Stentor Design Questions

[Albert H]

Unfortunately, your 4 Watt resistors will probably be wirewound types, so will be very inductive and therefore useless as an RF load. Your ideas about resistors in parallel are a bit wrong too - the formula is:

1/R1 + 1/R2 + 1/R3......etc = 1/Rtotal

This means that two equal resistors in parallel gives half the value, three equal resistors give one third of the value and so on. It's best to use the same power rating for all the resistors. Years ago, I made my dummy load from 1k resistors just like the one on this page:

http://www.k4eaa.com/dummy.html

Mine was in a "Golden Syrup" tin and I used it for years. I think that I might still have it somewhere. One well-known rig builder in South London filled his dummy load with chip oil to dissipate the heat. Unfortunately, it was second-hand oil and every time he fired up a rig, his workshop was filled with chip shop smells!

[mountainman]

Built radium's design. The bigger coil diameters made the taps a little easier to deal with. Really hoping this 1033 JFET is going to work in place of the BF245.....

IMG_20170930_202923.jpg

IMG_20170930_202912.jpg

I'll try to bang together a dummy load here in the next few minutes and give this lil guy a try. I also have a 12v 20W light bulb here which measures 2 ohms across the contacts. I thought I remembered people using light bulbs as dummy loads and using the brightness of the bulb as a signal strength indicator? They may have been larger bulbs though. I'm gonna measure a few more incandescent bulbs i have laying around.

[mountainman]

Got it going. Not sure on output frequency yet but going into the dummy load, I'm definitely getting 5w+. I built about the ugliest DL ever, (10) 56 ohm 1w resistors in parallel on a PCB split in half.

IMG_20170930_223012.jpg

First plugging it in I was showing about 1w, turning some caps brought it up to 5-7w, but the MRF237 heated up quickly after the tuning so I shut it off.

For those that don't notice, the power meter is a cheap CB unit with some 259 to BNC adapters on it. The wattage is approximate, but the meter at least shows increases and decreases in power. Need to make a WAY better heat sink and/or just find a happy medium between power output and heat from the final. Glad to see some watts coming out, hopefully it's close to on frequency.

[MiXiN]

10 x 56r Resistors in parallel will be about 5.6 Ohms or thereabouts, so well out from 50 ohms.

Make sure you use a better heatsink on the MRF237 or the thing will fry.

Well done on getting it going.

[mountainman]

Sorry that was 10x 560r. Measured it at 55.9ohm. Now my next issue is I'm transmitting at 115.2MHz with the pot turned all the way down. Started at 133MHz and starting smushing coils together to get down to where I am now.

Not sure where to go from here to get down in the 90s.

[Maximus]

You could try reducing the diameter of the coils. They look a bit big


Sent from my iPhone using Tapatalk

[MiXiN]

Maximus is right.

The Oscillator coil in particular looks far too big in diameter. Is it about 10mm or more?

If I remember rightly, mine were something like 5.5mm and 6.5mm and they landed me around 95Mhz.

[mountainman]

Maximus is right.

The Oscillator coil in particular looks far too big in diameter. Is it about 10mm or more?

If I remember rightly, mine were something like 5.5mm and 6.5mm and they landed me around 95Mhz.

I was trying to do them with a 8mm bolt per the schematic radium posted but instead used a 3/8" drill bit which works out to 9.5mm. I will re wind and re tap the coils.

[mountainman]

I may or may not have supplied opposite polarity power to the board. I rewound/installed the first coil and then set about buillding and external switch and power supply outlet to install into its box when built. After getting the power situation figured out, I am now getting zero watts out and the MRF is getting warm very quickly.

Did I fry the final?

[mountainman]

Disregard, finished re winding the second tapped coil and put back together. The tansmitter will go from 500mw to around 7w depending on where the caps are set. Now X-mitting at 96.1MHz with the pot centered. THe audio test earlier indicated that Bass is not something this transmits cleanly. It sounds great on highs and mids but bass sounds like static. Is this where the lpf would help out?

[Albert H]

Got it going. I built about the ugliest DL ever, (10) 56 ohm 1w resistors in parallel on a PCB split in half.

IMG_20170930_223012.jpg

You've built a 5.6Ω dummy load! No wonder the MRF 237 got hot - it's working into (almost) a short! Your light bulb that reads 2Ω when cold will have a much higher resistance when hot, but the in-rush of current would probably destroy your output transistor - that's why light bulbs make very poor dummy loads!

You also need a much bigger heatsink on the MRF237! It's an old fashioned device (from the late 70s) so is inefficient compared to modern devices. You really must get rid of the heat if you want to save it from an early death! When a conventional transistor like the '237 gets hot, it draws more current, so it gets hotter, draws more current, so it gets hotter.....etc until it melts internally!

One thing I notice about your coils - the wire is rather thin. I usually used 18 or even 16 SWG for coils in circuits like that. Remember - it's a free-running exciter, and adjusting anything will affect the operating frequency. It's essential to have rigid coils that aren't microphonic or likely to bend at the slightest touch! In a pinch, I'd use stripped twin-and-earth mains cable cores. They're copper and easy to solder!

