Craig Sawyers

Justin has taken most (or all) of the parts off his site, and only lists finished products. That does not mean that he does not have them, and that they are available for purchase if asked. Either PM him, or contact him via his site. I've bought several DACTs off him after he took them off his site.

Holy shit. I can't get my head around what you must be going through right now. Just hang on in there - you'll turn the corner pretty soon.

And just to add to eveyone's 2SC3675 paranoia, Sanyo are disconintuing the part: SANYO Transistor: 2SC3675 Only 1200 left in Sanyo's inventory. After that - nothing. Anyone in the US up for a group buy of 200 or more (200 is the MOQ) at 86c each? You could certainly count me in for a T2's worth of them (34) plus a bunch of spares.

Bloody hell. Your and my experience with this should sound warning klaxons to all T2 builders - measure any 2SC3675 that you are planning on using!

The 2SC3675 problem is solved. I hauled my Tek 577 curve tracer out with the aim of measuring the hFE of my batch of 2SC4686A. The answer was 25 +/-10%. So I thought I'd try out a 2SC3675 just for comparison. The one I connected up gave the answer ... 10. Say what? So I grabbed another and this was was 70. Another was 10. Then I spotted that there were two batches that looked quite different. One lot has an hFE of 10, the other was in the range 50 to 70. A picture of the two types is here http://www.tech-enterprise.com/tekstuff/2SC3675.jpg . If you zoom in you will see the markings on the packages. The one on the left gives the correct high value hFE, and the one on the left has a puny hFE of 10. All remaining ten samples of the right hand one have a gain of 10. The remaining four of the left hand one have gains of 50 to 70. Of course I have a random selection stuffed into my T2. And the ones that have failed in the batteries are (surprise surprise) the low gain ones. In forty years of designing and building electronics, this is a first for me. The right hand ones in the picture are either a faulty batch, or counterfeit. The duff ones are marked iSC, and may come from this outfit isc,inchange Semiconductor Company Limited Welcome You , and clearly Chinese crap. I'm going to order another batch of these sods, and check every last one before I replace the defective ones in my T2 lock stock and barrel.

Yeah - I can see that snow in Texas is a bit of a rare novelty. But in the UK we get snow on a pretty regular basis. Hell, Cairngorm in Scotland is a ski resort! Now sure, Scotland gets a lot more snow than here in the South (even though only a very few hundred miles separate us), but even there they have ground majestically to a halt - people stranded in their cars for ten hours on the motorways, quite a few fatal car accidents and so forth. This all happened 12 months ago - and their excuse was that it was freak weather which was not forecast. This year all this was predicted from climate modelling a week in advance. And yet again we have not got our act together. In just-OK-is-good-enough Britain.

Everywhere on the planet copes with snow better than the UK. With snow forecast for today in our area, we set off early to pick up the Christmas Tree (from the Christmas Tree Barn at Christmas Common! Seriously). -4C but no snow to start. There'd been a bit of snow during the night as we drove further. Collect tree and then it started to snow big time. Half an hour to drive there, hour and a half to drive back. With snow forcecast for days, we saw no sign of ploughs or gritting. Airports closed throughout Britain, chaos everywhere on the roads. You know how many time Helsinki airport in Finland has been closed in the last ten years? None. In the UK we get 6" of snow and we close the lot. Pathetic.

Well that was done under section 44 of the terrorism act (arbitrary stop and seach powers, regardless of whether there was reasonable suspicion of terrorist activities), which the European Court of Human Rights overturned earlier this year as unlawful. The current legal position is this: Metropolitan Police Service - About the Met - Photography advice . "Members of the public and the media do not need a permit to film or photograph in public places and police have no power to stop them filming or photographing incidents or police personnel" "Officers do not have the power to delete digital images or destroy film at any point during a search. Deletion or destruction may only take place following seizure if there is a lawful power (such as a court order) that permits such deletion or destruction" Which is all pretty clear. But lets not forget that the Metropolitan Police have a reputation rather lower than a snake's belly. In 1979 they killed Blair Peach with a truncheon blow at a protest. Last year they killed Ian Tomlinson at the G20 protest last year, and the officers had (illegally) taken off their identification badges - so the mobile phone records taken of the incident were not able to ID the officers. And most recently (last couple of weeks) they clubbed a student protester called Alfie Meadows so hard that he had to have brain surgery to save his life. Not forgetting Jean Charles de Menezes. It does seem to be positively draconian and downright dangerous in the US though - doing some searching about, a British guy on vacation recently came perilously close to getting shot when he took some photos of the Pentagon.

