Anyone know anything about this?

[MiXiN]

Dutchrfshop has decent cheap heatsinks - get it while you can because you all gonna cry when brexit kicks in...

Code: Select all

http://dutchrfshop.nl/koelmateriaal.html

Forgot about Dutch RF Shop, so thanks for the reminder mate.

I'll get an order in before Freddo Chocolates hit £87. Lol.

[Albert H]

Don't panic. The exchange rates will be better next week. All the bollocks will have died down by then.

[teckniqs]

I found out an interesting tip for buying from China on eBay. If you buy on eBay.CA (Canada) you can usually find the same thing at the same price in dollars as it is in pounds on UK eBay and save yourself some cash.

[shuffy]

Sorry I've been away for a week, you might have ordered already. I don't know what the delivery costs would be from Netherlands and you're only running 40W.

I think I've used both of these with 40W amps, I've certainly used the second one (1K/W), I've got one here. Either with a small fan should do, they're not bad value, and free delivery at the moment.

http://uk.rs-online.com/web/p/heatsinks/9033166/
http://uk.rs-online.com/web/p/heatsinks/5075194/

[MiXiN]

Sorry I've been away for a week, you might have ordered already. I don't know what the delivery costs would be from Netherlands and you're only running 40W.

I think I've used both of these with 40W amps, I've certainly used the second one (1K/W), I've got one here. Either with a small fan should do, they're not bad value, and free delivery at the moment.

http://uk.rs-online.com/web/p/heatsinks/9033166/
http://uk.rs-online.com/web/p/heatsinks/5075194/

Alright shuffy,

I didn't get around to buying from Dutch RF shop, so I bought the one you quoted on the bottom link with the channel.

My metal to connect an heatsink to is around 5.5 inches in length and 20mm in height so the heatsink won't quite cover all the metal length ways but it should be ample.

Nice one for the link mate.

[Albert H]

Years ago, I spotted some aluminium strips in a skip. These were about 4cm wide and formed as a U-shaped channel. They were used as frames for temporary room dividers (for offices) and when they were no longer wanted, they were discarded. With a bit of effort with a big hacksaw and a bit of bolting together, we made big, four-finned heatsinks. They were perfect for use with the 28V stud-mount transistors we used in those days (the 2N3375 into a 587BLY was a favourite 90 Watt+ combination!). My PA board was modified to fit against these homebrewed heatsinks, and we made dozens of them - 2N3866, 2N3375, 587BLY - roughly 100mW input for 100W output at 30V - which had fixed, home-made mica capacitors instead of trimmers. These PAs could be built for about £12 a go! The power supply used to cost more than the rest of the rig!

[shuffy]

My metal to connect an heatsink to is around 5.5 inches in length and 20mm in height so the heatsink won't quite cover all the metal length ways but it should be ample.
Nice one for the link mate.

You're welcome. I've used that heatsink on similar amps to yours and depending on your application you might need a fan to keep things cool, although I've got away with it when mounting the heatsink on the outside of the case. Here's the one I was talking about earlier

heatsink.jpg

I've seen amplifiers of Stephen's with that piece of metal on them before, I don't know if yours will be a help or a hindrance but obviously the bottom line is to have the best thermal path from your transistor through the heatsink material. In the amp in the photo there's also a 1971 as driver on the left which is mounted flat against the heatsink, but the main device is just mounted through a hole between the fins - I don't tend to use any kind of spreader if I can get away with it. It's slightly offset from the middle so that the heatsink and PCB will fit with the box design but closest to the middle is best. Good luck

[shuffy]

Years ago, I spotted some aluminium strips in a skip.

I know it wasn't you (and I think we both know who it was) but do you remember the "saucepan" transmitter from the early 90s? And the Eric Gotts "water tank" incident based on a similar principle... now I suspect that might have been something to do with you...

[teckniqs]

i'm sure I recall him mentioning it.....

[Albert H]

Guilty as charged! I never used a saucepan (except for cooking my dinner) but I did self-tap a PA to the underside of a water tank up a west London block!

I also used the inside of the aerial mast to mount a concealed PA a few times. They were tricky to build (and a bugger to tune up), but I drilled access holes for the critical trimmers which were tweaked after the whole mess was assembled. The holes were then covered up with that thin "silver foil" self-adhesive tape, and were invisible after a day or two up the block, and the transistor studs just looked like part of the construction of the mast.