[Albert H]

Disregard, finished re winding the second tapped coil and put back together. The tansmitter will go from 500mw to around 7w depending on where the caps are set. Now X-mitting at 96.1MHz with the pot centered. THe audio test earlier indicated that Bass is not something this transmits cleanly. It sounds great on highs and mids but bass sounds like static. Is this where the lpf would help out?

No. The lowpass filter is to get rid of the multiples of the output frequency - you really don't want to be radiating 192.2MHz (in the DAB band) and 288.3MHz (in the military aviation band) or 384.4MHz.......etc. The RF filter will get rid of these multiples (which would get you quickly nicked if you radiated them), but it won't affect the way the rig sounds.

The modulation is a slight problem, since there's no pre-emphasis provided. Here's a simple circuit:

Mono-pre-emphasis.png

Remember - you won't be able to set the modulation as loud as the commercial stations, because you don't have compression and limiting in your audio path. The pre-emphasis will make it sound a bit better, because you'll turn the mod down (to the point that the bass is undistorted) and the top end will sound right. Any low noise NPN transistor can be used - 2N5088, 2N3904, BC109C, BC183L etc... Just remember to check the pin-out of the device before soldering it in!

If you put this little circuit before the audio input of your board, you'll prevent any change in frequency when you plug in your audio source - just the audio lead could easily change the frequency of the rig!

[mountainman]

Got it going. I built about the ugliest DL ever, (10) 56 ohm 1w resistors in parallel on a PCB split in half.

IMG_20170930_223012.jpg

You've built a 5.6Ω dummy load! No wonder the MRF 237 got hot - it's working into (almost) a short! Your light bulb that reads 2Ω when cold will have a much higher resistance when hot, but the in-rush of current would probably destroy your output transistor - that's why light bulbs make very poor dummy loads!

You also need a much bigger heatsink on the MRF237! It's an old fashioned device (from the late 70s) so is inefficient compared to modern devices. You really must get rid of the heat if you want to save it from an early death! When a conventional transistor like the '237 gets hot, it draws more current, so it gets hotter, draws more current, so it gets hotter.....etc until it melts internally!

One thing I notice about your coils - the wire is rather thin. I usually used 18 or even 16 SWG for coils in circuits like that. Remember - it's a free-running exciter, and adjusting anything will affect the operating frequency. It's essential to have rigid coils that aren't microphonic or likely to bend at the slightest touch! In a pinch, I'd use stripped twin-and-earth mains cable cores. They're copper and easy to solder!

That was a typo:

Sorry that was 10x 560r. Measured it at 55.9ohm. Now my next issue is I'm transmitting at 115.2MHz with the pot turned all the way down. Started at 133MHz and starting smushing coils together to get down to where I am now.

Not sure where to go from here to get down in the 90s.

New update:

Added this LPF:

http://electronics-diy.com/electronic_s ... php?id=150

Now I'm only getting 62.6mhz out. I had to sub the 22pf for 20 pf and the 51pf for 56pf.

I think I'M cutting out the 96.1mhz that I was getting out before....

[mountainman]

Got it going. I built about the ugliest DL ever, (10) 56 ohm 1w resistors in parallel on a PCB split in half.

IMG_20170930_223012.jpg

One thing I notice about your coils - the wire is rather thin. I usually used 18 or even 16 SWG for coils in circuits like that. Remember - it's a free-running exciter, and adjusting anything will affect the operating frequency. It's essential to have rigid coils that aren't microphonic or likely to bend at the slightest touch! In a pinch, I'd use stripped twin-and-earth mains cable cores. They're copper and easy to solder!

Your right, the coils are made out of 22 gauge because that what I had on hand in enameled wire. I suppose making it out of naked heavier copper wire would work better assuming I don't have any coils touching?

[mountainman]

IMG_20171001_203231.jpg

Should I be running RCA or some other type of coax from the stentor to the lpf?

[Shedbuilt]

Increasing the diameter of an air cored inductor, increases its inductance. Increasing the inductance of an oscillator tank circuit coil, reduces the frequency of oscillation. For a 6 turn coil, 12mm long, increasing former diameter, from 6mm to 10mm, will (if my maths is correct), more than double its inductance. This alone, is enough to push the frequency of a 90 odd MHz oscillator, down into the 60 odd MHz range. So I think the frequency you are now "seeing", is most likely the real one; meaning that - in my opinion, you were quite likely picking up a harmonic before.

[Maximus]

I’d use some screened 50 ohm RG178 coax or similar from the tx to the lpf. You could use the cheaper RG58 coax but it’s trickier to bend due to the thickness.


Sent from my iPhone using Tapatalk

[MiXiN]

OP,

Have you got this running properly & on frequency yet?

[mountainman]

I just got the thin coax in the mail to re attempt building the LPF. I need to re-coil it with some thicker wires but I am currently away from the bench for the next week or so....