linear approximation to sine and cosine, essentially what the Phillips chips use. The approximations look remarkedly rough, but it turns out that you really are mostly interested in the ratio (the tangent of the desired pointer angle) of the currents in the two coils rather than absolute currents. The ratio (really, the tangent) stays within a fraction of a degree throughout a complete 360 degrees. The thought is to use the PIC to take a digital serial command in, convert it to piecewise approximations, and output a pulse width modulated pulse stream for the two air-core coils. External opamps will low pass filter the pulse stream and, using the same opamp with external transistors circuit as before, drive the air-core movement. The price of materials is quite attractive. The PIC is about 3 bucks, quad opamps are about a quarter, while transistors are a dime a piece. Actually, I already have all the hardware.On the other hand, I have not developed micro-code for years and I'm not (currently) up to speed on the PIC family. So, if it sounds like I know what I'm talking about, please take it all with a grain of salt. (My wife and kids certainly do.)RC servos undergo horrible, disfiguring transformations at the hands of pitiless robotics hobbyists. Actually looks like a lot of fun... I plan on following in their footsteps. One very attractive mod is to remove the servo's potentiometer and replace it with a fixed voltage divider. The output gear is then modified by removing the mechanical tab that limits its rotation. The robot guys do this to make a variable speed geared motor to drive the robot's wheels. I think a very similar approach can be used to drive an external multi-turn pot with a bit more gearing with an indicator pointer on it. This should get beyond the nominal 180 degree limtation of an unmodified servo, plus it should provide a finer degree of resolution. The servo has an internal deadband to stabilize its activity. Spead over 180 degrees this results in about 8 bits equivalent positioning accuracy. An external multi-turn pot and a bit of additional gearing should boost this by about three bits. Almost certainly the backlash in the gearing will eat this addition, but it will be fun to try.Mikewww.mikesflightdeck.com

Ah, availability. That's too bad. However, the circuit you describe sounds interesting. I've heard about those server mods for making them into speed-controlled motors. Looks like you've got plenty of things to tinker with. :-) I definitely want to implement some of these things at some point. Meanwhile, I need to get my panel built and I'll be using monitors behind them as a substitute for physical gauges.Peter http://bfu.avsim.net/sigpics/PeterR.gifBFU Forums Moderator