The RF ran up the inside of the mast from the exciter (usually about 100mW, up RG58, using the braid for the ground connection). There was a +30V lead taped to the coax, though for a few of them I found some 75Ω RF coax with twin inners, which was very convenient. The aerial was also fed up the inside of the mast, so it stopped certain people putting drawing pins in the coax!

Some rigs had the 100mW exciter feeding "through" a dummy PA (usually made with transistors blown up during PA development) with a small, throwaway heatsink. In fact, the RF passed beneath the "PA" board and the same came out as was put in. Sometimes there was a working LED on the "PA" just to lend further credence to it working!

The exciters we used in those days used a Plessey ÷100 prescaler (SP2689) and a 145106 PLL chip, a ½f oscillator, buffer, two-transistor doubler and amplifier, all on a board 3½" × 1½". They gave about 100 - 150mW out according to the supply voltage and were clean enough on their own for a little "round the block" rig. They were really cheap and quick to build. There was also an "at frequency" version (that had to be very carefully screened so that it didn't "see" the PA) that used even fewer parts!

[MiXiN]

My metal to connect an heatsink to is around 5.5 inches in length and 20mm in height so the heatsink won't quite cover all the metal length ways but it should be ample.
Nice one for the link mate.

You're welcome. I've used that heatsink on similar amps to yours and depending on your application you might need a fan to keep things cool, although I've got away with it when mounting the heatsink on the outside of the case. Here's the one I was talking about earlier

heatsink.jpg

I've seen amplifiers of Stephen's with that piece of metal on them before, I don't know if yours will be a help or a hindrance but obviously the bottom line is to have the best thermal path from your transistor through the heatsink material. In the amp in the photo there's also a 1971 as driver on the left which is mounted flat against the heatsink, but the main device is just mounted through a hole between the fins - I don't tend to use any kind of spreader if I can get away with it. It's slightly offset from the middle so that the heatsink and PCB will fit with the box design but closest to the middle is best. Good luck

Hi Shuffy,

The heatsink arrived from RS; The first arrived slightly damaged, so they sent a new one free of charge. What a great company and all sent free delivery via Parcelforce next day.

If you check the picture of the NRG TX, there's a piece of copper foil - and I'm wondering if its original intention was as some thermal interface pad? Should I leave this on and add some Arctic Silver heatsink compound between this and the RS heatsink?

20160706_131817.jpg

Not sure on whether to connect the heatsink as shown below (horizontal) making use of the wide flat channel, have it vertically mounted, again making use of the wide channel, or completely remove all the Aluminium and connect the RS heatsink directly to the MRF240?


Ideally, for nostalgic reasons, I'd like to keep the TX looking original - but whichever is the best way to keep the final cool is ultimately best.

20160706_131843.jpg

Sent from my SM-G900F using Tapatalk

[shuffy]

The heatsink arrived from RS; The first arrived slightly damaged, so they sent a new one free of charge. What a great company and all sent free delivery via Parcelforce next day

For stuff that's not too hard to get and where you can't get an equivalent product cheaper elsewhere, I love RS and CPC. They courier everything - if you order a single resistor on your trade account, they'll send it round in a van. (Yes, that's true; no, it wasn't me!)

If you check the picture of the NRG TX, there's a piece of copper foil - and I'm wondering if its original intention was as some thermal interface pad? Should I leave this on and add some Arctic Silver heatsink compound between this and the RS heatsink?

No - this won't do anything. I still think that's an earth strap to attach it to something else.

Not sure on whether to connect the heatsink as shown below (horizontal) making use of the wide flat channel, have it vertically mounted, again making use of the wide channel, or completely remove all the Aluminium and connect the RS heatsink directly to the MRF240?

Like I said, it's all about getting the heat away from the transistor as effectively as possible. I can see where you're coming from the wanting to keep the look of the thing though. Mounting bits of metal against each other isn't necessarily going to give you the best heat transfer away from the transistor, I think it would be better to discard the piece of metal and bolt the transistor directly through the new heatsink.

I don't know how thick that bit of Aluminium is, and I wouldn't do this myself, but if you've enough of the transistor stud to get through both the metal and the heatsink, and the rest of the metal will mount flush against the heatsink, you could maybe try that - you need to make good flat contact between the metal and the heatsink and use some transfer compound where the two pieces are held tightest together, most likely around the transistor stud. I can see there are some screws through the pcb into the bit of metal with spacers on them so I doubt it will go flush against the new heatsink.

[Albert H]

I used to tap the transistor mounting holes in my heatsinks, so that the transistors could bolt straight into the heatsink - with maximum heat transfer - and I didn't need the securing nut.