I find all this bewildering. If you were going to take clandestine photos (whatever that means), you would not use an expensive and physically obvious camera, you'd use a phone. Sure we did not have anything of the scale of 9/11 in the UK, but we had a bomb campaign on the London Underground and on London buses four or five years ago that killed many. Then there was a failed bombing attempt on the same theme shortly afterwards (incompetent bombers, luckily). That caused a bad mistake by the British SWAT guys when they incorrectly identified a Brazilian student called Jean Charles de Menezes as a terrorist bomber and shot the guy dead on a crowded tube train. They shot him 7 times in the head at point blank range just to make sure. There was an outcry about that, not surprisingly, and significant indications that the police supressed and even deleted evidence. And we had the whole Ireland troubles that killed thousands by the bomb and the bullet from the 70's to 90's, including many on the British mainland. They even launched mortars at 10 Downing Street. But there'd be rioting in the street if we had a policy of camera confiscation when people were quietly going about their hobby (or profession) taking photos of a tree.

Um - Mouser returns a No Results for 598-CDV19CF050D03F

Pics, Reks, we want to see the gory pics

That is bloody outrageous. I'd stir up such a shit storm if that happened to me the campus gestapo would be looking for alternative employment. I'm appalled.

No idea what shoulder washers for the Beta22 are like, so can offer no opinion. But if you check the thermal resistance of the 4171G you will find it is the same as for thermasil. Alumina has actually got way more thermal conductivity than thermasil, but the 4171G 2mm thick so the numbers balance out. The main thing is that if you draw things out on a piece of paper, you'll quickly see that the breakdown distance is the thickness of the washer, irrespective of the shoulder washer bush length. So although thermasil might have a breakdown potential of 1kV or more, the *actual* breakdown around the hole in the insulator is only a couple of hundred volts. The only way to make use of the native breakdown potenial is to use insulators without a hole, and use a spring clip to hold the transistor at the right pressure. The 4171G has 2mm, which is good for a couple of kV. Don't forget to use heatsink grease. The only screw technology that I know of that will take the torque needed for optimum heat transfer (1.1Nm for TO220 and 0.4Nm for TO126) is steel.

Not unless you use the right grade of stainless....

Having been down this track with the T2, what you need is an insulating bush that goes through the transistor tab, and then well into the 4171G. If it is a short bush, the transistor tab isolation is compromised. You have to draw a cross section picture to see this clearly. The ones that work well are the Aavid 7721-3PPSG. These can be a tight fit in some TO220 holes, but the bush length is 1/8" (3.18mm), which is more than enough with conductive (steel) screws and nuts. I prefer steel fixings because you can get the torque required - I tried plastic screws, but they always strip at a small fraction of the recommended TO220 package fixing torque. Future have 675 in reserve stock and 14,000 on order.

Say what? You're in Florida - or am I missing something about the Florida climate? Was in Copenhagen over the weekend visiting relatives. Minus 5C with a stiff wind straight off the sea, so was -20C with wind chill. Even the little mermaid looked frozen.

Jeez! That does not sound good, dude. Let us know how the surgery goes - sounds like plates to hold it together from what you say. Euch.

That is very helpful information. The low hfe of the 2SC4686A is certainly a concern, but I need to find a robust solution for the fact that my 2SC3675's die through second breakdown. So I either need to use a darlington configuration, or a higher current in the differential pair (by reducing the emitter resistor). Not very original, of course - but in the batteries I'm not massively concerned.

All you'll be listening to is the sound of the pot. I have no scientific rationale for why these sound just as awful as they do - tearing one apart the design principles and construction look superb. But they sound dreadful. For a more sane price than the RK50, go for a DACT - knocks the RK27 into a cocked hat.

There was a time in a land far far away that I could have understood that - but looking up Chinese Remainder Theorem made me realise that too much water had passed under the bridge to understand the first damned thing. But number theory was never one of my strong suits - I was always much too much a mathematical physicist. Back when the earth was young.

Um - nope. It means number 5 (an imperial size), 40 threads per inch, 0.25" thread length hexagonal cap head. Alas for those of us in metric countries, you need to source the right screws - the ones that go into tapped holes in the case have to absolutely right. Search the thread for European sources.

Is it me having a senior moment - or has there been a sort of mass change in people's avatars?

A couple of days ago (Sunday) went to see Meat Loaf at Birminham UK. Voice was a bit creaky, but that was probably a few decades of bodily misuse. Otherwise awesome.

That is not what I said - 60Hz transformers will always work with 50Hz. The problem arises because the primary inductance is fixed, so the primary current has a 1/f dependence. If you are trying to design a transformer that effectively utilises the core at 60Hz, you push the core with 20% more flux when using it at 50Hz. So a 50Hz transformer needs either more primary turns (to keep the NI product the same) or 20% more core. Which means that a 60Hz optimised transformer (doesn't matter whether EI or toroid) invariably hums when the primaries are connected for use on a 50Hz supply. My supplier for the T2 transformers, Paul Houlden, can and will specifically redesign and optimise the toroid for 50Hz (if he knows the spec). He also uses oversize cores (within space constraints) to reduce mechanical noise to imperceptible levels.

Future Electronics have 60 in stock. No problems with where they are supplied to. Craig