[shuffy]

I used to tap the transistor mounting holes in my heatsinks, so that the transistors could bolt straight into the heatsink - with maximum heat transfer - and I didn't need the securing nut.

Yes but how do you get your PCB to line up squarely with the heatsink when the transistor is flush against the metal? Or are we talking about the stud mounting transistors with legs!!

[Albert H]

The trick was to get the thread to start in the right place. I have a tapping jig that would ensure this. I was surprised at the improvement in heat transfer, when I measured it. I used "Redpoint" heatsinks, which had about a centimetre of aluminium on the base, and the fins were thicker than the usual too. I could frequently get away without using a fan - useful when a rig was being concealed on a rooftop and the noise of a fan would give it away!

Hiding rigs was an art - aerials that looked like CB twigs, cables hidden down wall cavities or ducts, mains supplies out of the back of mains sockets, link aerials that were old TV aerials, rigs hidden in dry risers or hydrant housings..... The "authorities" who were meant to take the gear away when they'd found it would usually locate it (with some difficulty), but rig thieves almost never got my gear. There were also some unusual counter-measures taken to make removing my gear very unpleasant!

The big trick when building gear for illicit broadcasting was to make it as technically good as it could possibly be - stable, spectrally pure, correctly modulated and so on - but as cheap as possible. I frequently used an output device that I could get for under £3 each, but would give over 100 Watts with a little more than 10dB gain. The driver transistor used was 75p, and the exciter (including the PLL) would cost around £10. The most expensive part was usually the mains transformer. My rigs were always within 1kHz of the required frequency, had properly limited modulation, and had no in-band spurs and better than -70dBc harmonic suppression, and no conducted RF down the mains supply. They were frequently better than commercial broadcast rigs, but were made as cheaply as possible so that their "loss" wasn't too hard to bear! Link receivers were based on TV receiver front-ends and an IF strip that was based on a cheap, widely used IC. They cost about £8 to build. We used to spend more on coax than anything else!

[shuffy]

I can see your motivation for this regarding the fan. I'm a bit of a metal basher and frequently use taps and dies but never imagined I'd be able to get a thread to start in the right place so I could screw the transistor through a heatsink and have it line up. I've sometimes got other devices on the same board which need bolting down (my BLW60 amp above is a case in point) or need spacers for various reasons, which might mean all bets are off for your threaded hole for the PA...

The big trick when building gear for illicit broadcasting was to make it as technically good as it could possibly be - stable, spectrally pure, correctly modulated and so on - but as cheap as possible.

That's a big piece of the fascination for me. Actually I was having pretty much this conversation with somebody else last week. The challenge is quality, price, difficulty of detection and the satisfaction that you've designed and built it yourself. Panders to the ego just like playing records on air and if you also like music and some mild peril... bargain!

[Albert H]

My PCB PA also had the board bolted to the heatsink, with spacers and tapped holes in the 'sink. On that series of PAs, the heatsink was the roof of the box. The rest of the box was folded aluminium.

[shuffy]

Well Albert I'm just not worthy!

[Albert H]

Shuffy - it was the cheapest way to get good heat transfer, a good solid housing and screening. The exciter used to go into a little box 3½" × 2" × 1½" deep, with the two or three PLL ICs "dead-bugged" to the underside of the PCB. I didn't let on that they were PLL controlled, and several people were absolutely convinced that I knew the secret of the truly stable VFO!

I used to attach the exciter to the side of the link receiver box, and the whole lot - including the PSU - fitted nicely inside a large plastic waste paper bin (with a lid) that could be bought from Staples or Rymans for about a fiver.

The more sophisticated version was assembled into a Pye base station case, and if it was a VHF Pye base version, the SD1147 / SD1019 28V PA and its associated PSU was "re-purposed" for between 50 and 120 Watts output. At the lower end of the power range, it didn't need a blower!

[rigmo]

OK guys, good news, but bad temperatures.

I was going to use a 2N4427 for missing Transistor TR5 as suggested but have misplaced them so used a 2N3553 instead, and an 18pF Capacitor for C28.

I'm getting 24 Watts at 13.8V, but TR5 (2N3553) is too hot to touch for more than a few seconds, and the MRF240 is inadequately heatsinked for this sort of power. The heatsink is only about 7 x 1 x 1 inch - length, width, and height.

Apart from dropping the supply voltage, can anyone suggest a solution to excessively reducing the power and the high temperatures?

depend of matching a lot... if is not match overheating is easy and rapidly fast on lower power... ...